* [PATCH v1 5/7] MIPS: BCM63XX: Remove one time use ephy_reset_gpio_flags
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
In-Reply-To: <20260715185112.1323510-1-andriy.shevchenko@linux.intel.com>
ephy_reset_gpio_flags exists solely to supply polarity to the reset GPIO.
But in practice it's kept all the same and currently only a single user
present. Drop the member and use hard coded GPIO flags instead.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
arch/mips/bcm63xx/boards/board_bcm963xx.c | 7 +++----
arch/mips/include/asm/mach-bcm63xx/board_bcm963xx.h | 3 ---
2 files changed, 3 insertions(+), 7 deletions(-)
diff --git a/arch/mips/bcm63xx/boards/board_bcm963xx.c b/arch/mips/bcm63xx/boards/board_bcm963xx.c
index c5617b889b1c..57e916c06792 100644
--- a/arch/mips/bcm63xx/boards/board_bcm963xx.c
+++ b/arch/mips/bcm63xx/boards/board_bcm963xx.c
@@ -42,7 +42,7 @@ static struct board_info __initdata board_cvg834g = {
.expected_cpu_id = 0x3368,
.ephy_reset_gpio = 36,
- .ephy_reset_gpio_flags = GPIOF_OUT_INIT_HIGH,
+
.has_pci = 1,
.has_uart0 = 1,
.has_uart1 = 1,
@@ -903,9 +903,8 @@ int __init board_register_devices(void)
platform_device_register(&bcm63xx_gpio_leds);
- if (board.ephy_reset_gpio && board.ephy_reset_gpio_flags)
- gpio_request_one(board.ephy_reset_gpio,
- board.ephy_reset_gpio_flags, "ephy-reset");
+ if (board.ephy_reset_gpio)
+ gpio_request_one(board.ephy_reset_gpio, GPIOF_OUT_INIT_HIGH, "ephy-reset");
return 0;
}
diff --git a/arch/mips/include/asm/mach-bcm63xx/board_bcm963xx.h b/arch/mips/include/asm/mach-bcm63xx/board_bcm963xx.h
index 830f53f28e3f..018a56cbc328 100644
--- a/arch/mips/include/asm/mach-bcm63xx/board_bcm963xx.h
+++ b/arch/mips/include/asm/mach-bcm63xx/board_bcm963xx.h
@@ -46,9 +46,6 @@ struct board_info {
/* External PHY reset GPIO */
unsigned int ephy_reset_gpio;
-
- /* External PHY reset GPIO flags from gpio.h */
- unsigned long ephy_reset_gpio_flags;
};
#endif /* ! BOARD_BCM963XX_H_ */
--
2.50.1
^ permalink raw reply related
* [PATCH v1 2/7] media: em28xx: Split em28xx_pctv_290e_set_lna_gpio() helper
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
In-Reply-To: <20260715185112.1323510-1-andriy.shevchenko@linux.intel.com>
Split em28xx_pctv_290e_set_lna_gpio() helper in order to deduplicate
the GPIO request and setting.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
drivers/media/usb/em28xx/em28xx-dvb.c | 31 ++++++++++++---------------
1 file changed, 14 insertions(+), 17 deletions(-)
diff --git a/drivers/media/usb/em28xx/em28xx-dvb.c b/drivers/media/usb/em28xx/em28xx-dvb.c
index 8482fc4045ea..ef148a27c94d 100644
--- a/drivers/media/usb/em28xx/em28xx-dvb.c
+++ b/drivers/media/usb/em28xx/em28xx-dvb.c
@@ -737,17 +737,14 @@ static void pctv_520e_init(struct em28xx *dev)
regs[i].r, regs[i].len);
};
-static int em28xx_pctv_290e_set_lna(struct dvb_frontend *fe)
+static int em28xx_pctv_290e_set_lna_gpio(struct em28xx *dev, u32 lna)
{
- struct dtv_frontend_properties *c = &fe->dtv_property_cache;
- struct em28xx_i2c_bus *i2c_bus = fe->dvb->priv;
- struct em28xx *dev = i2c_bus->dev;
#ifdef CONFIG_GPIOLIB_LEGACY
struct em28xx_dvb *dvb = dev->dvb;
int ret;
unsigned long flags;
- if (c->lna == 1)
+ if (lna == 1)
flags = GPIOF_OUT_INIT_HIGH; /* enable LNA */
else
flags = GPIOF_OUT_INIT_LOW; /* disable LNA */
@@ -761,11 +758,20 @@ static int em28xx_pctv_290e_set_lna(struct dvb_frontend *fe)
return ret;
#else
dev_warn(&dev->intf->dev, "%s: LNA control is disabled (lna=%u)\n",
- KBUILD_MODNAME, c->lna);
+ KBUILD_MODNAME, lna);
return 0;
#endif
}
+static int em28xx_pctv_290e_set_lna(struct dvb_frontend *fe)
+{
+ struct dtv_frontend_properties *c = &fe->dtv_property_cache;
+ struct em28xx_i2c_bus *i2c_bus = fe->dvb->priv;
+ struct em28xx *dev = i2c_bus->dev;
+
+ return em28xx_pctv_290e_set_lna_gpio(dev, c->lna);
+}
+
static int em28xx_pctv_292e_set_lna(struct dvb_frontend *fe)
{
struct dtv_frontend_properties *c = &fe->dtv_property_cache;
@@ -1920,19 +1926,10 @@ static int em28xx_dvb_init(struct em28xx *dev)
goto out_free;
}
-#ifdef CONFIG_GPIOLIB_LEGACY
/* enable LNA for DVB-T, DVB-T2 and DVB-C */
- result = gpio_request_one(dvb->lna_gpio,
- GPIOF_OUT_INIT_LOW, NULL);
- if (result)
- dev_err(&dev->intf->dev,
- "gpio request failed %d\n",
- result);
- else
- gpio_free(dvb->lna_gpio);
+ em28xx_pctv_290e_set_lna_gpio(dev, 0);
- result = 0; /* continue even set LNA fails */
-#endif
+ /* continue even set LNA fails */
dvb->fe[0]->ops.set_lna = em28xx_pctv_290e_set_lna;
}
--
2.50.1
^ permalink raw reply related
* [PATCH v1 7/7] gpiolib: Get rid of not used anymore gpio_request_one()
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
In-Reply-To: <20260715185112.1323510-1-andriy.shevchenko@linux.intel.com>
No more users.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
Documentation/driver-api/gpio/consumer.rst | 2 +-
drivers/gpio/gpiolib-legacy.c | 31 ----------------------
include/linux/gpio/legacy.h | 8 ------
3 files changed, 1 insertion(+), 40 deletions(-)
diff --git a/Documentation/driver-api/gpio/consumer.rst b/Documentation/driver-api/gpio/consumer.rst
index bb3366047fad..03edd9472f07 100644
--- a/Documentation/driver-api/gpio/consumer.rst
+++ b/Documentation/driver-api/gpio/consumer.rst
@@ -463,7 +463,7 @@ and vice-versa::
The GPIO number returned by desc_to_gpio() can safely be used as a parameter of
the gpio\_*() functions for as long as the GPIO descriptor `desc` is not freed.
All the same, a GPIO number passed to gpio_to_desc() must first be properly
-acquired using e.g. gpio_request_one(), and the returned GPIO descriptor is only
+acquired using e.g. gpio_request(), and the returned GPIO descriptor is only
considered valid until that GPIO number is released using gpio_free().
Freeing a GPIO obtained by one API with the other API is forbidden and an
diff --git a/drivers/gpio/gpiolib-legacy.c b/drivers/gpio/gpiolib-legacy.c
index ef3f2ef30cf2..c2b04605b203 100644
--- a/drivers/gpio/gpiolib-legacy.c
+++ b/drivers/gpio/gpiolib-legacy.c
@@ -21,37 +21,6 @@ void gpio_free(unsigned gpio)
}
EXPORT_SYMBOL_GPL(gpio_free);
-/**
- * gpio_request_one - request a single GPIO with initial configuration
- * @gpio: the GPIO number
- * @flags: GPIO configuration as specified by GPIOF_*
- * @label: a literal description string of this GPIO
- *
- * **DEPRECATED** This function is deprecated and must not be used in new code.
- *
- * Returns:
- * 0 on success, or negative errno on failure.
- */
-int gpio_request_one(unsigned gpio, unsigned long flags, const char *label)
-{
- int err;
-
- err = gpio_request(gpio, label);
- if (err)
- return err;
-
- if (flags & GPIOF_IN)
- err = gpio_direction_input(gpio);
- else
- err = gpio_direction_output(gpio, !!(flags & GPIOF_OUT_INIT_HIGH));
-
- if (err)
- gpio_free(gpio);
-
- return err;
-}
-EXPORT_SYMBOL_GPL(gpio_request_one);
-
/*
* **DEPRECATED** This function is deprecated and must not be used in new code.
*/
diff --git a/include/linux/gpio/legacy.h b/include/linux/gpio/legacy.h
index 557ef635935e..d4273162ed13 100644
--- a/include/linux/gpio/legacy.h
+++ b/include/linux/gpio/legacy.h
@@ -82,8 +82,6 @@ static inline int gpio_to_irq(unsigned gpio)
return gpiod_to_irq(gpio_to_desc(gpio));
}
-int gpio_request_one(unsigned gpio, unsigned long flags, const char *label);
-
int devm_gpio_request_one(struct device *dev, unsigned gpio,
unsigned long flags, const char *label);
@@ -104,12 +102,6 @@ static inline int gpio_request(unsigned gpio, const char *label)
return -ENOSYS;
}
-static inline int gpio_request_one(unsigned gpio,
- unsigned long flags, const char *label)
-{
- return -ENOSYS;
-}
-
static inline void gpio_free(unsigned gpio)
{
might_sleep();
--
2.50.1
^ permalink raw reply related
* [PATCH v1 3/7] media: em28xx: Open code gpio_request_one()
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
In-Reply-To: <20260715185112.1323510-1-andriy.shevchenko@linux.intel.com>
Open code the gpio_request_one() call to be able to kill that API
in the follow changes.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
drivers/media/usb/em28xx/em28xx-dvb.c | 14 +++++---------
1 file changed, 5 insertions(+), 9 deletions(-)
diff --git a/drivers/media/usb/em28xx/em28xx-dvb.c b/drivers/media/usb/em28xx/em28xx-dvb.c
index ef148a27c94d..d7f8dcf27900 100644
--- a/drivers/media/usb/em28xx/em28xx-dvb.c
+++ b/drivers/media/usb/em28xx/em28xx-dvb.c
@@ -742,18 +742,14 @@ static int em28xx_pctv_290e_set_lna_gpio(struct em28xx *dev, u32 lna)
#ifdef CONFIG_GPIOLIB_LEGACY
struct em28xx_dvb *dvb = dev->dvb;
int ret;
- unsigned long flags;
- if (lna == 1)
- flags = GPIOF_OUT_INIT_HIGH; /* enable LNA */
- else
- flags = GPIOF_OUT_INIT_LOW; /* disable LNA */
-
- ret = gpio_request_one(dvb->lna_gpio, flags, NULL);
- if (ret)
+ ret = gpio_request(dvb->lna_gpio, NULL);
+ if (ret) {
dev_err(&dev->intf->dev, "gpio request failed %d\n", ret);
- else
+ } else {
+ gpio_direction_output(dvb->lna_gpio, !!(lna == 1));
gpio_free(dvb->lna_gpio);
+ }
return ret;
#else
--
2.50.1
^ permalink raw reply related
* [PATCH v1 1/7] ARM: pxa: spitz: Open code gpio_request_one()
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
In-Reply-To: <20260715185112.1323510-1-andriy.shevchenko@linux.intel.com>
Open code the gpio_request_one() call to be able to kill that API
in the follow changes.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
arch/arm/mach-pxa/spitz_pm.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/arch/arm/mach-pxa/spitz_pm.c b/arch/arm/mach-pxa/spitz_pm.c
index 03b4b347f11a..f46a3fb7e07c 100644
--- a/arch/arm/mach-pxa/spitz_pm.c
+++ b/arch/arm/mach-pxa/spitz_pm.c
@@ -108,7 +108,8 @@ static void spitz_presuspend(void)
PGSR2 |= GPIO_bit(SPITZ_GPIO_KEY_STROBE0);
pxa2xx_mfp_config(&gpio18_config, 1);
- gpio_request_one(18, GPIOF_OUT_INIT_HIGH, "Unknown");
+ gpio_request(18, "Unknown");
+ gpio_direction_output(18, 1);
gpio_free(18);
PRER = GPIO_bit(SPITZ_GPIO_KEY_INT);
--
2.50.1
^ permalink raw reply related
* [PATCH v1 6/7] MIPS: BCM63XX: Open code gpio_request_one()
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
In-Reply-To: <20260715185112.1323510-1-andriy.shevchenko@linux.intel.com>
Open code the gpio_request_one() call to be able to kill that API
in the follow changes.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
arch/mips/bcm63xx/boards/board_bcm963xx.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/arch/mips/bcm63xx/boards/board_bcm963xx.c b/arch/mips/bcm63xx/boards/board_bcm963xx.c
index 57e916c06792..993341604439 100644
--- a/arch/mips/bcm63xx/boards/board_bcm963xx.c
+++ b/arch/mips/bcm63xx/boards/board_bcm963xx.c
@@ -903,8 +903,11 @@ int __init board_register_devices(void)
platform_device_register(&bcm63xx_gpio_leds);
- if (board.ephy_reset_gpio)
- gpio_request_one(board.ephy_reset_gpio, GPIOF_OUT_INIT_HIGH, "ephy-reset");
+ if (board.ephy_reset_gpio) {
+ gpio_request(board.ephy_reset_gpio, "ephy-reset");
+ gpio_direction_output(board.ephy_reset_gpio, 1);
+ gpio_free(board.ephy_reset_gpio);
+ }
return 0;
}
--
2.50.1
^ permalink raw reply related
* [PATCH v1 4/7] MIPS: BCM47XX: Open code gpio_request_one()
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
In-Reply-To: <20260715185112.1323510-1-andriy.shevchenko@linux.intel.com>
Open code the gpio_request_one() call to be able to kill that API
in the follow changes.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
arch/mips/bcm47xx/workarounds.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/arch/mips/bcm47xx/workarounds.c b/arch/mips/bcm47xx/workarounds.c
index 745c6228eb2c..9148d2c219d1 100644
--- a/arch/mips/bcm47xx/workarounds.c
+++ b/arch/mips/bcm47xx/workarounds.c
@@ -9,11 +9,13 @@ static void __init bcm47xx_workarounds_enable_usb_power(int usb_power)
{
int err;
- err = gpio_request_one(usb_power, GPIOF_OUT_INIT_HIGH, "usb_power");
- if (err)
+ err = gpio_request(usb_power, "usb_power");
+ if (err) {
pr_err("Failed to request USB power gpio: %d\n", err);
- else
+ } else {
+ gpio_direction_output(usb_power, 1);
gpio_free(usb_power);
+ }
}
void __init bcm47xx_workarounds(void)
--
2.50.1
^ permalink raw reply related
* [PATCH v1 0/7] gpiolib: kill gpio_request_one()
From: Andy Shevchenko @ 2026-07-15 18:46 UTC (permalink / raw)
To: Andy Shevchenko, Bradford Love, linux-gpio, linux-doc,
linux-kernel, linux-arm-kernel, linux-mips, linux-media
Cc: Linus Walleij, Bartosz Golaszewski, Jonathan Corbet, Shuah Khan,
Daniel Mack, Haojian Zhuang, Robert Jarzmik, Russell King,
Hauke Mehrtens, Rafał Miłecki, Thomas Bogendoerfer,
Mauro Carvalho Chehab
This is the finilazation of the work announced in the commit 61e1fd2abca4
("gpiolib: legacy: Make sure we kill gpio_request_one() first"). With this
small series we kill gpio_request_one() and leave only two legacy calls
devm_gpio_request_one() and gpio_request(), where the managed one has only
a few users and high chances to be killed soon as well.
The series is supposed to go via GPIO tree (due to nature of the last patch)
with possible immutable branch for others. Please, ack.
Andy Shevchenko (7):
ARM: pxa: spitz: Open code gpio_request_one()
media: em28xx: Split em28xx_pctv_290e_set_lna_gpio() helper
media: em28xx: Open code gpio_request_one()
MIPS: BCM47XX: Open code gpio_request_one()
MIPS: BCM63XX: Remove one time use ephy_reset_gpio_flags
MIPS: BCM63XX: Open code gpio_request_one()
gpiolib: Get rid of not used anymore gpio_request_one()
Documentation/driver-api/gpio/consumer.rst | 2 +-
arch/arm/mach-pxa/spitz_pm.c | 3 +-
arch/mips/bcm47xx/workarounds.c | 8 ++-
arch/mips/bcm63xx/boards/board_bcm963xx.c | 10 ++--
.../include/asm/mach-bcm63xx/board_bcm963xx.h | 3 -
drivers/gpio/gpiolib-legacy.c | 31 ----------
drivers/media/usb/em28xx/em28xx-dvb.c | 57 ++++++++-----------
include/linux/gpio/legacy.h | 8 ---
8 files changed, 39 insertions(+), 83 deletions(-)
--
2.50.1
^ permalink raw reply
* [PATCH v9 8/8] drm/gpusvm: Use hmm_range_fault_unlocked_timeout() for range faults
From: Stanislav Kinsburskii @ 2026-07-15 18:16 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
Several GPU SVM paths take mmap_read_lock() only to call hmm_range_fault()
and open-code mmu interval sequence setup before each HMM walk. They also
retry -EBUSY until HMM_RANGE_DEFAULT_TIMEOUT expires.
Use hmm_range_fault_unlocked_timeout() for those faults. The HMM helper now
owns mmap_lock acquisition and refreshes range->notifier_seq for its
internal retries, while GPU SVM keeps its existing driver-lock validation
with mmu_interval_read_retry() after a successful fault.
Pass HMM_RANGE_DEFAULT_TIMEOUT as the helper retry budget for each HMM
fault attempt. This scopes the timeout to repeated HMM notifier retries
while preserving the outer retry loops that restart when the interval is
invalidated before GPU SVM updates or consumes the mapping state.
Leave drm_gpusvm_check_pages() on hmm_range_fault() because that path is
called with the mmap lock already held by its caller.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
drivers/gpu/drm/drm_gpusvm.c | 61 +++++-------------------------------------
1 file changed, 7 insertions(+), 54 deletions(-)
diff --git a/drivers/gpu/drm/drm_gpusvm.c b/drivers/gpu/drm/drm_gpusvm.c
index 958cb605aedd..de5bbfe58ee9 100644
--- a/drivers/gpu/drm/drm_gpusvm.c
+++ b/drivers/gpu/drm/drm_gpusvm.c
@@ -773,8 +773,7 @@ enum drm_gpusvm_scan_result drm_gpusvm_scan_mm(struct drm_gpusvm_range *range,
.end = end,
.dev_private_owner = dev_private_owner,
};
- unsigned long timeout =
- jiffies + msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
+ unsigned long timeout = msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
enum drm_gpusvm_scan_result state = DRM_GPUSVM_SCAN_UNPOPULATED, new_state;
unsigned long *pfns;
unsigned long npages = npages_in_range(start, end);
@@ -788,22 +787,7 @@ enum drm_gpusvm_scan_result drm_gpusvm_scan_mm(struct drm_gpusvm_range *range,
hmm_range.hmm_pfns = pfns;
retry:
- hmm_range.notifier_seq = mmu_interval_read_begin(notifier);
- mmap_read_lock(range->gpusvm->mm);
-
- while (true) {
- err = hmm_range_fault(&hmm_range);
- if (err == -EBUSY) {
- if (time_after(jiffies, timeout))
- break;
-
- hmm_range.notifier_seq =
- mmu_interval_read_begin(notifier);
- continue;
- }
- break;
- }
- mmap_read_unlock(range->gpusvm->mm);
+ err = hmm_range_fault_unlocked_timeout(&hmm_range, timeout);
if (err)
goto err_free;
@@ -1406,8 +1390,7 @@ int drm_gpusvm_get_pages(struct drm_gpusvm *gpusvm,
.dev_private_owner = ctx->device_private_page_owner,
};
void *zdd;
- unsigned long timeout =
- jiffies + msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
+ unsigned long timeout = msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
unsigned long i, j;
unsigned long npages = npages_in_range(pages_start, pages_end);
unsigned long num_dma_mapped;
@@ -1422,9 +1405,6 @@ int drm_gpusvm_get_pages(struct drm_gpusvm *gpusvm,
struct dma_iova_state *state = &svm_pages->state;
retry:
- if (time_after(jiffies, timeout))
- return -EBUSY;
-
hmm_range.notifier_seq = mmu_interval_read_begin(notifier);
if (drm_gpusvm_pages_valid_unlocked(gpusvm, svm_pages))
goto set_seqno;
@@ -1439,21 +1419,7 @@ int drm_gpusvm_get_pages(struct drm_gpusvm *gpusvm,
}
hmm_range.hmm_pfns = pfns;
- while (true) {
- mmap_read_lock(mm);
- err = hmm_range_fault(&hmm_range);
- mmap_read_unlock(mm);
-
- if (err == -EBUSY) {
- if (time_after(jiffies, timeout))
- break;
-
- hmm_range.notifier_seq =
- mmu_interval_read_begin(notifier);
- continue;
- }
- break;
- }
+ err = hmm_range_fault_unlocked_timeout(&hmm_range, timeout);
mmput(mm);
if (err)
goto err_free;
@@ -1720,8 +1686,7 @@ int drm_gpusvm_range_evict(struct drm_gpusvm *gpusvm,
.end = drm_gpusvm_range_end(range),
.dev_private_owner = NULL,
};
- unsigned long timeout =
- jiffies + msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
+ unsigned long timeout = msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
unsigned long *pfns;
unsigned long npages = npages_in_range(drm_gpusvm_range_start(range),
drm_gpusvm_range_end(range));
@@ -1736,24 +1701,12 @@ int drm_gpusvm_range_evict(struct drm_gpusvm *gpusvm,
return -ENOMEM;
hmm_range.hmm_pfns = pfns;
- while (!time_after(jiffies, timeout)) {
- hmm_range.notifier_seq = mmu_interval_read_begin(notifier);
- if (time_after(jiffies, timeout)) {
- err = -ETIME;
- break;
- }
-
- mmap_read_lock(mm);
- err = hmm_range_fault(&hmm_range);
- mmap_read_unlock(mm);
- if (err != -EBUSY)
- break;
- }
+ err = hmm_range_fault_unlocked_timeout(&hmm_range, timeout);
kvfree(pfns);
mmput(mm);
- return err;
+ return err == -EBUSY ? -ETIME : err;
}
EXPORT_SYMBOL_GPL(drm_gpusvm_range_evict);
^ permalink raw reply related
* [PATCH v9 7/8] accel/amdxdna: Use hmm_range_fault_unlocked_timeout() for range population
From: Stanislav Kinsburskii @ 2026-07-15 18:16 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
aie2_populate_range() takes mmap_read_lock() only around hmm_range_fault().
It also open-codes the mmu interval sequence setup before each HMM walk and
retries -EBUSY until HMM_RANGE_DEFAULT_TIMEOUT expires.
Use hmm_range_fault_unlocked_timeout() instead. The HMM helper now owns
the mmap lock and refreshes mapp->range.notifier_seq for its internal
retries, so the driver only needs to call the helper and then validate the
sequence before marking the mapping populated.
Pass HMM_RANGE_DEFAULT_TIMEOUT as the helper retry budget for each HMM
population attempt. This scopes the timeout to repeated HMM notifier
retries while preserving the existing outer loop that moves between invalid
mappings and restarts when the interval is invalidated before the driver
updates its mapping state.
Keep returning -ETIME when the HMM retry budget expires, matching the
driver's existing timeout error convention.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
drivers/accel/amdxdna/aie2_ctx.c | 23 ++++-------------------
1 file changed, 4 insertions(+), 19 deletions(-)
diff --git a/drivers/accel/amdxdna/aie2_ctx.c b/drivers/accel/amdxdna/aie2_ctx.c
index 54486960cbf5..b5b4ca263002 100644
--- a/drivers/accel/amdxdna/aie2_ctx.c
+++ b/drivers/accel/amdxdna/aie2_ctx.c
@@ -1034,7 +1034,7 @@ static int aie2_populate_range(struct amdxdna_gem_obj *abo)
bool found;
int ret;
- timeout = jiffies + msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
+ timeout = msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
again:
found = false;
down_write(&xdna->notifier_lock);
@@ -1061,24 +1061,9 @@ static int aie2_populate_range(struct amdxdna_gem_obj *abo)
return -EFAULT;
}
- mapp->range.notifier_seq = mmu_interval_read_begin(&mapp->notifier);
- mmap_read_lock(mm);
- ret = hmm_range_fault(&mapp->range);
- mmap_read_unlock(mm);
- if (ret) {
- if (time_after(jiffies, timeout)) {
- ret = -ETIME;
- goto put_mm;
- }
-
- if (ret == -EBUSY) {
- amdxdna_umap_put(mapp);
- mmput(mm);
- goto again;
- }
-
+ ret = hmm_range_fault_unlocked_timeout(&mapp->range, timeout);
+ if (ret)
goto put_mm;
- }
down_write(&xdna->notifier_lock);
if (mmu_interval_read_retry(&mapp->notifier, mapp->range.notifier_seq)) {
@@ -1096,7 +1081,7 @@ static int aie2_populate_range(struct amdxdna_gem_obj *abo)
put_mm:
amdxdna_umap_put(mapp);
mmput(mm);
- return ret;
+ return ret == -EBUSY ? -ETIME : ret;
}
int aie2_cmd_submit(struct amdxdna_hwctx *hwctx, struct amdxdna_sched_job *job, u64 *seq)
^ permalink raw reply related
* [PATCH v9 6/8] RDMA/umem: Use hmm_range_fault_unlocked_timeout() for ODP faults
From: Stanislav Kinsburskii @ 2026-07-15 18:16 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
ib_umem_odp_map_dma_and_lock() takes mmap_read_lock() only around
hmm_range_fault(), then retries -EBUSY until HMM_RANGE_DEFAULT_TIMEOUT
expires.
Use hmm_range_fault_unlocked_timeout() instead. The HMM helper now owns
the mmap lock and refreshes range->notifier_seq for its internal retries.
ODP keeps using HMM_RANGE_DEFAULT_TIMEOUT for each HMM fault attempt,
while interval invalidation retries continue to be handled by the existing
outer loop.
ODP still validates the interval notifier sequence while holding umem_mutex
before DMA mapping pages.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
drivers/infiniband/core/umem_odp.c | 18 +++++-------------
1 file changed, 5 insertions(+), 13 deletions(-)
diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c
index 404fa1cc3254..9cc21cd762d9 100644
--- a/drivers/infiniband/core/umem_odp.c
+++ b/drivers/infiniband/core/umem_odp.c
@@ -329,7 +329,7 @@ int ib_umem_odp_map_dma_and_lock(struct ib_umem_odp *umem_odp, u64 user_virt,
struct mm_struct *owning_mm = umem_odp->umem.owning_mm;
int pfn_index, dma_index, ret = 0, start_idx;
unsigned int page_shift, hmm_order, pfn_start_idx;
- unsigned long num_pfns, current_seq;
+ unsigned long num_pfns;
struct hmm_range range = {};
unsigned long timeout;
@@ -363,26 +363,18 @@ int ib_umem_odp_map_dma_and_lock(struct ib_umem_odp *umem_odp, u64 user_virt,
}
range.hmm_pfns = &(umem_odp->map.pfn_list[pfn_start_idx]);
- timeout = jiffies + msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
+ timeout = msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
retry:
- current_seq = range.notifier_seq =
- mmu_interval_read_begin(&umem_odp->notifier);
-
- mmap_read_lock(owning_mm);
- ret = hmm_range_fault(&range);
- mmap_read_unlock(owning_mm);
- if (unlikely(ret)) {
- if (ret == -EBUSY && !time_after(jiffies, timeout))
- goto retry;
+ ret = hmm_range_fault_unlocked_timeout(&range, timeout);
+ if (unlikely(ret))
goto out_put_mm;
- }
start_idx = (range.start - ib_umem_start(umem_odp)) >> page_shift;
dma_index = start_idx;
mutex_lock(&umem_odp->umem_mutex);
- if (mmu_interval_read_retry(&umem_odp->notifier, current_seq)) {
+ if (mmu_interval_read_retry(&umem_odp->notifier, range.notifier_seq)) {
mutex_unlock(&umem_odp->umem_mutex);
goto retry;
}
^ permalink raw reply related
* [PATCH v9 5/8] drm/nouveau: Use hmm_range_fault_unlocked_timeout() for SVM faults
From: Stanislav Kinsburskii @ 2026-07-15 18:16 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
nouveau_range_fault() takes mmap_read_lock() only to call
hmm_range_fault(). It also keeps a single HMM_RANGE_DEFAULT_TIMEOUT
deadline across both HMM -EBUSY retries and post-fault
mmu_interval_read_retry() retries.
Use hmm_range_fault_unlocked_timeout() instead. The HMM helper now owns
the mmap lock and refreshes range->notifier_seq for its internal retries.
Nouveau keeps its existing absolute deadline in the outer loop and passes
the remaining jiffies to the helper for each fault attempt, so retries
caused by mmu_interval_read_retry() do not reset the overall retry budget.
Nouveau still validates the interval notifier sequence while holding
svmm->mutex before programming the GPU mapping.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
drivers/gpu/drm/nouveau/nouveau_svm.c | 20 +++++++++++---------
1 file changed, 11 insertions(+), 9 deletions(-)
diff --git a/drivers/gpu/drm/nouveau/nouveau_svm.c b/drivers/gpu/drm/nouveau/nouveau_svm.c
index dcc92131488e..58735446d783 100644
--- a/drivers/gpu/drm/nouveau/nouveau_svm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_svm.c
@@ -678,20 +678,22 @@ static int nouveau_range_fault(struct nouveau_svmm *svmm,
range.end = notifier->notifier.interval_tree.last + 1;
while (true) {
- if (time_after(jiffies, timeout)) {
+ long remaining = timeout - jiffies;
+
+ /*
+ * The HMM timeout only bounds retries while HMM is walking and
+ * faulting the range. This fault is handled by a kernel worker,
+ * so fatal signals from the faulting process cannot stop an
+ * endless stream of invalidations here.
+ */
+ if (time_after_eq(jiffies, timeout)) {
ret = -EBUSY;
goto out;
}
- range.notifier_seq = mmu_interval_read_begin(range.notifier);
- mmap_read_lock(mm);
- ret = hmm_range_fault(&range);
- mmap_read_unlock(mm);
- if (ret) {
- if (ret == -EBUSY)
- continue;
+ ret = hmm_range_fault_unlocked_timeout(&range, remaining);
+ if (ret)
goto out;
- }
mutex_lock(&svmm->mutex);
if (mmu_interval_read_retry(range.notifier,
^ permalink raw reply related
* [PATCH v9 4/8] mshv: Use hmm_range_fault_unlocked_timeout() for region faults
From: Stanislav Kinsburskii @ 2026-07-15 18:16 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
MSHV currently faults movable memory regions by taking mmap_read_lock()
around hmm_range_fault(). That prevents the fault path from handling VMAs
whose fault handlers need to drop mmap_lock, such as userfaultfd-backed
mappings.
Use hmm_range_fault_unlocked_timeout() instead. Passing a timeout of 0
preserves MSHV's existing unbounded retry behavior while letting the HMM
helper own mmap_lock acquisition and refresh range->notifier_seq internally
before walking the range. After the fault succeeds, MSHV still takes
mreg_mutex and checks mmu_interval_read_retry() before installing the pages
into the region, so the existing invalidation synchronization is preserved.
Fold the small fault-and-lock helper into mshv_region_range_fault(), since
the remaining retry path is just the standard "fault, take the driver lock,
check the interval notifier sequence" pattern.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
drivers/hv/mshv_regions.c | 54 ++++++++-------------------------------------
1 file changed, 10 insertions(+), 44 deletions(-)
diff --git a/drivers/hv/mshv_regions.c b/drivers/hv/mshv_regions.c
index 6d65e5b42152..dddaade31b5d 100644
--- a/drivers/hv/mshv_regions.c
+++ b/drivers/hv/mshv_regions.c
@@ -381,46 +381,6 @@ int mshv_region_get(struct mshv_mem_region *region)
return kref_get_unless_zero(®ion->mreg_refcount);
}
-/**
- * mshv_region_hmm_fault_and_lock - Handle HMM faults and lock the memory region
- * @region: Pointer to the memory region structure
- * @range: Pointer to the HMM range structure
- *
- * This function performs the following steps:
- * 1. Reads the notifier sequence for the HMM range.
- * 2. Acquires a read lock on the memory map.
- * 3. Handles HMM faults for the specified range.
- * 4. Releases the read lock on the memory map.
- * 5. If successful, locks the memory region mutex.
- * 6. Verifies if the notifier sequence has changed during the operation.
- * If it has, releases the mutex and returns -EBUSY to match with
- * hmm_range_fault() return code for repeating.
- *
- * Return: 0 on success, a negative error code otherwise.
- */
-static int mshv_region_hmm_fault_and_lock(struct mshv_mem_region *region,
- struct hmm_range *range)
-{
- int ret;
-
- range->notifier_seq = mmu_interval_read_begin(range->notifier);
- mmap_read_lock(region->mreg_mni.mm);
- ret = hmm_range_fault(range);
- mmap_read_unlock(region->mreg_mni.mm);
- if (ret)
- return ret;
-
- mutex_lock(®ion->mreg_mutex);
-
- if (mmu_interval_read_retry(range->notifier, range->notifier_seq)) {
- mutex_unlock(®ion->mreg_mutex);
- cond_resched();
- return -EBUSY;
- }
-
- return 0;
-}
-
/**
* mshv_region_range_fault - Handle memory range faults for a given region.
* @region: Pointer to the memory region structure.
@@ -452,13 +412,19 @@ static int mshv_region_range_fault(struct mshv_mem_region *region,
range.start = region->start_uaddr + page_offset * HV_HYP_PAGE_SIZE;
range.end = range.start + page_count * HV_HYP_PAGE_SIZE;
- do {
- ret = mshv_region_hmm_fault_and_lock(region, &range);
- } while (ret == -EBUSY);
-
+again:
+ ret = hmm_range_fault_unlocked_timeout(&range, 0);
if (ret)
goto out;
+ mutex_lock(®ion->mreg_mutex);
+
+ if (mmu_interval_read_retry(range.notifier, range.notifier_seq)) {
+ mutex_unlock(®ion->mreg_mutex);
+ cond_resched();
+ goto again;
+ }
+
for (i = 0; i < page_count; i++)
region->mreg_pages[page_offset + i] = hmm_pfn_to_page(pfns[i]);
^ permalink raw reply related
* [PATCH v9 3/8] selftests/mm: add HMM test for mmap lock-dropping faults
From: Stanislav Kinsburskii @ 2026-07-15 18:16 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
Add test_hmm coverage for the HMM lock-dropping fault path. The test module
gets a new HMM_DMIRROR_READ_UNLOCKED ioctl that calls
hmm_range_fault_unlocked_timeout() with a timeout of 0, exercising the
unbounded retry mode while allowing the mmap lock to be dropped during
fault handling.
Add a userfaultfd_read selftest that registers an anonymous mapping with
UFFDIO_REGISTER_MODE_MISSING, services the faults from a handler thread
with UFFDIO_COPY, and verifies that HMM can read back the data supplied by
the handler. This exercises the path where handle_mm_fault() drops
mmap_lock and hmm_range_fault_unlocked_timeout() restarts the walk
internally.
Assisted-by: GitHub-Copilot:claude-opus-4.6
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
lib/test_hmm.c | 107 +++++++++++++++++++++++
lib/test_hmm_uapi.h | 1
tools/testing/selftests/mm/hmm-tests.c | 150 ++++++++++++++++++++++++++++++++
3 files changed, 257 insertions(+), 1 deletion(-)
diff --git a/lib/test_hmm.c b/lib/test_hmm.c
index 45c0cb992218..6205fb313bd0 100644
--- a/lib/test_hmm.c
+++ b/lib/test_hmm.c
@@ -389,6 +389,67 @@ static int dmirror_range_fault(struct dmirror *dmirror,
return ret;
}
+static int dmirror_range_fault_unlocked(struct dmirror *dmirror,
+ struct hmm_range *range,
+ unsigned long timeout)
+{
+ int ret;
+
+ while (true) {
+ ret = hmm_range_fault_unlocked_timeout(range, timeout);
+ if (ret)
+ goto out;
+
+ mutex_lock(&dmirror->mutex);
+ if (mmu_interval_read_retry(range->notifier,
+ range->notifier_seq)) {
+ mutex_unlock(&dmirror->mutex);
+ continue;
+ }
+ break;
+ }
+
+ ret = dmirror_do_fault(dmirror, range);
+
+ mutex_unlock(&dmirror->mutex);
+out:
+ return ret;
+}
+
+static int dmirror_fault_unlocked(struct dmirror *dmirror,
+ unsigned long start,
+ unsigned long end, bool write,
+ unsigned long timeout)
+{
+ struct mm_struct *mm = dmirror->notifier.mm;
+ unsigned long addr;
+ unsigned long pfns[32];
+ struct hmm_range range = {
+ .notifier = &dmirror->notifier,
+ .hmm_pfns = pfns,
+ .pfn_flags_mask = 0,
+ .default_flags =
+ HMM_PFN_REQ_FAULT | (write ? HMM_PFN_REQ_WRITE : 0),
+ .dev_private_owner = dmirror->mdevice,
+ };
+ int ret = 0;
+
+ if (!mmget_not_zero(mm))
+ return -EFAULT;
+
+ for (addr = start; addr < end; addr = range.end) {
+ range.start = addr;
+ range.end = min(addr + (ARRAY_SIZE(pfns) << PAGE_SHIFT), end);
+
+ ret = dmirror_range_fault_unlocked(dmirror, &range, timeout);
+ if (ret)
+ break;
+ }
+
+ mmput(mm);
+ return ret;
+}
+
static int dmirror_fault(struct dmirror *dmirror, unsigned long start,
unsigned long end, bool write)
{
@@ -488,6 +549,48 @@ static int dmirror_read(struct dmirror *dmirror, struct hmm_dmirror_cmd *cmd)
return ret;
}
+static int dmirror_read_unlocked(struct dmirror *dmirror,
+ struct hmm_dmirror_cmd *cmd,
+ unsigned long timeout)
+{
+ struct dmirror_bounce bounce;
+ unsigned long start, end;
+ unsigned long size = cmd->npages << PAGE_SHIFT;
+ int ret;
+
+ start = cmd->addr;
+ end = start + size;
+ if (end < start)
+ return -EINVAL;
+
+ ret = dmirror_bounce_init(&bounce, start, size);
+ if (ret)
+ return ret;
+
+ while (1) {
+ mutex_lock(&dmirror->mutex);
+ ret = dmirror_do_read(dmirror, start, end, &bounce);
+ mutex_unlock(&dmirror->mutex);
+ if (ret != -ENOENT)
+ break;
+
+ start = cmd->addr + (bounce.cpages << PAGE_SHIFT);
+ ret = dmirror_fault_unlocked(dmirror, start, end, false, timeout);
+ if (ret)
+ break;
+ cmd->faults++;
+ }
+
+ if (ret == 0) {
+ if (copy_to_user(u64_to_user_ptr(cmd->ptr), bounce.ptr,
+ bounce.size))
+ ret = -EFAULT;
+ }
+ cmd->cpages = bounce.cpages;
+ dmirror_bounce_fini(&bounce);
+ return ret;
+}
+
static int dmirror_do_write(struct dmirror *dmirror, unsigned long start,
unsigned long end, struct dmirror_bounce *bounce)
{
@@ -1572,7 +1675,9 @@ static long dmirror_fops_unlocked_ioctl(struct file *filp,
dmirror->flags = cmd.npages;
ret = 0;
break;
-
+ case HMM_DMIRROR_READ_UNLOCKED:
+ ret = dmirror_read_unlocked(dmirror, &cmd, 0);
+ break;
default:
return -EINVAL;
}
diff --git a/lib/test_hmm_uapi.h b/lib/test_hmm_uapi.h
index f94c6d457338..ea9b0ec404fb 100644
--- a/lib/test_hmm_uapi.h
+++ b/lib/test_hmm_uapi.h
@@ -38,6 +38,7 @@ struct hmm_dmirror_cmd {
#define HMM_DMIRROR_CHECK_EXCLUSIVE _IOWR('H', 0x06, struct hmm_dmirror_cmd)
#define HMM_DMIRROR_RELEASE _IOWR('H', 0x07, struct hmm_dmirror_cmd)
#define HMM_DMIRROR_FLAGS _IOWR('H', 0x08, struct hmm_dmirror_cmd)
+#define HMM_DMIRROR_READ_UNLOCKED _IOWR('H', 0x09, struct hmm_dmirror_cmd)
#define HMM_DMIRROR_FLAG_FAIL_ALLOC (1ULL << 0)
diff --git a/tools/testing/selftests/mm/hmm-tests.c b/tools/testing/selftests/mm/hmm-tests.c
index 6fccbdab02ee..5acb728666f8 100644
--- a/tools/testing/selftests/mm/hmm-tests.c
+++ b/tools/testing/selftests/mm/hmm-tests.c
@@ -29,6 +29,10 @@
#include <sys/mman.h>
#include <sys/ioctl.h>
#include <sys/time.h>
+#include <sys/syscall.h>
+#include <sys/eventfd.h>
+#include <linux/userfaultfd.h>
+#include <poll.h>
/*
* This is a private UAPI to the kernel test module so it isn't exported
@@ -2952,4 +2956,150 @@ TEST_F_TIMEOUT(hmm, benchmark_thp_migration, 120)
&thp_results, ®ular_results);
}
}
+/*
+ * Test that HMM can fault in pages backed by userfaultfd using the
+ * hmm_range_fault_unlocked_timeout() path with no timeout. This exercises
+ * the lock-drop retry logic in the HMM framework.
+ */
+struct uffd_thread_args {
+ int uffd;
+ int stop_fd;
+ void *page_buffer;
+ unsigned long page_size;
+};
+
+static void *uffd_handler_thread(void *arg)
+{
+ struct uffd_thread_args *args = arg;
+ struct uffd_msg msg;
+ struct uffdio_copy copy;
+ struct pollfd pollfd[2];
+ int ret;
+
+ pollfd[0].fd = args->uffd;
+ pollfd[0].events = POLLIN;
+ pollfd[1].fd = args->stop_fd;
+ pollfd[1].events = POLLIN;
+
+ while (1) {
+ ret = poll(pollfd, 2, -1);
+ if (ret <= 0)
+ break;
+ if (pollfd[1].revents)
+ break;
+ if (!(pollfd[0].revents & POLLIN))
+ break;
+
+ ret = read(args->uffd, &msg, sizeof(msg));
+ if (ret != sizeof(msg))
+ break;
+
+ if (msg.event != UFFD_EVENT_PAGEFAULT)
+ break;
+
+ /* Fill the page with a known pattern */
+ memset(args->page_buffer, 0xAB, args->page_size);
+
+ copy.dst = msg.arg.pagefault.address & ~(args->page_size - 1);
+ copy.src = (unsigned long)args->page_buffer;
+ copy.len = args->page_size;
+ copy.mode = 0;
+ copy.copy = 0;
+
+ ret = ioctl(args->uffd, UFFDIO_COPY, ©);
+ if (ret < 0)
+ break;
+ }
+
+ return NULL;
+}
+
+TEST_F(hmm, userfaultfd_read)
+{
+ struct hmm_buffer *buffer;
+ struct uffd_thread_args uffd_args;
+ unsigned long npages;
+ unsigned long size;
+ unsigned long i;
+ unsigned char *ptr;
+ pthread_t thread;
+ int uffd;
+ int stop_fd;
+ int ret;
+ struct uffdio_api api;
+ struct uffdio_register reg;
+ uint64_t stop = 1;
+ ssize_t nwrite;
+
+ npages = 4;
+ size = npages << self->page_shift;
+
+ /* Create userfaultfd */
+ uffd = syscall(__NR_userfaultfd, O_CLOEXEC | O_NONBLOCK);
+ if (uffd < 0)
+ SKIP(return, "userfaultfd not available");
+
+ api.api = UFFD_API;
+ api.features = 0;
+ ret = ioctl(uffd, UFFDIO_API, &api);
+ ASSERT_EQ(ret, 0);
+
+ buffer = malloc(sizeof(*buffer));
+ ASSERT_NE(buffer, NULL);
+
+ buffer->fd = -1;
+ buffer->size = size;
+ buffer->mirror = malloc(size);
+ ASSERT_NE(buffer->mirror, NULL);
+
+ /* Create anonymous mapping */
+ buffer->ptr = mmap(NULL, size,
+ PROT_READ | PROT_WRITE,
+ MAP_PRIVATE | MAP_ANONYMOUS,
+ -1, 0);
+ ASSERT_NE(buffer->ptr, MAP_FAILED);
+
+ /* Register the region with userfaultfd */
+ reg.range.start = (unsigned long)buffer->ptr;
+ reg.range.len = size;
+ reg.mode = UFFDIO_REGISTER_MODE_MISSING;
+ ret = ioctl(uffd, UFFDIO_REGISTER, ®);
+ ASSERT_EQ(ret, 0);
+
+ /* Set up the handler thread */
+ uffd_args.uffd = uffd;
+ stop_fd = eventfd(0, EFD_CLOEXEC);
+ ASSERT_GE(stop_fd, 0);
+ uffd_args.stop_fd = stop_fd;
+ uffd_args.page_buffer = malloc(self->page_size);
+ ASSERT_NE(uffd_args.page_buffer, NULL);
+ uffd_args.page_size = self->page_size;
+
+ ret = pthread_create(&thread, NULL, uffd_handler_thread, &uffd_args);
+ ASSERT_EQ(ret, 0);
+
+ /*
+ * Use the unlocked read path which allows the mmap lock to be
+ * dropped during the fault, enabling userfaultfd resolution.
+ */
+ ret = hmm_dmirror_cmd(self->fd, HMM_DMIRROR_READ_UNLOCKED,
+ buffer, npages);
+ ASSERT_EQ(ret, 0);
+ ASSERT_EQ(buffer->cpages, npages);
+
+ /* Verify the device read the data filled by the uffd handler */
+ ptr = buffer->mirror;
+ for (i = 0; i < size; ++i)
+ ASSERT_EQ(ptr[i], (unsigned char)0xAB);
+
+ nwrite = write(stop_fd, &stop, sizeof(stop));
+ ASSERT_EQ(nwrite, sizeof(stop));
+ pthread_join(thread, NULL);
+ close(stop_fd);
+ free(uffd_args.page_buffer);
+ close(uffd);
+ hmm_buffer_free(buffer);
+}
+
+
TEST_HARNESS_MAIN
^ permalink raw reply related
* [PATCH v9 2/8] mm/hmm: add hmm_range_fault_unlocked_timeout() for mmap lock-drop support
From: Stanislav Kinsburskii @ 2026-07-15 18:16 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
hmm_range_fault() requires the caller to hold the mmap read lock for the
duration of the call. This is incompatible with mappings whose fault
handler may release the mmap lock, notably userfaultfd-managed regions,
where handle_mm_fault() can return VM_FAULT_RETRY or VM_FAULT_COMPLETED
after dropping the lock. Drivers that need to populate device page tables
for such mappings have no way to do so today.
Add hmm_range_fault_unlocked_timeout() for callers that do not need to hold
mmap_lock across any work outside the HMM fault itself. The helper takes
mmap_read_lock_killable() internally, calls the common HMM fault
implementation, and releases the lock before returning if it is still held.
The timeout is specified in jiffies; passing 0 retries indefinitely, while
a non-zero timeout makes the helper return -EBUSY when the retry budget
expires.
When handle_mm_fault() drops mmap_lock, or when the range is invalidated,
hmm_range_fault_unlocked_timeout() refreshes range->notifier_seq and
retries the walk internally. If the lock was dropped, the retry deadline is
also restarted because a lock-dropping fault handler made progress.
Ordinary -EBUSY retries keep the existing deadline, preserving the caller's
timeout policy for repeated mmu-notifier invalidations.
The caller only needs to perform the usual post-success
mmu_interval_read_retry() check while holding its update lock before
consuming the pfns. If mmap_lock acquisition is interrupted or a fatal
signal is pending during retry handling, -EINTR is returned instead.
The common implementation conditionally sets FAULT_FLAG_ALLOW_RETRY and
FAULT_FLAG_KILLABLE only for hmm_range_fault_unlocked_timeout(). The
existing hmm_range_fault() path still passes no locked state, does not
allow handle_mm_fault() to drop mmap_lock, and remains a thin wrapper
preserving the existing API contract for current callers.
The previous refactor that moved page fault handling out of the page-table
walk callbacks is what makes this change small. Faults now run after
walk_page_range() has unwound, with only mmap_lock held, so dropping it
does not interact with the walker's pte spinlock or hugetlb_vma_lock.
Hugetlb regions therefore participate in the unlocked path uniformly with
PTE- and PMD-level mappings; no special case is required.
Documentation/mm/hmm.rst is updated with a description of the new API and
the recommended caller pattern.
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
Documentation/mm/hmm.rst | 79 ++++++++++++++++------
include/linux/hmm.h | 2 +
mm/hmm.c | 165 ++++++++++++++++++++++++++++++++++++++--------
3 files changed, 194 insertions(+), 52 deletions(-)
diff --git a/Documentation/mm/hmm.rst b/Documentation/mm/hmm.rst
index 7d61b7a8b65b..e021218ada58 100644
--- a/Documentation/mm/hmm.rst
+++ b/Documentation/mm/hmm.rst
@@ -156,42 +156,57 @@ During the ops->invalidate() callback the device driver must perform the
update action to the range (mark range read only, or fully unmap, etc.). The
device must complete the update before the driver callback returns.
-When the device driver wants to populate a range of virtual addresses, it can
-use::
+When the device driver wants to populate a range of virtual addresses, the
+normal interface is::
- int hmm_range_fault(struct hmm_range *range);
+ int hmm_range_fault_unlocked_timeout(struct hmm_range *range,
+ unsigned long timeout);
It will trigger a page fault on missing or read-only entries if write access is
requested (see below). Page faults use the generic mm page fault code path just
-like a CPU page fault. The usage pattern is::
+like a CPU page fault.
+
+The caller must not hold ``mmap_read_lock`` before the call.
+``hmm_range_fault_unlocked_timeout()`` takes the mmap read lock internally and
+allows ``handle_mm_fault()`` to drop it during fault handling. This is required
+for VMAs whose fault handlers may release the mmap lock, for example regions
+managed by ``userfaultfd``.
+
+If the mmap lock is dropped or the range is invalidated, the function refreshes
+``range->notifier_seq`` and restarts the walk internally. ``-EINTR`` is returned
+if mmap lock acquisition is interrupted or a fatal signal is pending during
+retry handling.
+
+The timeout is specified in jiffies; passing ``0`` means retry indefinitely. The
+timeout exists to preserve caller policy for repeated mmu-notifier invalidation
+and is checked between retry attempts. HMM does not interrupt page fault
+handling when the timeout expires, but returns ``-EBUSY`` if the retry budget is
+exhausted before a stable range is obtained.
+
+The usage pattern is::
int driver_populate_range(...)
{
struct hmm_range range;
+ unsigned long timeout;
...
+ timeout = msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
range.notifier = &interval_sub;
range.start = ...;
range.end = ...;
range.hmm_pfns = ...;
- if (!mmget_not_zero(interval_sub->notifier.mm))
+ if (!mmget_not_zero(interval_sub.mm))
return -EFAULT;
again:
- range.notifier_seq = mmu_interval_read_begin(&interval_sub);
- mmap_read_lock(mm);
- ret = hmm_range_fault(&range);
- if (ret) {
- mmap_read_unlock(mm);
- if (ret == -EBUSY)
- goto again;
- return ret;
- }
- mmap_read_unlock(mm);
+ ret = hmm_range_fault_unlocked_timeout(&range, timeout);
+ if (ret)
+ goto out_put;
take_lock(driver->update);
- if (mmu_interval_read_retry(&ni, range.notifier_seq) {
+ if (mmu_interval_read_retry(range.notifier, range.notifier_seq)) {
release_lock(driver->update);
goto again;
}
@@ -200,13 +215,31 @@ like a CPU page fault. The usage pattern is::
* under the update lock */
release_lock(driver->update);
- return 0;
+ ret = 0;
+
+ out_put:
+ mmput(interval_sub.mm);
+ return ret;
}
The driver->update lock is the same lock that the driver takes inside its
invalidate() callback. That lock must be held before calling
mmu_interval_read_retry() to avoid any race with a concurrent CPU page table
-update.
+update. The retry check must use the same notifier and sequence number stored
+in ``range`` by ``hmm_range_fault_unlocked_timeout()``.
+
+Holding the mmap lock across HMM faults
+=======================================
+
+Most callers should use ``hmm_range_fault_unlocked_timeout()``. If a driver
+really needs to hold the mmap lock across work outside HMM, it can use::
+
+ int hmm_range_fault(struct hmm_range *range);
+
+The mmap lock must be held by the caller and will remain held on return. This
+interface cannot support VMAs whose fault handlers need to drop the mmap lock.
+New callers should prefer ``hmm_range_fault_unlocked_timeout()`` unless they
+have a specific requirement to keep the mmap lock held across the call.
Leverage default_flags and pfn_flags_mask
=========================================
@@ -221,8 +254,8 @@ permission, it sets::
range->default_flags = HMM_PFN_REQ_FAULT;
range->pfn_flags_mask = 0;
-and calls hmm_range_fault() as described above. This will fill fault all pages
-in the range with at least read permission.
+and calls the HMM range fault helper as described above. This will fault
+all pages in the range with at least read permission.
Now let's say the driver wants to do the same except for one page in the range for
which it wants to have write permission. Now driver set::
@@ -236,9 +269,9 @@ address == range->start + (index_of_write << PAGE_SHIFT) it will fault with
write permission i.e., if the CPU pte does not have write permission set then HMM
will call handle_mm_fault().
-After hmm_range_fault completes the flag bits are set to the current state of
-the page tables, ie HMM_PFN_VALID | HMM_PFN_WRITE will be set if the page is
-writable.
+After the HMM range fault helper completes the flag bits are set to the
+current state of the page tables, ie HMM_PFN_VALID | HMM_PFN_WRITE will be
+set if the page is writable.
Represent and manage device memory from core kernel point of view
diff --git a/include/linux/hmm.h b/include/linux/hmm.h
index db75ffc949a7..6f04e3932f5b 100644
--- a/include/linux/hmm.h
+++ b/include/linux/hmm.h
@@ -123,6 +123,8 @@ struct hmm_range {
* Please see Documentation/mm/hmm.rst for how to use the range API.
*/
int hmm_range_fault(struct hmm_range *range);
+int hmm_range_fault_unlocked_timeout(struct hmm_range *range,
+ unsigned long timeout);
/*
* HMM_RANGE_DEFAULT_TIMEOUT - default timeout (ms) when waiting for a range
diff --git a/mm/hmm.c b/mm/hmm.c
index bc9361a715fa..fc2e1cd0cb22 100644
--- a/mm/hmm.c
+++ b/mm/hmm.c
@@ -32,6 +32,7 @@
struct hmm_vma_walk {
struct hmm_range *range;
+ bool *locked;
unsigned long last;
unsigned long end;
unsigned int required_fault;
@@ -44,6 +45,14 @@ struct hmm_vma_walk {
*/
#define HMM_FAULT_PENDING -EAGAIN
+/*
+ * Internal sentinel returned by hmm_do_fault() when handle_mm_fault()
+ * completes a page fault with the mmap lock dropped. hmm_do_fault() sets
+ * *locked = false; the outer loop consumes the sentinel and never propagates
+ * it to the caller.
+ */
+#define HMM_FAULT_UNLOCKED -ENOLCK
+
enum {
HMM_NEED_FAULT = 1 << 0,
HMM_NEED_WRITE_FAULT = 1 << 1,
@@ -73,9 +82,9 @@ static int hmm_pfns_fill(unsigned long addr, unsigned long end,
*
* Called by the walk callbacks when they discover that part of the range
* needs a page fault. The callback records what to fault and returns
- * HMM_FAULT_PENDING; the outer loop in hmm_range_fault() drops back out of
- * walk_page_range() and invokes handle_mm_fault() from a context where no
- * page-table or hugetlb_vma_lock is held.
+ * HMM_FAULT_PENDING; the outer loop in hmm_range_fault_locked() drops
+ * back out of walk_page_range() and invokes handle_mm_fault() from a context
+ * where no page-table or hugetlb_vma_lock is held.
*/
static int hmm_record_fault(unsigned long addr, unsigned long end,
unsigned int required_fault,
@@ -624,7 +633,7 @@ static const struct mm_walk_ops hmm_walk_ops = {
/*
* hmm_do_fault - fault in a range recorded by a walk callback
*
- * Called from the outer loop in hmm_range_fault() after a callback
+ * Called from the outer loop in hmm_range_fault_locked() after a callback
* returned HMM_FAULT_PENDING. At this point we hold only mmap_lock;
* the page-table spinlock and any hugetlb_vma_lock acquired by the walk
* framework have already been released by the unwind.
@@ -641,6 +650,9 @@ static int hmm_do_fault(struct mm_struct *mm,
unsigned int fault_flags = FAULT_FLAG_REMOTE;
struct vm_area_struct *vma;
+ if (hmm_vma_walk->locked)
+ fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
+
vma = vma_lookup(mm, addr);
if (!vma)
return -EFAULT;
@@ -651,37 +663,34 @@ static int hmm_do_fault(struct mm_struct *mm,
fault_flags |= FAULT_FLAG_WRITE;
}
- for (; addr < end; addr += PAGE_SIZE)
- if (handle_mm_fault(vma, addr, fault_flags, NULL) &
- VM_FAULT_ERROR)
- return -EFAULT;
+ for (; addr < end; addr += PAGE_SIZE) {
+ vm_fault_t ret;
+
+ ret = handle_mm_fault(vma, addr, fault_flags, NULL);
+
+ if (ret & (VM_FAULT_COMPLETED | VM_FAULT_RETRY)) {
+ *hmm_vma_walk->locked = false;
+ return HMM_FAULT_UNLOCKED;
+ }
+
+ if (ret & VM_FAULT_ERROR) {
+ int err = vm_fault_to_errno(ret, 0);
+
+ if (WARN_ON(!err))
+ err = -EINVAL;
+
+ return err;
+ }
+ }
return -EBUSY;
}
-/**
- * hmm_range_fault - try to fault some address in a virtual address range
- * @range: argument structure
- *
- * Returns 0 on success or one of the following error codes:
- *
- * -EINVAL: Invalid arguments or mm or virtual address is in an invalid vma
- * (e.g., device file vma).
- * -ENOMEM: Out of memory.
- * -EPERM: Invalid permission (e.g., asking for write and range is read
- * only).
- * -EBUSY: The range has been invalidated and the caller needs to wait for
- * the invalidation to finish.
- * -EFAULT: A page was requested to be valid and could not be made valid
- * ie it has no backing VMA or it is illegal to access
- *
- * This is similar to get_user_pages(), except that it can read the page tables
- * without mutating them (ie causing faults).
- */
-int hmm_range_fault(struct hmm_range *range)
+static int hmm_range_fault_locked(struct hmm_range *range, bool *locked)
{
struct hmm_vma_walk hmm_vma_walk = {
.range = range,
+ .locked = locked,
.last = range->start,
};
struct mm_struct *mm = range->notifier->mm;
@@ -704,8 +713,14 @@ int hmm_range_fault(struct hmm_range *range)
* returns -EBUSY so the loop re-walks and picks up the
* now-present entries.
*/
- if (ret == HMM_FAULT_PENDING)
+ if (ret == HMM_FAULT_PENDING) {
ret = hmm_do_fault(mm, &hmm_vma_walk);
+ if (ret == HMM_FAULT_UNLOCKED) {
+ if (fatal_signal_pending(current))
+ return -EINTR;
+ return -EBUSY;
+ }
+ }
/*
* When -EBUSY is returned the loop restarts with
* hmm_vma_walk.last set to an address that has not been stored
@@ -715,8 +730,100 @@ int hmm_range_fault(struct hmm_range *range)
} while (ret == -EBUSY);
return ret;
}
+
+/**
+ * hmm_range_fault - try to fault some address in a virtual address range
+ * @range: argument structure
+ *
+ * Returns 0 on success or one of the following error codes:
+ *
+ * -EINVAL: Invalid arguments or mm or virtual address is in an invalid vma
+ * (e.g., device file vma).
+ * -ENOMEM: Out of memory.
+ * -EPERM: Invalid permission (e.g., asking for write and range is read
+ * only).
+ * -EBUSY: The range has been invalidated and the caller needs to wait for
+ * the invalidation to finish.
+ * -EFAULT: A page was requested to be valid and could not be made valid
+ * ie it has no backing VMA or it is illegal to access
+ *
+ * This is similar to get_user_pages(), except that it can read the page tables
+ * without mutating them (ie causing faults).
+ *
+ * The mmap lock must be held by the caller and will remain held on return.
+ * New users should prefer hmm_range_fault_unlocked_timeout() unless they
+ * specifically need to keep the mmap lock held across the call. This helper
+ * cannot support VMAs whose fault handlers need to drop the mmap lock.
+ */
+int hmm_range_fault(struct hmm_range *range)
+{
+ return hmm_range_fault_locked(range, NULL);
+}
EXPORT_SYMBOL(hmm_range_fault);
+/**
+ * hmm_range_fault_unlocked_timeout - fault in a range with a retry timeout
+ * @range: argument structure
+ * @timeout: timeout in jiffies for internal -EBUSY retries, or 0 to retry
+ * indefinitely
+ *
+ * The caller must not hold the mmap lock. The function takes the mmap read
+ * lock internally and allows handle_mm_fault() to drop it during faults. If
+ * the mmap lock is dropped or the range is invalidated, the function refreshes
+ * range->notifier_seq and restarts the walk internally.
+ *
+ * Passing 0 for @timeout retries indefinitely. A non-zero @timeout is a caller
+ * policy limit for repeated mmu-notifier invalidation retries. HMM does not
+ * interrupt page fault handling when the timeout expires, but returns -EBUSY
+ * if the retry budget is exhausted before a stable range is obtained.
+ *
+ * Returns 0 on success or one of the error codes documented for
+ * hmm_range_fault(). -EINTR is returned if mmap_lock acquisition is
+ * interrupted or a fatal signal is pending during retry handling.
+ */
+int hmm_range_fault_unlocked_timeout(struct hmm_range *range,
+ unsigned long timeout)
+{
+ struct mm_struct *mm = range->notifier->mm;
+ unsigned long deadline = 0;
+ bool locked = false;
+ int ret;
+
+ do {
+ if (fatal_signal_pending(current))
+ return -EINTR;
+
+ if (timeout) {
+ /*
+ * If the previous fault dropped mmap_lock, then the fault
+ * handler made progress. Restart the retry timeout in that
+ * case, but keep the existing deadline for ordinary -EBUSY
+ * retries.
+ */
+ if (!locked)
+ deadline = jiffies + timeout;
+
+ if (time_after(jiffies, deadline))
+ return -EBUSY;
+ }
+
+ range->notifier_seq =
+ mmu_interval_read_begin(range->notifier);
+
+ ret = mmap_read_lock_killable(mm);
+ if (ret)
+ return ret;
+
+ locked = true;
+ ret = hmm_range_fault_locked(range, &locked);
+ if (locked)
+ mmap_read_unlock(mm);
+ } while (ret == -EBUSY);
+
+ return ret;
+}
+EXPORT_SYMBOL(hmm_range_fault_unlocked_timeout);
+
/**
* hmm_dma_map_alloc - Allocate HMM map structure
* @dev: device to allocate structure for
^ permalink raw reply related
* [PATCH v9 1/8] mm/hmm: move page fault handling out of walk callbacks
From: Stanislav Kinsburskii @ 2026-07-15 18:15 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
In-Reply-To: <178413903133.1155966.3904063656020521607.stgit@skinsburskii>
hmm_range_fault() currently triggers page faults from inside the page-table
walk callbacks: hmm_vma_walk_pmd(), hmm_vma_walk_pud(),
hmm_vma_walk_hugetlb_entry() and the pte-level helper all call
hmm_vma_fault(), which in turn calls handle_mm_fault() while the walker
still holds nested locks. The pte spinlock is dropped explicitly by each
caller, and the hugetlb path manually drops and retakes
hugetlb_vma_lock_read around the fault to dodge a deadlock against the walk
framework's unconditional unlock.
This layering does not extend cleanly to fault handlers that may release
mmap_lock (VM_FAULT_RETRY, VM_FAULT_COMPLETED). If the lock is dropped
while walk_page_range() is mid-traversal, the VMA can be freed before the
walk framework's matching hugetlb_vma_unlock_read(), turning that unlock
into a use-after-free.
Split the responsibilities the way get_user_pages() does. Walk callbacks
become inspect-only: when they detect a range that needs to be faulted in,
they record it in struct hmm_vma_walk and return a private sentinel
(HMM_FAULT_PENDING). The outer loop in hmm_range_fault() then drops out of
walk_page_range(), invokes a new helper hmm_do_fault() that calls
handle_mm_fault() with only mmap_lock held, and restarts the walk so the
now-present entries are collected into hmm_pfns.
No functional change for existing callers. As a side effect the hugetlb
callback no longer needs the hugetlb_vma_{un}lock_read dance, and every
fault-path exit from the callbacks now releases the pte spinlock on a
single, common path. This refactor is also a precursor for adding an
unlockable variant of hmm_range_fault() in a follow-up patch.
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Stanislav Kinsburskii <skinsburskii@gmail.com>
---
mm/hmm.c | 118 +++++++++++++++++++++++++++++++++++++++-----------------------
1 file changed, 75 insertions(+), 43 deletions(-)
diff --git a/mm/hmm.c b/mm/hmm.c
index e5c1f4deed24..bc9361a715fa 100644
--- a/mm/hmm.c
+++ b/mm/hmm.c
@@ -33,8 +33,17 @@
struct hmm_vma_walk {
struct hmm_range *range;
unsigned long last;
+ unsigned long end;
+ unsigned int required_fault;
};
+/*
+ * Internal sentinel returned by walk callbacks when they need a page fault.
+ * The callback stores end/required_fault in hmm_vma_walk; the outer loop
+ * consumes the sentinel and never propagates it to the caller.
+ */
+#define HMM_FAULT_PENDING -EAGAIN
+
enum {
HMM_NEED_FAULT = 1 << 0,
HMM_NEED_WRITE_FAULT = 1 << 1,
@@ -60,37 +69,25 @@ static int hmm_pfns_fill(unsigned long addr, unsigned long end,
}
/*
- * hmm_vma_fault() - fault in a range lacking valid pmd or pte(s)
- * @addr: range virtual start address (inclusive)
- * @end: range virtual end address (exclusive)
- * @required_fault: HMM_NEED_* flags
- * @walk: mm_walk structure
- * Return: -EBUSY after page fault, or page fault error
+ * hmm_record_fault() - record a range that needs to be faulted in
*
- * This function will be called whenever pmd_none() or pte_none() returns true,
- * or whenever there is no page directory covering the virtual address range.
+ * Called by the walk callbacks when they discover that part of the range
+ * needs a page fault. The callback records what to fault and returns
+ * HMM_FAULT_PENDING; the outer loop in hmm_range_fault() drops back out of
+ * walk_page_range() and invokes handle_mm_fault() from a context where no
+ * page-table or hugetlb_vma_lock is held.
*/
-static int hmm_vma_fault(unsigned long addr, unsigned long end,
- unsigned int required_fault, struct mm_walk *walk)
+static int hmm_record_fault(unsigned long addr, unsigned long end,
+ unsigned int required_fault,
+ struct mm_walk *walk)
{
struct hmm_vma_walk *hmm_vma_walk = walk->private;
- struct vm_area_struct *vma = walk->vma;
- unsigned int fault_flags = FAULT_FLAG_REMOTE;
WARN_ON_ONCE(!required_fault);
hmm_vma_walk->last = addr;
-
- if (required_fault & HMM_NEED_WRITE_FAULT) {
- if (!(vma->vm_flags & VM_WRITE))
- return -EPERM;
- fault_flags |= FAULT_FLAG_WRITE;
- }
-
- for (; addr < end; addr += PAGE_SIZE)
- if (handle_mm_fault(vma, addr, fault_flags, NULL) &
- VM_FAULT_ERROR)
- return -EFAULT;
- return -EBUSY;
+ hmm_vma_walk->end = end;
+ hmm_vma_walk->required_fault = required_fault;
+ return HMM_FAULT_PENDING;
}
static unsigned int hmm_pte_need_fault(const struct hmm_vma_walk *hmm_vma_walk,
@@ -174,7 +171,7 @@ static int hmm_vma_walk_hole(unsigned long addr, unsigned long end,
return hmm_pfns_fill(addr, end, range, HMM_PFN_ERROR);
}
if (required_fault)
- return hmm_vma_fault(addr, end, required_fault, walk);
+ return hmm_record_fault(addr, end, required_fault, walk);
return hmm_pfns_fill(addr, end, range, 0);
}
@@ -209,7 +206,7 @@ static int hmm_vma_handle_pmd(struct mm_walk *walk, unsigned long addr,
required_fault =
hmm_range_need_fault(hmm_vma_walk, hmm_pfns, npages, cpu_flags);
if (required_fault)
- return hmm_vma_fault(addr, end, required_fault, walk);
+ return hmm_record_fault(addr, end, required_fault, walk);
pfn = pmd_pfn(pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
for (i = 0; addr < end; addr += PAGE_SIZE, i++, pfn++) {
@@ -328,7 +325,7 @@ static int hmm_vma_handle_pte(struct mm_walk *walk, unsigned long addr,
fault:
pte_unmap(ptep);
/* Fault any virtual address we were asked to fault */
- return hmm_vma_fault(addr, end, required_fault, walk);
+ return hmm_record_fault(addr, end, required_fault, walk);
}
#ifdef CONFIG_ARCH_HAS_PMD_SOFTLEAVES
@@ -371,7 +368,7 @@ static int hmm_vma_handle_absent_pmd(struct mm_walk *walk, unsigned long start,
npages, 0);
if (required_fault) {
if (softleaf_is_device_private(entry))
- return hmm_vma_fault(addr, end, required_fault, walk);
+ return hmm_record_fault(addr, end, required_fault, walk);
else
return -EFAULT;
}
@@ -517,7 +514,7 @@ static int hmm_vma_walk_pud(pud_t *pudp, unsigned long start, unsigned long end,
npages, cpu_flags);
if (required_fault) {
spin_unlock(ptl);
- return hmm_vma_fault(addr, end, required_fault, walk);
+ return hmm_record_fault(addr, end, required_fault, walk);
}
pfn = pud_pfn(pud) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
@@ -564,21 +561,8 @@ static int hmm_vma_walk_hugetlb_entry(pte_t *pte, unsigned long hmask,
required_fault =
hmm_pte_need_fault(hmm_vma_walk, pfn_req_flags, cpu_flags);
if (required_fault) {
- int ret;
-
spin_unlock(ptl);
- hugetlb_vma_unlock_read(vma);
- /*
- * Avoid deadlock: drop the vma lock before calling
- * hmm_vma_fault(), which will itself potentially take and
- * drop the vma lock. This is also correct from a
- * protection point of view, because there is no further
- * use here of either pte or ptl after dropping the vma
- * lock.
- */
- ret = hmm_vma_fault(addr, end, required_fault, walk);
- hugetlb_vma_lock_read(vma);
- return ret;
+ return hmm_record_fault(addr, end, required_fault, walk);
}
pfn = pte_pfn(entry) + ((start & ~hmask) >> PAGE_SHIFT);
@@ -637,6 +621,44 @@ static const struct mm_walk_ops hmm_walk_ops = {
.walk_lock = PGWALK_RDLOCK,
};
+/*
+ * hmm_do_fault - fault in a range recorded by a walk callback
+ *
+ * Called from the outer loop in hmm_range_fault() after a callback
+ * returned HMM_FAULT_PENDING. At this point we hold only mmap_lock;
+ * the page-table spinlock and any hugetlb_vma_lock acquired by the walk
+ * framework have already been released by the unwind.
+ *
+ * Returns -EBUSY on success (all pages faulted, caller should re-walk).
+ * Returns a negative errno on failure.
+ */
+static int hmm_do_fault(struct mm_struct *mm,
+ struct hmm_vma_walk *hmm_vma_walk)
+{
+ unsigned long addr = hmm_vma_walk->last;
+ unsigned long end = hmm_vma_walk->end;
+ unsigned int required_fault = hmm_vma_walk->required_fault;
+ unsigned int fault_flags = FAULT_FLAG_REMOTE;
+ struct vm_area_struct *vma;
+
+ vma = vma_lookup(mm, addr);
+ if (!vma)
+ return -EFAULT;
+
+ if (required_fault & HMM_NEED_WRITE_FAULT) {
+ if (!(vma->vm_flags & VM_WRITE))
+ return -EPERM;
+ fault_flags |= FAULT_FLAG_WRITE;
+ }
+
+ for (; addr < end; addr += PAGE_SIZE)
+ if (handle_mm_fault(vma, addr, fault_flags, NULL) &
+ VM_FAULT_ERROR)
+ return -EFAULT;
+
+ return -EBUSY;
+}
+
/**
* hmm_range_fault - try to fault some address in a virtual address range
* @range: argument structure
@@ -674,6 +696,16 @@ int hmm_range_fault(struct hmm_range *range)
return -EBUSY;
ret = walk_page_range(mm, hmm_vma_walk.last, range->end,
&hmm_walk_ops, &hmm_vma_walk);
+ /*
+ * When HMM_FAULT_PENDING is returned a walk callback
+ * recorded a range that needs handle_mm_fault();
+ * hmm_do_fault() runs the fault outside walk_page_range()
+ * (so no page-table or hugetlb_vma_lock is held) and
+ * returns -EBUSY so the loop re-walks and picks up the
+ * now-present entries.
+ */
+ if (ret == HMM_FAULT_PENDING)
+ ret = hmm_do_fault(mm, &hmm_vma_walk);
/*
* When -EBUSY is returned the loop restarts with
* hmm_vma_walk.last set to an address that has not been stored
^ permalink raw reply related
* [PATCH v9 0/8] mm/hmm: Add mmap lock-drop support for userfaultfd-backed mappings
From: Stanislav Kinsburskii @ 2026-07-15 18:15 UTC (permalink / raw)
To: airlied, akhilesh, akpm, corbet, dakr, david, decui, haiyangz,
jgg, kees, kys, leon, liam, lizhi.hou, ljs, longli, lyude,
maarten.lankhorst, mamin506, mhocko, mripard, nouveau, ogabbay,
oleg, rppt, shuah, simona, skhan, skinsburskii, surenb,
tzimmermann, vbabka, wei.liu, skinsburskii
Cc: dri-devel, intel-xe, linux-mm, linux-doc, linux-hyperv,
linux-kernel, linux-kselftest, linux-rdma
This series extends the HMM framework to support userfaultfd-backed memory
by allowing the mmap read lock to be dropped during hmm_range_fault().
Some page fault handlers — most notably userfaultfd — require the mmap lock
to be released so that userspace can resolve the fault. The current HMM
interface never sets FAULT_FLAG_ALLOW_RETRY, making it impossible to fault
in pages from userfaultfd-registered regions.
This series follows the established int *locked pattern from
get_user_pages_remote() in mm/gup.c. A new helper function,
hmm_range_fault_locked(), accepts an int *locked parameter. When the
mmap lock is dropped during fault resolution (VM_FAULT_RETRY or
VM_FAULT_COMPLETED), the function returns 0 with *locked = 0, signalling
the caller to restart its walk. The existing hmm_range_fault() is
refactored into a thin wrapper that passes NULL, preserving current
behavior for all existing callers.
Possible approaches to lift this limitation are documented in
Documentation/mm/hmm.rst.
Changes in v9:
- Folded the fixups into the full 8-patch series instead of sending a
separate fixup series.
- Clarified that the HMM timeout bounds repeated HMM/mmu-notifier retry
attempts, with the helper refreshing range->notifier_seq internally.
- Kept nouveau’s explicit outer deadline because its retry loop runs in a GPU
fault worker and cannot rely on fatal signals from the faulting process.
- Converted amdxdna, and GPU SVM callers to pass the shole timeout budget to
hmm_range_fault_unlocked_timeout().
Changes in v8:
- Make hmm_range_fault_unlocked_timeout() the primary documented HMM
range-fault API, and move hmm_range_fault() into the “use only if the
caller really must hold mmap_lock” category.
- Clarify that the timeout is a retry budget for repeated mmu-notifier
invalidation retries. HMM does not interrupt an in-progress page fault
when the timeout expires.
- Restart the retry timeout only when handle_mm_fault() dropped
mmap_lock, because that indicates a lock-dropping fault handler such as
userfaultfd made progress. Ordinary -EBUSY retries keep the existing
deadline.
- Remove the attempted timeout selftest. The remaining selftest covers the
intended userfaultfd path by resolving missing-page faults through
HMM_DMIRROR_READ_UNLOCKED and hmm_range_fault_unlocked_timeout(..., 0).
Changes in v7:
- Replaced the unlocked HMM API with
hmm_range_fault_unlocked_timeout(). The helper now takes a timeout in
jiffies, with 0 meaning retry indefinitely.
- Moved -EBUSY retry handling into the HMM helper for the unlocked path.
The helper refreshes range->notifier_seq internally before each retry.
- Switched the unlocked path to mmap_read_lock_killable() and return
-EINTR if mmap lock acquisition is interrupted or a fatal signal is
pending during retry handling.
- Removed the redundant non-timeout hmm_range_fault_unlocked() interface.
- Updated Documentation/mm/hmm.rst and kernel-doc to describe the timeout API
and the intended caller pattern.
- Updated the HMM selftests to use hmm_range_fault_unlocked_timeout()
only, including coverage for the finite-timeout path.
- Added in-tree users of the new helper:
- mshv
- nouveau
- RDMA/umem
- amdxdna
- drm/gpusvm
- Preserved each converted driver’s existing timeout convention:
- unbounded retry where the old code retried indefinitely,
- HMM_RANGE_DEFAULT_TIMEOUT where the old code used that budget,
- existing driver-specific timeout return values such as -ETIME.
- Used max_t(long, timeout - jiffies, 1) when passing remaining time from
absolute jiffies deadlines to avoid unsigned underflow while keeping a
minimum one-jiffy retry window.
- Left callers on hmm_range_fault() when they already need to hold
mmap_lock across surrounding work, such as drm_gpusvm_check_pages().
Changes in v6:
- Reworked the new API from the external int *locked pattern to
hmm_range_fault_unlocked(), which owns mmap_read_lock() internally.
- Changed the dropped-lock contract: hmm_range_fault_unlocked() now returns
-EBUSY when the mmap lock is dropped, and callers restart with a fresh
mmu_interval_read_begin() sequence.
- Kept hmm_range_fault() as the locked variant for existing users, preserving
its caller-held mmap lock contract.
- Added an in-tree user by converting the MSHV region fault path to
hmm_range_fault_unlocked().
- Updated Documentation/mm/hmm.rst and kernel-doc to describe the unlocked
helper and retry pattern.
- Updated commit messages to match the new API and return semantics.
- Kept the userfaultfd HMM selftest using the test_hmm unlocked read ioctl
path.
Changes in v5:
- Rework hmm_range_fault_unlockable() retry handling to retry
VM_FAULT_RETRY internally with FAULT_FLAG_TRIED set, matching the
fixup_user_fault() pattern and avoiding repeated first-retry lock drops.
- Distinguish VM_FAULT_RETRY from VM_FAULT_COMPLETED: retry faults now
reacquire the mmap lock internally, while completed faults return to the
caller with *locked = 0 so the caller can restart with a fresh notifier
sequence.
- Document the two *locked return states, including the -EINTR case when a
fatal signal is pending after the mmap lock has already been dropped.
- Update comments around HMM_FAULT_UNLOCKED and the HMM fault loop to match
the current hmm_range_fault_unlockable() implementation.
Changes in v4:
- Rebased on 7.2-rc1
Changes in v3:
- Return -EFAULT from dmirror_fault_unlockable() when the mirrored mm can
no longer be pinned.
- Add an eventfd stop signal for the userfaultfd handler thread to avoid
waiting for the poll timeout on successful test completion.
Changes in v2:
- Split into a preparatory refactor (new patch 1) that moves
handle_mm_fault() out of the walk callbacks, plus a smaller feature
patch on top. Suggested by David Hildenbrand.
- Hugetlb regions are now supported on the unlockable path; the v1
-EFAULT short-circuit and the hugetlb_vma_lock_read drop/retake
dance are gone.
- Distinct internal sentinels for "needs fault" (HMM_FAULT_PENDING)
and "lock dropped" (HMM_FAULT_UNLOCKED).
- Outer loop now re-walks after a successful internal fault so the
faulted pfns end up in range->hmm_pfns.
- Kernel-doc on hmm_range_fault_unlockable() and the
Documentation/mm/hmm.rst example match the implementation.
- Dropped the mshv driver conversion (v1 patch 2); will post
separately.
- Selftest converted to drive the path through test_hmm with a
userfaultfd handler (new HMM_DMIRROR_READ_UNLOCKABLE ioctl).
---
Stanislav Kinsburskii (8):
mm/hmm: move page fault handling out of walk callbacks
mm/hmm: add hmm_range_fault_unlocked_timeout() for mmap lock-drop support
selftests/mm: add HMM test for mmap lock-dropping faults
mshv: Use hmm_range_fault_unlocked_timeout() for region faults
drm/nouveau: Use hmm_range_fault_unlocked_timeout() for SVM faults
RDMA/umem: Use hmm_range_fault_unlocked_timeout() for ODP faults
accel/amdxdna: Use hmm_range_fault_unlocked_timeout() for range population
drm/gpusvm: Use hmm_range_fault_unlocked_timeout() for range faults
Documentation/mm/hmm.rst | 79 +++++++---
drivers/accel/amdxdna/aie2_ctx.c | 23 ---
drivers/gpu/drm/drm_gpusvm.c | 61 +-------
drivers/gpu/drm/nouveau/nouveau_svm.c | 20 +-
drivers/hv/mshv_regions.c | 54 +------
drivers/infiniband/core/umem_odp.c | 18 +-
include/linux/hmm.h | 2
lib/test_hmm.c | 107 +++++++++++++
lib/test_hmm_uapi.h | 1
mm/hmm.c | 259 +++++++++++++++++++++++++-------
tools/testing/selftests/mm/hmm-tests.c | 150 +++++++++++++++++++
11 files changed, 551 insertions(+), 223 deletions(-)
^ permalink raw reply
* [PATCH net-next v5 15/15] ixd: add devlink support
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Amritha Nambiar, larysa.zaremba, przemyslaw.kitszel,
aleksander.lobakin, sridhar.samudrala, michal.swiatkowski,
maciej.fijalkowski, emil.s.tantilov, madhu.chittim, joshua.a.hay,
jacob.e.keller, jayaprakash.shanmugam, jiri, horms, corbet,
richardcochran, skhan, linux-doc, Bharath R
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Amritha Nambiar <amritha.nambiar@intel.com>
Enable initial support for the devlink interface with the ixd driver. The
ixd hardware is a single function PCIe device. So, the PCIe adapter gets
its own devlink instance to manage device-wide resources or configuration.
$ devlink dev show
pci/0000:83:00.6
$ devlink dev info pci/0000:83:00.6
pci/0000:83:00.6:
driver ixd
serial_number 00-a0-c9-ff-ff-23-45-67
versions:
fixed:
device.type MEV
running:
virtchnl 2.0
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
Documentation/networking/devlink/index.rst | 1 +
Documentation/networking/devlink/ixd.rst | 30 ++++++
drivers/net/ethernet/intel/ixd/Kconfig | 1 +
drivers/net/ethernet/intel/ixd/Makefile | 1 +
drivers/net/ethernet/intel/ixd/ixd.h | 1 +
drivers/net/ethernet/intel/ixd/ixd_devlink.c | 97 ++++++++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_devlink.h | 50 ++++++++++
drivers/net/ethernet/intel/ixd/ixd_lib.c | 3 +
drivers/net/ethernet/intel/ixd/ixd_main.c | 14 ++-
9 files changed, 195 insertions(+), 3 deletions(-)
create mode 100644 Documentation/networking/devlink/ixd.rst
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_devlink.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_devlink.h
diff --git a/Documentation/networking/devlink/index.rst b/Documentation/networking/devlink/index.rst
index 4745148fecf4..e371be30b99e 100644
--- a/Documentation/networking/devlink/index.rst
+++ b/Documentation/networking/devlink/index.rst
@@ -88,6 +88,7 @@ parameters, info versions, and other features it supports.
ionic
iosm
ixgbe
+ ixd
kvaser_pciefd
kvaser_usb
mlx4
diff --git a/Documentation/networking/devlink/ixd.rst b/Documentation/networking/devlink/ixd.rst
new file mode 100644
index 000000000000..17b63c8425aa
--- /dev/null
+++ b/Documentation/networking/devlink/ixd.rst
@@ -0,0 +1,30 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+===================
+ixd devlink support
+===================
+
+This document describes the devlink features implemented by the ``ixd``
+device driver.
+
+Info versions
+=============
+
+The ``ixd`` driver reports the following versions
+
+.. list-table:: devlink info versions implemented
+ :widths: 5 5 5 90
+
+ * - Name
+ - Type
+ - Example
+ - Description
+ * - ``device.type``
+ - fixed
+ - MEV
+ - The hardware type for this device
+ * - ``virtchnl``
+ - running
+ - 2.0
+ - 2-digit version number (major.minor) of the communication channel
+ (virtchnl) used by the device.
diff --git a/drivers/net/ethernet/intel/ixd/Kconfig b/drivers/net/ethernet/intel/ixd/Kconfig
index 24510c50070e..34181c59dcdc 100644
--- a/drivers/net/ethernet/intel/ixd/Kconfig
+++ b/drivers/net/ethernet/intel/ixd/Kconfig
@@ -7,6 +7,7 @@ config IXD
select LIBETH
select LIBIE_CP
select LIBIE_PCI
+ select NET_DEVLINK
help
This driver supports Intel(R) Control Plane PCI Function
of Intel E2100 and later IPUs and FNICs.
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
index 90abf231fb16..03760a2580b9 100644
--- a/drivers/net/ethernet/intel/ixd/Makefile
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -8,5 +8,6 @@ obj-$(CONFIG_IXD) += ixd.o
ixd-y := ixd_main.o
ixd-y += ixd_ctlq.o
ixd-y += ixd_dev.o
+ixd-y += ixd_devlink.o
ixd-y += ixd_lib.o
ixd-y += ixd_virtchnl.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd.h b/drivers/net/ethernet/intel/ixd/ixd.h
index 8e3cd5bc5a84..343c5053045f 100644
--- a/drivers/net/ethernet/intel/ixd/ixd.h
+++ b/drivers/net/ethernet/intel/ixd/ixd.h
@@ -23,6 +23,7 @@ struct ixd_adapter {
struct delayed_work init_work;
u8 reset_retries;
u8 vc_retries;
+ bool success;
} init_task;
struct delayed_work mbx_task;
struct libie_ctlq_xn_manager *xnm;
diff --git a/drivers/net/ethernet/intel/ixd/ixd_devlink.c b/drivers/net/ethernet/intel/ixd/ixd_devlink.c
new file mode 100644
index 000000000000..23ab11226978
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_devlink.c
@@ -0,0 +1,97 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2025, Intel Corporation. */
+
+#include "ixd.h"
+#include "ixd_devlink.h"
+
+#define IXD_DEVLINK_INFO_LEN 128
+
+/**
+ * ixd_fill_dsn - Get the serial number for the ixd device
+ * @adapter: adapter to query
+ * @buf: storage buffer for the info request
+ */
+static void ixd_fill_dsn(struct ixd_adapter *adapter, char *buf)
+{
+ u8 dsn[8];
+
+ /* Copy the DSN into an array in Big Endian format */
+ put_unaligned_be64(pci_get_dsn(adapter->cp_ctx.mmio_info.pdev), dsn);
+
+ snprintf(buf, IXD_DEVLINK_INFO_LEN, "%8phD", dsn);
+}
+
+/**
+ * ixd_fill_device_name - Get the name of the underlying hardware
+ * @adapter: adapter to query
+ * @buf: storage buffer for the info request
+ * @buf_size: size of the storage buffer
+ */
+static void ixd_fill_device_name(struct ixd_adapter *adapter, char *buf,
+ size_t buf_size)
+{
+ if (adapter->caps.device_type == cpu_to_le32(VIRTCHNL2_MEV_DEVICE))
+ snprintf(buf, buf_size, "%s", "MEV");
+ else
+ snprintf(buf, buf_size, "%s", "UNKNOWN");
+}
+
+/**
+ * ixd_devlink_info_get - .info_get devlink handler
+ * @devlink: devlink instance structure
+ * @req: the devlink info request
+ * @extack: extended netdev ack structure
+ *
+ * Callback for the devlink .info_get operation. Reports information about the
+ * device.
+ *
+ * Return: zero on success or an error code on failure.
+ */
+static int ixd_devlink_info_get(struct devlink *devlink,
+ struct devlink_info_req *req,
+ struct netlink_ext_ack *extack)
+{
+ struct ixd_adapter *adapter = devlink_priv(devlink);
+ char buf[IXD_DEVLINK_INFO_LEN];
+ int err;
+
+ ixd_fill_dsn(adapter, buf);
+ err = devlink_info_serial_number_put(req, buf);
+ if (err)
+ return err;
+
+ ixd_fill_device_name(adapter, buf, IXD_DEVLINK_INFO_LEN);
+ err = devlink_info_version_fixed_put(req, "device.type", buf);
+ if (err)
+ return err;
+
+ snprintf(buf, sizeof(buf), "%u.%u",
+ adapter->vc_ver.major, adapter->vc_ver.minor);
+
+ return devlink_info_version_running_put(req, "virtchnl", buf);
+}
+
+static const struct devlink_ops ixd_devlink_ops = {
+ .info_get = ixd_devlink_info_get,
+};
+
+/**
+ * ixd_adapter_alloc - Allocate devlink and return adapter pointer
+ * @dev: the device to allocate for
+ *
+ * Allocate a devlink instance for this device and return the private area as
+ * the adapter structure.
+ *
+ * Return: adapter structure on success, NULL on failure
+ */
+struct ixd_adapter *ixd_adapter_alloc(struct device *dev)
+{
+ struct devlink *devlink;
+
+ devlink = devlink_alloc(&ixd_devlink_ops, sizeof(struct ixd_adapter),
+ dev);
+ if (!devlink)
+ return NULL;
+
+ return devlink_priv(devlink);
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_devlink.h b/drivers/net/ethernet/intel/ixd/ixd_devlink.h
new file mode 100644
index 000000000000..b23a1b37aebc
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_devlink.h
@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (c) 2025, Intel Corporation. */
+
+#ifndef _IXD_DEVLINK_H_
+#define _IXD_DEVLINK_H_
+
+#include <net/devlink.h>
+
+#include "ixd.h"
+
+struct ixd_adapter *ixd_adapter_alloc(struct device *dev);
+
+/**
+ * ixd_devlink_free - teardown the devlink
+ * @adapter: the adapter structure to free
+ *
+ */
+static inline void ixd_devlink_free(struct ixd_adapter *adapter)
+{
+ struct devlink *devlink = priv_to_devlink(adapter);
+
+ devlink_free(devlink);
+}
+
+/**
+ * ixd_devlink_unregister - Unregister devlink for this adapter.
+ * @adapter: the adapter structure to cleanup
+ *
+ * Init task must be completed or cancelled beforehand.
+ */
+static inline void ixd_devlink_unregister(struct ixd_adapter *adapter)
+{
+ if (!adapter->init_task.success)
+ return;
+
+ devlink_unregister(priv_to_devlink(adapter));
+}
+
+/**
+ * ixd_devlink_register - Register devlink interface for this adapter
+ * @adapter: pointer to ixd adapter structure to be associated with devlink
+ *
+ * Register the devlink instance associated with this adapter
+ */
+static inline void ixd_devlink_register(struct ixd_adapter *adapter)
+{
+ devlink_register(priv_to_devlink(adapter));
+}
+
+#endif /* _IXD_DEVLINK_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lib.c b/drivers/net/ethernet/intel/ixd/ixd_lib.c
index ce870aba143d..6f2b39cd985c 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_lib.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_lib.c
@@ -3,6 +3,7 @@
#include "ixd.h"
#include "ixd_ctlq.h"
+#include "ixd_devlink.h"
#include "ixd_virtchnl.h"
#define IXD_DFLT_MBX_Q_LEN 64
@@ -154,6 +155,8 @@ void ixd_init_task(struct work_struct *work)
if (!ixd_vc_dev_init(adapter)) {
adapter->init_task.vc_retries = 0;
+ adapter->init_task.success = true;
+ ixd_devlink_register(adapter);
return;
}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
index eb8e410300c2..973bc480e25a 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_main.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -4,6 +4,7 @@
#include "ixd.h"
#include "ixd_ctlq.h"
#include "ixd_lan_regs.h"
+#include "ixd_devlink.h"
MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
MODULE_IMPORT_NS("LIBIE_CP");
@@ -21,6 +22,8 @@ static void ixd_remove(struct pci_dev *pdev)
/* Do not mix removal with (re)initialization */
cancel_delayed_work_sync(&adapter->init_task.init_work);
+ ixd_devlink_unregister(adapter);
+
/* Leave the device clean on exit */
if (adapter->xnm)
libie_ctlq_xn_shutdown(adapter->xnm);
@@ -28,6 +31,7 @@ static void ixd_remove(struct pci_dev *pdev)
ixd_deinit_dflt_mbx(adapter);
libie_pci_unmap_all_mmio_regions(&adapter->cp_ctx.mmio_info);
+ ixd_devlink_free(adapter);
}
/**
@@ -95,7 +99,7 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
if (WARN_ON(ent->device != IXD_DEV_ID_CPF))
return -EINVAL;
- adapter = devm_kzalloc(&pdev->dev, sizeof(*adapter), GFP_KERNEL);
+ adapter = ixd_adapter_alloc(&pdev->dev);
if (!adapter)
return -ENOMEM;
@@ -104,13 +108,13 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
err = libie_pci_init_dev(pdev);
if (err)
- return err;
+ goto free_adapter;
pci_set_drvdata(pdev, adapter);
err = ixd_iomap_regions(adapter);
if (err)
- return err;
+ goto free_adapter;
INIT_DELAYED_WORK(&adapter->init_task.init_work,
ixd_init_task);
@@ -121,6 +125,10 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
msecs_to_jiffies(500));
return 0;
+
+free_adapter:
+ ixd_devlink_free(adapter);
+ return err;
}
static const struct pci_device_id ixd_pci_tbl[] = {
--
2.47.1
^ permalink raw reply related
* [PATCH net-next v5 14/15] ixd: add the core initialization
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Larysa Zaremba, anthony.l.nguyen, przemyslaw.kitszel,
aleksander.lobakin, sridhar.samudrala, michal.swiatkowski,
maciej.fijalkowski, emil.s.tantilov, madhu.chittim, joshua.a.hay,
jacob.e.keller, jayaprakash.shanmugam, jiri, horms, corbet,
richardcochran, skhan, linux-doc, Bharath R
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Larysa Zaremba <larysa.zaremba@intel.com>
As the mailbox is setup, initialize the core. This makes use of the send
and receive mailbox message framework for virtchnl communication between
the driver and device Control Plane (CP).
To start with, driver confirms the virtchnl version with the CP. Once that
is done, it requests and gets the required capabilities and resources
needed such as max vectors, queues, vports etc.
Use a unified way of handling the virtchnl messages, where a single
function handles all related memory management and the caller only provides
the callbacks to fill the send buffer and to handle the response.
Place generic control queue message handling separately to facilitate the
addition of protocols other than virtchannel in the future.
Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/ixd/Makefile | 2 +
drivers/net/ethernet/intel/ixd/ixd.h | 10 +
drivers/net/ethernet/intel/ixd/ixd_ctlq.c | 146 ++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_ctlq.h | 34 ++++
drivers/net/ethernet/intel/ixd/ixd_lib.c | 30 ++-
drivers/net/ethernet/intel/ixd/ixd_main.c | 3 +
drivers/net/ethernet/intel/ixd/ixd_virtchnl.c | 190 ++++++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_virtchnl.h | 12 ++
8 files changed, 426 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_ctlq.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_ctlq.h
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_virtchnl.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_virtchnl.h
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
index 164b2c86952f..90abf231fb16 100644
--- a/drivers/net/ethernet/intel/ixd/Makefile
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -6,5 +6,7 @@
obj-$(CONFIG_IXD) += ixd.o
ixd-y := ixd_main.o
+ixd-y += ixd_ctlq.o
ixd-y += ixd_dev.o
ixd-y += ixd_lib.o
+ixd-y += ixd_virtchnl.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd.h b/drivers/net/ethernet/intel/ixd/ixd.h
index c86c2b05c5b4..8e3cd5bc5a84 100644
--- a/drivers/net/ethernet/intel/ixd/ixd.h
+++ b/drivers/net/ethernet/intel/ixd/ixd.h
@@ -10,19 +10,29 @@
* struct ixd_adapter - Data structure representing a CPF
* @cp_ctx: Control plane communication context
* @init_task: Delayed initialization after reset
+ * @mbx_task: Control queue Rx handling
* @xnm: virtchnl transaction manager
* @asq: Send control queue info
* @arq: Receive control queue info
+ * @vc_ver: Negotiated virtchnl version
+ * @caps: Negotiated virtchnl capabilities
*/
struct ixd_adapter {
struct libie_ctlq_ctx cp_ctx;
struct {
struct delayed_work init_work;
u8 reset_retries;
+ u8 vc_retries;
} init_task;
+ struct delayed_work mbx_task;
struct libie_ctlq_xn_manager *xnm;
struct libie_ctlq_info *asq;
struct libie_ctlq_info *arq;
+ struct {
+ u32 major;
+ u32 minor;
+ } vc_ver;
+ struct virtchnl2_get_capabilities caps;
};
/**
diff --git a/drivers/net/ethernet/intel/ixd/ixd_ctlq.c b/drivers/net/ethernet/intel/ixd/ixd_ctlq.c
new file mode 100644
index 000000000000..6d658be0695d
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_ctlq.c
@@ -0,0 +1,146 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_ctlq.h"
+#include "ixd_virtchnl.h"
+
+/**
+ * ixd_ctlq_clean_sq - Clean the send control queue after sending the message
+ * @adapter: The adapter that sent the messages
+ * @force: Clean regardless of send status
+ *
+ * Free the libie send resources after sending the message and handling
+ * the response.
+ */
+void ixd_ctlq_clean_sq(struct ixd_adapter *adapter, bool force)
+{
+ struct libie_ctlq_xn_clean_params params = {
+ .ctlq = adapter->asq,
+ .force = force,
+ .num_msgs = adapter->asq->ring_len,
+ .rel_tx_buf = kfree,
+ };
+
+ libie_ctlq_xn_send_clean(¶ms);
+}
+
+/**
+ * ixd_ctlq_init_sparams - Initialize control queue send parameters
+ * @adapter: The adapter with initialized mailbox
+ * @sparams: Parameters to initialize
+ * @msg_buf: DMA-mappable pointer to the message being sent
+ * @msg_size: Message size
+ */
+static void ixd_ctlq_init_sparams(struct ixd_adapter *adapter,
+ struct libie_ctlq_xn_send_params *sparams,
+ void *msg_buf, size_t msg_size)
+{
+ *sparams = (struct libie_ctlq_xn_send_params) {
+ .rel_tx_buf = kfree,
+ .xnm = adapter->xnm,
+ .ctlq = adapter->asq,
+ .timeout_ms = IXD_CTLQ_TIMEOUT,
+ .send_buf = (struct kvec) {
+ .iov_base = msg_buf,
+ .iov_len = msg_size,
+ },
+ };
+}
+
+/**
+ * ixd_ctlq_do_req - Perform a standard virtchnl request
+ * @adapter: The adapter with initialized mailbox
+ * @req: virtchnl request description
+ *
+ * Return: %0 if a message was sent and received a response
+ * that was successfully handled by the custom callback,
+ * negative error otherwise.
+ */
+int ixd_ctlq_do_req(struct ixd_adapter *adapter, const struct ixd_ctlq_req *req)
+{
+ u8 onstack_send_buff[LIBIE_CP_TX_COPYBREAK] __aligned_largest = {};
+ struct libie_ctlq_xn_send_params send_params = {};
+ struct kvec *recv_mem;
+ void *send_buff;
+ int err;
+
+ send_buff = libie_cp_can_send_onstack(req->send_size) ?
+ &onstack_send_buff : kzalloc(req->send_size, GFP_KERNEL);
+ if (!send_buff)
+ return -ENOMEM;
+
+ ixd_ctlq_init_sparams(adapter, &send_params, send_buff,
+ req->send_size);
+
+ send_params.chnl_opcode = req->opcode;
+
+ if (req->send_buff_init)
+ req->send_buff_init(adapter, send_buff, req->ctx);
+
+ ixd_ctlq_clean_sq(adapter, false);
+ err = libie_ctlq_xn_send(&send_params);
+ if (err)
+ return err;
+
+ recv_mem = &send_params.recv_mem;
+ if (req->recv_process)
+ err = req->recv_process(adapter, recv_mem->iov_base,
+ recv_mem->iov_len, req->ctx);
+
+ libie_ctlq_release_rx_buf(recv_mem);
+
+ return err;
+}
+
+/**
+ * ixd_ctlq_handle_msg - Default control queue message handler
+ * @ctx: Control plane communication context
+ * @msg: Message received
+ */
+static void ixd_ctlq_handle_msg(struct libie_ctlq_ctx *ctx,
+ struct libie_ctlq_msg *msg)
+{
+ struct ixd_adapter *adapter = pci_get_drvdata(ctx->mmio_info.pdev);
+
+ if (ixd_vc_can_handle_msg(msg))
+ ixd_vc_recv_event_msg(adapter, msg);
+ else
+ dev_dbg_ratelimited(ixd_to_dev(adapter),
+ "Received an unsupported opcode 0x%x from the CP\n",
+ msg->chnl_opcode);
+
+ libie_ctlq_release_rx_buf(&msg->recv_mem);
+}
+
+/**
+ * ixd_ctlq_recv_mb_msg - Receive a potential message over mailbox periodically
+ * @adapter: The adapter with initialized mailbox
+ */
+static void ixd_ctlq_recv_mb_msg(struct ixd_adapter *adapter)
+{
+ struct libie_ctlq_xn_recv_params xn_params = {
+ .xnm = adapter->xnm,
+ .ctlq = adapter->arq,
+ .ctlq_msg_handler = ixd_ctlq_handle_msg,
+ .budget = LIBIE_CTLQ_MAX_XN_ENTRIES,
+ };
+
+ libie_ctlq_xn_recv(&xn_params);
+}
+
+/**
+ * ixd_ctlq_rx_task - Periodically check for mailbox responses and events
+ * @work: work handle
+ */
+void ixd_ctlq_rx_task(struct work_struct *work)
+{
+ struct ixd_adapter *adapter;
+
+ adapter = container_of(work, struct ixd_adapter, mbx_task.work);
+
+ queue_delayed_work(system_unbound_wq, &adapter->mbx_task,
+ msecs_to_jiffies(300));
+
+ ixd_ctlq_recv_mb_msg(adapter);
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_ctlq.h b/drivers/net/ethernet/intel/ixd/ixd_ctlq.h
new file mode 100644
index 000000000000..8839f9f8f6d5
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_ctlq.h
@@ -0,0 +1,34 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_CTLQ_H_
+#define _IXD_CTLQ_H_
+
+#include <linux/net/intel/virtchnl2.h>
+
+#define IXD_CTLQ_TIMEOUT 2000
+
+/**
+ * struct ixd_ctlq_req - Standard virtchnl request description
+ * @opcode: protocol opcode, only virtchnl2 is needed for now
+ * @send_size: required length of the send buffer
+ * @send_buff_init: function to initialize the allocated send buffer
+ * @recv_process: function to handle the CP response
+ * @ctx: additional context for callbacks
+ */
+struct ixd_ctlq_req {
+ enum virtchnl2_op opcode;
+ size_t send_size;
+ void (*send_buff_init)(struct ixd_adapter *adapter, void *send_buff,
+ void *ctx);
+ int (*recv_process)(struct ixd_adapter *adapter, void *recv_buff,
+ size_t recv_size, void *ctx);
+ void *ctx;
+};
+
+void ixd_ctlq_clean_sq(struct ixd_adapter *adapter, bool force);
+int ixd_ctlq_do_req(struct ixd_adapter *adapter,
+ const struct ixd_ctlq_req *req);
+void ixd_ctlq_rx_task(struct work_struct *work);
+
+#endif /* _IXD_CTLQ_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lib.c b/drivers/net/ethernet/intel/ixd/ixd_lib.c
index afc413d3650f..ce870aba143d 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_lib.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_lib.c
@@ -2,6 +2,8 @@
/* Copyright (C) 2025 Intel Corporation */
#include "ixd.h"
+#include "ixd_ctlq.h"
+#include "ixd_virtchnl.h"
#define IXD_DFLT_MBX_Q_LEN 64
@@ -67,6 +69,11 @@ static void ixd_adapter_fill_dflt_ctlqs(struct ixd_adapter *adapter)
*/
void ixd_deinit_dflt_mbx(struct ixd_adapter *adapter)
{
+ cancel_delayed_work_sync(&adapter->mbx_task);
+
+ if (adapter->asq)
+ ixd_ctlq_clean_sq(adapter, true);
+
if (adapter->xnm)
libie_ctlq_xn_deinit(adapter->xnm, &adapter->cp_ctx);
@@ -108,6 +115,8 @@ int ixd_init_dflt_mbx(struct ixd_adapter *adapter)
return -ENOENT;
}
+ queue_delayed_work(system_unbound_wq, &adapter->mbx_task, 0);
+
return 0;
}
@@ -136,8 +145,27 @@ void ixd_init_task(struct work_struct *work)
adapter->init_task.reset_retries = 0;
err = ixd_init_dflt_mbx(adapter);
- if (err)
+ if (err) {
dev_err(ixd_to_dev(adapter),
"Failed to initialize the default mailbox: %pe\n",
ERR_PTR(err));
+ return;
+ }
+
+ if (!ixd_vc_dev_init(adapter)) {
+ adapter->init_task.vc_retries = 0;
+ return;
+ }
+
+ libie_ctlq_xn_shutdown(adapter->xnm);
+ ixd_trigger_reset(adapter);
+ ixd_deinit_dflt_mbx(adapter);
+ if (++adapter->init_task.vc_retries > 5) {
+ dev_err(ixd_to_dev(adapter),
+ "Failed to establish mailbox communications with the hardware\n");
+ return;
+ }
+
+ queue_delayed_work(system_unbound_wq, &adapter->init_task.init_work,
+ msecs_to_jiffies(500));
}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
index 35d0ff5c8d98..eb8e410300c2 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_main.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -2,6 +2,7 @@
/* Copyright (C) 2025 Intel Corporation */
#include "ixd.h"
+#include "ixd_ctlq.h"
#include "ixd_lan_regs.h"
MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
@@ -19,6 +20,7 @@ static void ixd_remove(struct pci_dev *pdev)
/* Do not mix removal with (re)initialization */
cancel_delayed_work_sync(&adapter->init_task.init_work);
+
/* Leave the device clean on exit */
if (adapter->xnm)
libie_ctlq_xn_shutdown(adapter->xnm);
@@ -112,6 +114,7 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
INIT_DELAYED_WORK(&adapter->init_task.init_work,
ixd_init_task);
+ INIT_DELAYED_WORK(&adapter->mbx_task, ixd_ctlq_rx_task);
ixd_trigger_reset(adapter);
queue_delayed_work(system_unbound_wq, &adapter->init_task.init_work,
diff --git a/drivers/net/ethernet/intel/ixd/ixd_virtchnl.c b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.c
new file mode 100644
index 000000000000..fc3b6d2e28c5
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.c
@@ -0,0 +1,190 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_ctlq.h"
+#include "ixd_virtchnl.h"
+
+/**
+ * ixd_vc_recv_event_msg - Handle virtchnl event message
+ * @adapter: The adapter handling the message
+ * @ctlq_msg: Message received
+ */
+void ixd_vc_recv_event_msg(struct ixd_adapter *adapter,
+ struct libie_ctlq_msg *ctlq_msg)
+{
+ int payload_size = ctlq_msg->data_len;
+ struct virtchnl2_event *v2e;
+
+ if (payload_size < sizeof(*v2e)) {
+ dev_warn_ratelimited(ixd_to_dev(adapter),
+ "Failed to receive valid payload for event msg (op 0x%X len %u)\n",
+ ctlq_msg->chnl_opcode,
+ payload_size);
+ return;
+ }
+
+ v2e = (struct virtchnl2_event *)ctlq_msg->recv_mem.iov_base;
+
+ dev_dbg(ixd_to_dev(adapter), "Got event 0x%X from the CP\n",
+ le32_to_cpu(v2e->event));
+}
+
+/**
+ * ixd_vc_can_handle_msg - Decide if an event has to be handled by virtchnl code
+ * @ctlq_msg: Message received
+ *
+ * Return: %true if virtchnl code can handle the event, %false otherwise
+ */
+bool ixd_vc_can_handle_msg(struct libie_ctlq_msg *ctlq_msg)
+{
+ return ctlq_msg->chnl_opcode == VIRTCHNL2_OP_EVENT;
+}
+
+/**
+ * ixd_handle_caps - Handle VIRTCHNL2_OP_GET_CAPS response
+ * @adapter: The adapter for which the capabilities are being updated
+ * @recv_buff: Buffer containing the response
+ * @recv_size: Response buffer size
+ * @ctx: unused
+ *
+ * Return: %0 if the response format is correct and was handled as expected,
+ * negative error otherwise.
+ */
+static int ixd_handle_caps(struct ixd_adapter *adapter, void *recv_buff,
+ size_t recv_size, void *ctx)
+{
+ if (recv_size < sizeof(adapter->caps))
+ return -EBADMSG;
+
+ adapter->caps = *(typeof(adapter->caps) *)recv_buff;
+
+ return 0;
+}
+
+/**
+ * ixd_req_vc_caps - Request and save device capability
+ * @adapter: The adapter to get the capabilities for
+ *
+ * Return: success or error if sending the get capability message fails
+ */
+static int ixd_req_vc_caps(struct ixd_adapter *adapter)
+{
+ const struct ixd_ctlq_req req = {
+ .opcode = VIRTCHNL2_OP_GET_CAPS,
+ .send_size = sizeof(struct virtchnl2_get_capabilities),
+ .ctx = NULL,
+ .send_buff_init = NULL,
+ .recv_process = ixd_handle_caps,
+ };
+
+ return ixd_ctlq_do_req(adapter, &req);
+}
+
+/**
+ * ixd_get_vc_ver - Get version info from adapter
+ *
+ * Return: filled in virtchannel2 version info, ready for sending
+ */
+static struct virtchnl2_version_info ixd_get_vc_ver(void)
+{
+ return (struct virtchnl2_version_info) {
+ .major = cpu_to_le32(VIRTCHNL2_VERSION_MAJOR_2),
+ .minor = cpu_to_le32(VIRTCHNL2_VERSION_MINOR_0),
+ };
+}
+
+static void ixd_fill_vc_ver(struct ixd_adapter *adapter, void *send_buff,
+ void *ctx)
+{
+ *(struct virtchnl2_version_info *)send_buff = ixd_get_vc_ver();
+}
+
+/**
+ * ixd_handle_vc_ver - Handle VIRTCHNL2_OP_VERSION response
+ * @adapter: The adapter for which the version is being updated
+ * @recv_buff: Buffer containing the response
+ * @recv_size: Response buffer size
+ * @ctx: Unused
+ *
+ * Return: %0 if the response format is correct and was handled as expected,
+ * negative error otherwise.
+ */
+static int ixd_handle_vc_ver(struct ixd_adapter *adapter, void *recv_buff,
+ size_t recv_size, void *ctx)
+{
+ struct virtchnl2_version_info need_ver = ixd_get_vc_ver();
+ struct virtchnl2_version_info *recv_ver;
+
+ if (recv_size < sizeof(need_ver))
+ return -EBADMSG;
+
+ recv_ver = recv_buff;
+ if (le32_to_cpu(need_ver.major) != le32_to_cpu(recv_ver->major) ||
+ le32_to_cpu(need_ver.minor) != le32_to_cpu(recv_ver->minor))
+ dev_warn(ixd_to_dev(adapter),
+ "Virtchnl version does not match (expected %u.%u, received %u.%u)\n",
+ le32_to_cpu(need_ver.major),
+ le32_to_cpu(need_ver.minor),
+ le32_to_cpu(recv_ver->major),
+ le32_to_cpu(recv_ver->minor));
+
+ if (le32_to_cpu(need_ver.major) != le32_to_cpu(recv_ver->major)) {
+ dev_err(ixd_to_dev(adapter),
+ "Device initialization failed due to virtchnl major version mismatch\n");
+ return -EOPNOTSUPP;
+ }
+
+ adapter->vc_ver.major = le32_to_cpu(recv_ver->major);
+ adapter->vc_ver.minor = le32_to_cpu(recv_ver->minor);
+
+ return 0;
+}
+
+/**
+ * ixd_req_vc_version - Request and save Virtchannel2 version
+ * @adapter: The adapter to get the version for
+ *
+ * Return: success or error if sending fails or the response was not as expected
+ */
+static int ixd_req_vc_version(struct ixd_adapter *adapter)
+{
+ const struct ixd_ctlq_req req = {
+ .opcode = VIRTCHNL2_OP_VERSION,
+ .send_size = sizeof(struct virtchnl2_version_info),
+ .ctx = NULL,
+ .send_buff_init = ixd_fill_vc_ver,
+ .recv_process = ixd_handle_vc_ver,
+ };
+
+ return ixd_ctlq_do_req(adapter, &req);
+}
+
+/**
+ * ixd_vc_dev_init - virtchnl device core initialization
+ * @adapter: device information
+ *
+ * Return: %0 on success or error if any step of the initialization fails
+ */
+int ixd_vc_dev_init(struct ixd_adapter *adapter)
+{
+ int err;
+
+ err = ixd_req_vc_version(adapter);
+ if (err) {
+ dev_warn(ixd_to_dev(adapter),
+ "Getting virtchnl version failed, error=%pe\n",
+ ERR_PTR(err));
+ return err;
+ }
+
+ err = ixd_req_vc_caps(adapter);
+ if (err) {
+ dev_warn(ixd_to_dev(adapter),
+ "Getting virtchnl capabilities failed, error=%pe\n",
+ ERR_PTR(err));
+ return err;
+ }
+
+ return err;
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_virtchnl.h b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.h
new file mode 100644
index 000000000000..1a53da8b545c
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.h
@@ -0,0 +1,12 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_VIRTCHNL_H_
+#define _IXD_VIRTCHNL_H_
+
+int ixd_vc_dev_init(struct ixd_adapter *adapter);
+bool ixd_vc_can_handle_msg(struct libie_ctlq_msg *ctlq_msg);
+void ixd_vc_recv_event_msg(struct ixd_adapter *adapter,
+ struct libie_ctlq_msg *ctlq_msg);
+
+#endif /* _IXD_VIRTCHNL_H_ */
--
2.47.1
^ permalink raw reply related
* [PATCH net-next v5 13/15] ixd: add reset checks and initialize the mailbox
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Larysa Zaremba, anthony.l.nguyen, przemyslaw.kitszel,
aleksander.lobakin, sridhar.samudrala, michal.swiatkowski,
maciej.fijalkowski, emil.s.tantilov, madhu.chittim, joshua.a.hay,
jacob.e.keller, jayaprakash.shanmugam, jiri, horms, corbet,
richardcochran, skhan, linux-doc, Aleksandr Loktionov, Bharath R
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Larysa Zaremba <larysa.zaremba@intel.com>
At the end of the probe, trigger hard reset, initialize and schedule the
after-reset task. If the reset is complete in a pre-determined time,
initialize the default mailbox, through which other resources will be
negotiated.
Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/ixd/Kconfig | 1 +
drivers/net/ethernet/intel/ixd/Makefile | 2 +
drivers/net/ethernet/intel/ixd/ixd.h | 28 +++-
drivers/net/ethernet/intel/ixd/ixd_dev.c | 89 +++++++++++
drivers/net/ethernet/intel/ixd/ixd_lan_regs.h | 40 +++++
drivers/net/ethernet/intel/ixd/ixd_lib.c | 143 ++++++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_main.c | 34 ++++-
7 files changed, 328 insertions(+), 9 deletions(-)
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_dev.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_lib.c
diff --git a/drivers/net/ethernet/intel/ixd/Kconfig b/drivers/net/ethernet/intel/ixd/Kconfig
index f5594efe292c..24510c50070e 100644
--- a/drivers/net/ethernet/intel/ixd/Kconfig
+++ b/drivers/net/ethernet/intel/ixd/Kconfig
@@ -5,6 +5,7 @@ config IXD
tristate "Intel(R) Control Plane Function Support"
depends on PCI_MSI
select LIBETH
+ select LIBIE_CP
select LIBIE_PCI
help
This driver supports Intel(R) Control Plane PCI Function
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
index 3849bc240600..164b2c86952f 100644
--- a/drivers/net/ethernet/intel/ixd/Makefile
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -6,3 +6,5 @@
obj-$(CONFIG_IXD) += ixd.o
ixd-y := ixd_main.o
+ixd-y += ixd_dev.o
+ixd-y += ixd_lib.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd.h b/drivers/net/ethernet/intel/ixd/ixd.h
index 1b918c5d31cd..c86c2b05c5b4 100644
--- a/drivers/net/ethernet/intel/ixd/ixd.h
+++ b/drivers/net/ethernet/intel/ixd/ixd.h
@@ -4,14 +4,25 @@
#ifndef _IXD_H_
#define _IXD_H_
-#include <linux/net/intel/libie/pci.h>
+#include <linux/net/intel/libie/controlq.h>
/**
* struct ixd_adapter - Data structure representing a CPF
- * @hw: Device access data
+ * @cp_ctx: Control plane communication context
+ * @init_task: Delayed initialization after reset
+ * @xnm: virtchnl transaction manager
+ * @asq: Send control queue info
+ * @arq: Receive control queue info
*/
struct ixd_adapter {
- struct libie_mmio_info hw;
+ struct libie_ctlq_ctx cp_ctx;
+ struct {
+ struct delayed_work init_work;
+ u8 reset_retries;
+ } init_task;
+ struct libie_ctlq_xn_manager *xnm;
+ struct libie_ctlq_info *asq;
+ struct libie_ctlq_info *arq;
};
/**
@@ -22,7 +33,16 @@ struct ixd_adapter {
*/
static inline struct device *ixd_to_dev(struct ixd_adapter *adapter)
{
- return &adapter->hw.pdev->dev;
+ return &adapter->cp_ctx.mmio_info.pdev->dev;
}
+void ixd_ctlq_reg_init(struct ixd_adapter *adapter,
+ struct libie_ctlq_reg *ctlq_reg_tx,
+ struct libie_ctlq_reg *ctlq_reg_rx);
+void ixd_trigger_reset(struct ixd_adapter *adapter);
+bool ixd_check_reset_complete(struct ixd_adapter *adapter);
+void ixd_init_task(struct work_struct *work);
+int ixd_init_dflt_mbx(struct ixd_adapter *adapter);
+void ixd_deinit_dflt_mbx(struct ixd_adapter *adapter);
+
#endif /* _IXD_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_dev.c b/drivers/net/ethernet/intel/ixd/ixd_dev.c
new file mode 100644
index 000000000000..cdd5477cc1f4
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_dev.c
@@ -0,0 +1,89 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_lan_regs.h"
+
+/**
+ * ixd_ctlq_reg_init - Initialize default mailbox registers
+ * @adapter: PCI device driver-specific private data
+ * @ctlq_reg_tx: Transmit queue registers info to be filled
+ * @ctlq_reg_rx: Receive queue registers info to be filled
+ */
+void ixd_ctlq_reg_init(struct ixd_adapter *adapter,
+ struct libie_ctlq_reg *ctlq_reg_tx,
+ struct libie_ctlq_reg *ctlq_reg_rx)
+{
+ struct libie_mmio_info *mmio_info = &adapter->cp_ctx.mmio_info;
+ *ctlq_reg_tx = (struct libie_ctlq_reg) {
+ .head = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQH),
+ .tail = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQT),
+ .len = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQLEN),
+ .addr_high = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQBAH),
+ .addr_low = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQBAL),
+ .len_mask = PF_FW_ATQLEN_ATQLEN_M,
+ .len_ena_mask = PF_FW_ATQLEN_ATQENABLE_M,
+ .head_mask = PF_FW_ATQH_ATQH_M,
+ };
+
+ *ctlq_reg_rx = (struct libie_ctlq_reg) {
+ .head = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQH),
+ .tail = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQT),
+ .len = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQLEN),
+ .addr_high = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQBAH),
+ .addr_low = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQBAL),
+ .len_mask = PF_FW_ARQLEN_ARQLEN_M,
+ .len_ena_mask = PF_FW_ARQLEN_ARQENABLE_M,
+ .head_mask = PF_FW_ARQH_ARQH_M,
+ };
+}
+
+static const struct ixd_reset_reg ixd_reset_reg = {
+ .rstat = PFGEN_RSTAT,
+ .rstat_m = PFGEN_RSTAT_PFR_STATE_M,
+ .rstat_ok_v = 0b01,
+ .rtrigger = PFGEN_CTRL,
+ .rtrigger_m = PFGEN_CTRL_PFSWR,
+};
+
+/**
+ * ixd_trigger_reset - Trigger PFR reset
+ * @adapter: the device with mapped reset register
+ */
+void ixd_trigger_reset(struct ixd_adapter *adapter)
+{
+ void __iomem *addr;
+ u32 reg_val;
+
+ addr = libie_pci_get_mmio_addr(&adapter->cp_ctx.mmio_info,
+ ixd_reset_reg.rtrigger);
+ reg_val = readl(addr);
+ writel(reg_val | ixd_reset_reg.rtrigger_m, addr);
+}
+
+/**
+ * ixd_check_reset_complete - Check if the PFR reset is completed
+ * @adapter: CPF being reset
+ *
+ * Return: %true if the register read indicates reset has been finished,
+ * %false otherwise
+ */
+bool ixd_check_reset_complete(struct ixd_adapter *adapter)
+{
+ u32 reg_val, reset_status;
+ void __iomem *addr;
+
+ addr = libie_pci_get_mmio_addr(&adapter->cp_ctx.mmio_info,
+ ixd_reset_reg.rstat);
+ reg_val = readl(addr);
+ reset_status = reg_val & ixd_reset_reg.rstat_m;
+
+ /* 0xFFFFFFFF might be read if the other side hasn't cleared
+ * the register for us yet.
+ */
+ if (reg_val != GENMASK(31, 0) &&
+ reset_status == ixd_reset_reg.rstat_ok_v)
+ return true;
+
+ return false;
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
index fbb88929d0de..58e58c75981b 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
+++ b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
@@ -11,9 +11,33 @@
#define PF_FW_MBX_REG_LEN 4096
#define PF_FW_MBX 0x08400000
+#define PF_FW_ARQBAL (PF_FW_MBX)
+#define PF_FW_ARQBAH (PF_FW_MBX + 0x4)
+#define PF_FW_ARQLEN (PF_FW_MBX + 0x8)
+#define PF_FW_ARQLEN_ARQLEN_M GENMASK(12, 0)
+#define PF_FW_ARQLEN_ARQENABLE_S 31
+#define PF_FW_ARQLEN_ARQENABLE_M BIT(PF_FW_ARQLEN_ARQENABLE_S)
+#define PF_FW_ARQH_ARQH_M GENMASK(12, 0)
+#define PF_FW_ARQH (PF_FW_MBX + 0xC)
+#define PF_FW_ARQT (PF_FW_MBX + 0x10)
+
+#define PF_FW_ATQBAL (PF_FW_MBX + 0x14)
+#define PF_FW_ATQBAH (PF_FW_MBX + 0x18)
+#define PF_FW_ATQLEN (PF_FW_MBX + 0x1C)
+#define PF_FW_ATQLEN_ATQLEN_M GENMASK(9, 0)
+#define PF_FW_ATQLEN_ATQENABLE_S 31
+#define PF_FW_ATQLEN_ATQENABLE_M BIT(PF_FW_ATQLEN_ATQENABLE_S)
+#define PF_FW_ATQH_ATQH_M GENMASK(9, 0)
+#define PF_FW_ATQH (PF_FW_MBX + 0x20)
+#define PF_FW_ATQT (PF_FW_MBX + 0x24)
+
/* Reset registers */
#define PFGEN_RTRIG_REG_LEN 2048
#define PFGEN_RTRIG 0x08407000 /* Device resets */
+#define PFGEN_RSTAT 0x08407008 /* PFR status */
+#define PFGEN_RSTAT_PFR_STATE_M GENMASK(1, 0)
+#define PFGEN_CTRL 0x0840700C /* PFR trigger */
+#define PFGEN_CTRL_PFSWR BIT(0)
/**
* struct ixd_bar_region - BAR region description
@@ -25,4 +49,20 @@ struct ixd_bar_region {
resource_size_t size;
};
+/**
+ * struct ixd_reset_reg - structure for reset registers
+ * @rstat: offset of status in register
+ * @rstat_m: status mask
+ * @rstat_ok_v: value that indicates PFR completed status
+ * @rtrigger: offset of reset trigger in register
+ * @rtrigger_m: reset trigger mask
+ */
+struct ixd_reset_reg {
+ u32 rstat;
+ u32 rstat_m;
+ u32 rstat_ok_v;
+ u32 rtrigger;
+ u32 rtrigger_m;
+};
+
#endif /* _IXD_LAN_REGS_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lib.c b/drivers/net/ethernet/intel/ixd/ixd_lib.c
new file mode 100644
index 000000000000..afc413d3650f
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_lib.c
@@ -0,0 +1,143 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+
+#define IXD_DFLT_MBX_Q_LEN 64
+
+/**
+ * ixd_init_ctlq_create_info - Initialize control queue info for creation
+ * @info: destination
+ * @type: type of the queue to create
+ * @ctlq_reg: register assigned to the control queue
+ */
+static void ixd_init_ctlq_create_info(struct libie_ctlq_create_info *info,
+ enum virtchnl2_queue_type type,
+ const struct libie_ctlq_reg *ctlq_reg)
+{
+ *info = (struct libie_ctlq_create_info) {
+ .type = type,
+ .id = -1,
+ .reg = *ctlq_reg,
+ .len = IXD_DFLT_MBX_Q_LEN,
+ };
+}
+
+/**
+ * ixd_init_libie_xn_params - Initialize xn transaction manager creation info
+ * @params: destination
+ * @adapter: adapter info struct
+ * @ctlqs: list of the managed queues to create
+ * @num_queues: length of the queue list
+ */
+static void ixd_init_libie_xn_params(struct libie_ctlq_xn_init_params *params,
+ struct ixd_adapter *adapter,
+ struct libie_ctlq_create_info *ctlqs,
+ uint num_queues)
+{
+ *params = (struct libie_ctlq_xn_init_params){
+ .cctlq_info = ctlqs,
+ .ctx = &adapter->cp_ctx,
+ .num_qs = num_queues,
+ };
+}
+
+/**
+ * ixd_adapter_fill_dflt_ctlqs - Find default control queues and store them
+ * @adapter: adapter info struct
+ */
+static void ixd_adapter_fill_dflt_ctlqs(struct ixd_adapter *adapter)
+{
+ guard(spinlock)(&adapter->cp_ctx.ctlqs_lock);
+ struct libie_ctlq_info *cq;
+
+ list_for_each_entry(cq, &adapter->cp_ctx.ctlqs, list) {
+ if (cq->qid != -1)
+ continue;
+ if (cq->type == LIBIE_CTLQ_TYPE_RX)
+ adapter->arq = cq;
+ else if (cq->type == LIBIE_CTLQ_TYPE_TX)
+ adapter->asq = cq;
+ }
+}
+
+/**
+ * ixd_deinit_dflt_mbx - Deinitialize default mailbox
+ * @adapter: adapter info struct
+ */
+void ixd_deinit_dflt_mbx(struct ixd_adapter *adapter)
+{
+ if (adapter->xnm)
+ libie_ctlq_xn_deinit(adapter->xnm, &adapter->cp_ctx);
+
+ adapter->arq = NULL;
+ adapter->asq = NULL;
+ adapter->xnm = NULL;
+}
+
+/**
+ * ixd_init_dflt_mbx - Setup default mailbox parameters and make request
+ * @adapter: adapter info struct
+ *
+ * Return: %0 on success, negative errno code on failure
+ */
+int ixd_init_dflt_mbx(struct ixd_adapter *adapter)
+{
+ struct libie_ctlq_create_info ctlqs_info[2];
+ struct libie_ctlq_xn_init_params xn_params;
+ struct libie_ctlq_reg ctlq_reg_tx;
+ struct libie_ctlq_reg ctlq_reg_rx;
+ int err;
+
+ ixd_ctlq_reg_init(adapter, &ctlq_reg_tx, &ctlq_reg_rx);
+ ixd_init_ctlq_create_info(&ctlqs_info[0], LIBIE_CTLQ_TYPE_TX,
+ &ctlq_reg_tx);
+ ixd_init_ctlq_create_info(&ctlqs_info[1], LIBIE_CTLQ_TYPE_RX,
+ &ctlq_reg_rx);
+ ixd_init_libie_xn_params(&xn_params, adapter, ctlqs_info,
+ ARRAY_SIZE(ctlqs_info));
+ err = libie_ctlq_xn_init(&xn_params);
+ if (err)
+ return err;
+ adapter->xnm = xn_params.xnm;
+
+ ixd_adapter_fill_dflt_ctlqs(adapter);
+
+ if (!adapter->asq || !adapter->arq) {
+ ixd_deinit_dflt_mbx(adapter);
+ return -ENOENT;
+ }
+
+ return 0;
+}
+
+/**
+ * ixd_init_task - Initialize after reset
+ * @work: init work struct
+ */
+void ixd_init_task(struct work_struct *work)
+{
+ struct ixd_adapter *adapter;
+ int err;
+
+ adapter = container_of(work, struct ixd_adapter,
+ init_task.init_work.work);
+
+ if (!ixd_check_reset_complete(adapter)) {
+ if (++adapter->init_task.reset_retries < 10)
+ queue_delayed_work(system_unbound_wq,
+ &adapter->init_task.init_work,
+ msecs_to_jiffies(500));
+ else
+ dev_err(ixd_to_dev(adapter),
+ "Device reset failed. The driver was unable to contact the device's firmware. Check that the FW is running.\n");
+ return;
+ }
+
+ adapter->init_task.reset_retries = 0;
+ err = ixd_init_dflt_mbx(adapter);
+ if (err)
+ dev_err(ixd_to_dev(adapter),
+ "Failed to initialize the default mailbox: %pe\n",
+ ERR_PTR(err));
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
index 75ee53152e61..35d0ff5c8d98 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_main.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -5,6 +5,7 @@
#include "ixd_lan_regs.h"
MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
+MODULE_IMPORT_NS("LIBIE_CP");
MODULE_IMPORT_NS("LIBIE_PCI");
MODULE_LICENSE("GPL");
@@ -16,7 +17,15 @@ static void ixd_remove(struct pci_dev *pdev)
{
struct ixd_adapter *adapter = pci_get_drvdata(pdev);
- libie_pci_unmap_all_mmio_regions(&adapter->hw);
+ /* Do not mix removal with (re)initialization */
+ cancel_delayed_work_sync(&adapter->init_task.init_work);
+ /* Leave the device clean on exit */
+ if (adapter->xnm)
+ libie_ctlq_xn_shutdown(adapter->xnm);
+ ixd_trigger_reset(adapter);
+ ixd_deinit_dflt_mbx(adapter);
+
+ libie_pci_unmap_all_mmio_regions(&adapter->cp_ctx.mmio_info);
}
/**
@@ -51,7 +60,7 @@ static int ixd_iomap_regions(struct ixd_adapter *adapter)
};
for (int i = 0; i < ARRAY_SIZE(regions); i++) {
- struct libie_mmio_info *mmio_info = &adapter->hw;
+ struct libie_mmio_info *mmio_info = &adapter->cp_ctx.mmio_info;
bool map_ok;
map_ok = libie_pci_map_mmio_region(mmio_info,
@@ -81,11 +90,15 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
struct ixd_adapter *adapter;
int err;
+ if (WARN_ON(ent->device != IXD_DEV_ID_CPF))
+ return -EINVAL;
+
adapter = devm_kzalloc(&pdev->dev, sizeof(*adapter), GFP_KERNEL);
if (!adapter)
return -ENOMEM;
- adapter->hw.pdev = pdev;
- INIT_LIST_HEAD(&adapter->hw.mmio_list);
+
+ adapter->cp_ctx.mmio_info.pdev = pdev;
+ INIT_LIST_HEAD(&adapter->cp_ctx.mmio_info.mmio_list);
err = libie_pci_init_dev(pdev);
if (err)
@@ -93,7 +106,18 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
pci_set_drvdata(pdev, adapter);
- return ixd_iomap_regions(adapter);
+ err = ixd_iomap_regions(adapter);
+ if (err)
+ return err;
+
+ INIT_DELAYED_WORK(&adapter->init_task.init_work,
+ ixd_init_task);
+
+ ixd_trigger_reset(adapter);
+ queue_delayed_work(system_unbound_wq, &adapter->init_task.init_work,
+ msecs_to_jiffies(500));
+
+ return 0;
}
static const struct pci_device_id ixd_pci_tbl[] = {
--
2.47.1
^ permalink raw reply related
* [PATCH net-next v5 12/15] ixd: add basic driver framework for Intel(R) Control Plane Function
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Larysa Zaremba, anthony.l.nguyen, przemyslaw.kitszel,
aleksander.lobakin, sridhar.samudrala, michal.swiatkowski,
maciej.fijalkowski, emil.s.tantilov, madhu.chittim, joshua.a.hay,
jacob.e.keller, jayaprakash.shanmugam, jiri, horms, corbet,
richardcochran, skhan, linux-doc, Bharath R
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Larysa Zaremba <larysa.zaremba@intel.com>
Add module register and probe functionality. Add the required support to
register IXD PCI driver, as well as probe and remove call backs. Enable the
PCI device and request the kernel to reserve the memory resources that will
be used by the driver. Finally map the BAR0 address space.
For now, use devm_kzalloc() to allocate adapter, as it requires the least
amount of code. In a later commit, it will be replaced with a devlink
alternative.
Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
.../device_drivers/ethernet/index.rst | 1 +
.../device_drivers/ethernet/intel/ixd.rst | 39 ++++++
drivers/net/ethernet/intel/Kconfig | 2 +
drivers/net/ethernet/intel/Makefile | 1 +
drivers/net/ethernet/intel/ixd/Kconfig | 13 ++
drivers/net/ethernet/intel/ixd/Makefile | 8 ++
drivers/net/ethernet/intel/ixd/ixd.h | 28 +++++
drivers/net/ethernet/intel/ixd/ixd_lan_regs.h | 28 +++++
drivers/net/ethernet/intel/ixd/ixd_main.c | 112 ++++++++++++++++++
9 files changed, 232 insertions(+)
create mode 100644 Documentation/networking/device_drivers/ethernet/intel/ixd.rst
create mode 100644 drivers/net/ethernet/intel/ixd/Kconfig
create mode 100644 drivers/net/ethernet/intel/ixd/Makefile
create mode 100644 drivers/net/ethernet/intel/ixd/ixd.h
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_main.c
diff --git a/Documentation/networking/device_drivers/ethernet/index.rst b/Documentation/networking/device_drivers/ethernet/index.rst
index 786a23c84b90..afaa1de55b50 100644
--- a/Documentation/networking/device_drivers/ethernet/index.rst
+++ b/Documentation/networking/device_drivers/ethernet/index.rst
@@ -37,6 +37,7 @@ Contents:
intel/igbvf
intel/ixgbe
intel/ixgbevf
+ intel/ixd
intel/i40e
intel/iavf
intel/ice
diff --git a/Documentation/networking/device_drivers/ethernet/intel/ixd.rst b/Documentation/networking/device_drivers/ethernet/intel/ixd.rst
new file mode 100644
index 000000000000..1387626e5d20
--- /dev/null
+++ b/Documentation/networking/device_drivers/ethernet/intel/ixd.rst
@@ -0,0 +1,39 @@
+.. SPDX-License-Identifier: GPL-2.0+
+
+==========================================================================
+iXD Linux* Base Driver for the Intel(R) Control Plane Function
+==========================================================================
+
+Intel iXD Linux driver.
+Copyright(C) 2025 Intel Corporation.
+
+.. contents::
+
+For questions related to hardware requirements, refer to the documentation
+supplied with your Intel adapter. All hardware requirements listed apply to use
+with Linux.
+
+
+Identifying Your Adapter
+========================
+For information on how to identify your adapter, and for the latest Intel
+network drivers, refer to the Intel Support website:
+http://www.intel.com/support
+
+
+Support
+=======
+For general information, go to the Intel support website at:
+http://www.intel.com/support/
+
+If an issue is identified with the released source code on a supported kernel
+with a supported adapter, email the specific information related to the issue
+to intel-wired-lan@lists.osuosl.org.
+
+
+Trademarks
+==========
+Intel is a trademark or registered trademark of Intel Corporation or its
+subsidiaries in the United States and/or other countries.
+
+* Other names and brands may be claimed as the property of others.
diff --git a/drivers/net/ethernet/intel/Kconfig b/drivers/net/ethernet/intel/Kconfig
index 288fa8ce53af..780f113986ea 100644
--- a/drivers/net/ethernet/intel/Kconfig
+++ b/drivers/net/ethernet/intel/Kconfig
@@ -398,4 +398,6 @@ config IGC_LEDS
source "drivers/net/ethernet/intel/idpf/Kconfig"
+source "drivers/net/ethernet/intel/ixd/Kconfig"
+
endif # NET_VENDOR_INTEL
diff --git a/drivers/net/ethernet/intel/Makefile b/drivers/net/ethernet/intel/Makefile
index 9a37dc76aef0..08b29f3b6801 100644
--- a/drivers/net/ethernet/intel/Makefile
+++ b/drivers/net/ethernet/intel/Makefile
@@ -19,3 +19,4 @@ obj-$(CONFIG_IAVF) += iavf/
obj-$(CONFIG_FM10K) += fm10k/
obj-$(CONFIG_ICE) += ice/
obj-$(CONFIG_IDPF) += idpf/
+obj-$(CONFIG_IXD) += ixd/
diff --git a/drivers/net/ethernet/intel/ixd/Kconfig b/drivers/net/ethernet/intel/ixd/Kconfig
new file mode 100644
index 000000000000..f5594efe292c
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/Kconfig
@@ -0,0 +1,13 @@
+# SPDX-License-Identifier: GPL-2.0-only
+# Copyright (C) 2025 Intel Corporation
+
+config IXD
+ tristate "Intel(R) Control Plane Function Support"
+ depends on PCI_MSI
+ select LIBETH
+ select LIBIE_PCI
+ help
+ This driver supports Intel(R) Control Plane PCI Function
+ of Intel E2100 and later IPUs and FNICs.
+ It facilitates a centralized control over multiple IDPF PFs/VFs/SFs
+ exposed by the same card.
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
new file mode 100644
index 000000000000..3849bc240600
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -0,0 +1,8 @@
+# SPDX-License-Identifier: GPL-2.0-only
+# Copyright (C) 2025 Intel Corporation
+
+# Intel(R) Control Plane Function Linux Driver
+
+obj-$(CONFIG_IXD) += ixd.o
+
+ixd-y := ixd_main.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd.h b/drivers/net/ethernet/intel/ixd/ixd.h
new file mode 100644
index 000000000000..1b918c5d31cd
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd.h
@@ -0,0 +1,28 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_H_
+#define _IXD_H_
+
+#include <linux/net/intel/libie/pci.h>
+
+/**
+ * struct ixd_adapter - Data structure representing a CPF
+ * @hw: Device access data
+ */
+struct ixd_adapter {
+ struct libie_mmio_info hw;
+};
+
+/**
+ * ixd_to_dev - Get the corresponding device struct from an adapter
+ * @adapter: PCI device driver-specific private data
+ *
+ * Return: struct device corresponding to the given adapter
+ */
+static inline struct device *ixd_to_dev(struct ixd_adapter *adapter)
+{
+ return &adapter->hw.pdev->dev;
+}
+
+#endif /* _IXD_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
new file mode 100644
index 000000000000..fbb88929d0de
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
@@ -0,0 +1,28 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_LAN_REGS_H_
+#define _IXD_LAN_REGS_H_
+
+/* Control Plane Function PCI ID */
+#define IXD_DEV_ID_CPF 0x1efe
+
+/* Control Queue (Mailbox) */
+#define PF_FW_MBX_REG_LEN 4096
+#define PF_FW_MBX 0x08400000
+
+/* Reset registers */
+#define PFGEN_RTRIG_REG_LEN 2048
+#define PFGEN_RTRIG 0x08407000 /* Device resets */
+
+/**
+ * struct ixd_bar_region - BAR region description
+ * @offset: BAR region offset
+ * @size: BAR region size
+ */
+struct ixd_bar_region {
+ resource_size_t offset;
+ resource_size_t size;
+};
+
+#endif /* _IXD_LAN_REGS_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
new file mode 100644
index 000000000000..75ee53152e61
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -0,0 +1,112 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_lan_regs.h"
+
+MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
+MODULE_IMPORT_NS("LIBIE_PCI");
+MODULE_LICENSE("GPL");
+
+/**
+ * ixd_remove - remove a CPF PCI device
+ * @pdev: PCI device being removed
+ */
+static void ixd_remove(struct pci_dev *pdev)
+{
+ struct ixd_adapter *adapter = pci_get_drvdata(pdev);
+
+ libie_pci_unmap_all_mmio_regions(&adapter->hw);
+}
+
+/**
+ * ixd_shutdown - shut down a CPF PCI device
+ * @pdev: PCI device being shut down
+ */
+static void ixd_shutdown(struct pci_dev *pdev)
+{
+ ixd_remove(pdev);
+
+ if (system_state == SYSTEM_POWER_OFF)
+ pci_set_power_state(pdev, PCI_D3hot);
+}
+
+/**
+ * ixd_iomap_regions - iomap PCI BARs
+ * @adapter: adapter to map memory regions for
+ *
+ * Returns: %0 on success, negative on failure
+ */
+static int ixd_iomap_regions(struct ixd_adapter *adapter)
+{
+ const struct ixd_bar_region regions[] = {
+ {
+ .offset = PFGEN_RTRIG,
+ .size = PFGEN_RTRIG_REG_LEN,
+ },
+ {
+ .offset = PF_FW_MBX,
+ .size = PF_FW_MBX_REG_LEN,
+ },
+ };
+
+ for (int i = 0; i < ARRAY_SIZE(regions); i++) {
+ struct libie_mmio_info *mmio_info = &adapter->hw;
+ bool map_ok;
+
+ map_ok = libie_pci_map_mmio_region(mmio_info,
+ regions[i].offset,
+ regions[i].size);
+ if (!map_ok) {
+ dev_err(ixd_to_dev(adapter),
+ "Failed to map PCI device MMIO region\n");
+
+ libie_pci_unmap_all_mmio_regions(mmio_info);
+ return -EIO;
+ }
+ }
+
+ return 0;
+}
+
+/**
+ * ixd_probe - probe a CPF PCI device
+ * @pdev: corresponding PCI device
+ * @ent: entry in ixd_pci_tbl
+ *
+ * Returns: %0 on success, negative errno code on failure
+ */
+static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+ struct ixd_adapter *adapter;
+ int err;
+
+ adapter = devm_kzalloc(&pdev->dev, sizeof(*adapter), GFP_KERNEL);
+ if (!adapter)
+ return -ENOMEM;
+ adapter->hw.pdev = pdev;
+ INIT_LIST_HEAD(&adapter->hw.mmio_list);
+
+ err = libie_pci_init_dev(pdev);
+ if (err)
+ return err;
+
+ pci_set_drvdata(pdev, adapter);
+
+ return ixd_iomap_regions(adapter);
+}
+
+static const struct pci_device_id ixd_pci_tbl[] = {
+ { PCI_VDEVICE(INTEL, IXD_DEV_ID_CPF) },
+ { }
+};
+MODULE_DEVICE_TABLE(pci, ixd_pci_tbl);
+
+static struct pci_driver ixd_driver = {
+ .name = KBUILD_MODNAME,
+ .id_table = ixd_pci_tbl,
+ .probe = ixd_probe,
+ .remove = ixd_remove,
+ .shutdown = ixd_shutdown,
+};
+module_pci_driver(ixd_driver);
--
2.47.1
^ permalink raw reply related
* [PATCH net-next v5 11/15] idpf: print a debug message and bail in case of non-event ctlq message
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Larysa Zaremba, anthony.l.nguyen, przemyslaw.kitszel,
aleksander.lobakin, sridhar.samudrala, michal.swiatkowski,
maciej.fijalkowski, emil.s.tantilov, madhu.chittim, joshua.a.hay,
jacob.e.keller, jayaprakash.shanmugam, jiri, horms, corbet,
richardcochran, skhan, linux-doc, Aleksandr Loktionov,
Samuel Salin
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Larysa Zaremba <larysa.zaremba@intel.com>
Unlike previous internal idpf ctlq implementation, idpf calls the default
message handler for all received messages that do not have a matching xn
transaction, not only for VIRTCHNL2_OP_EVENT. This leads to many error
messages printing garbage, because the parsing expected a valid event
message, but got e.g. a delayed response for a timed-out transaction.
The information about timed-out transactions and otherwise unhandleable
messages can still be valuable for developers, so print the information
with dynamic debug and exit the function, so the following functions can
parse valid events in peace.
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Tested-by: Samuel Salin <Samuel.salin@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/idpf/idpf_virtchnl.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
index 7c166651f90d..13f38d03411a 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
@@ -84,6 +84,13 @@ void idpf_recv_event_msg(struct libie_ctlq_ctx *ctx,
u32 event;
adapter = container_of(ctx, struct idpf_adapter, ctlq_ctx);
+ if (ctlq_msg->chnl_opcode != VIRTCHNL2_OP_EVENT) {
+ dev_dbg(&adapter->pdev->dev,
+ "Unhandled message with opcode %u from CP\n",
+ ctlq_msg->chnl_opcode);
+ goto free_rx_buf;
+ }
+
if (payload_size < sizeof(*v2e)) {
dev_err_ratelimited(&adapter->pdev->dev, "Failed to receive valid payload for event msg (op %d len %d)\n",
ctlq_msg->chnl_opcode,
--
2.47.1
^ permalink raw reply related
* [PATCH net-next v5 09/15] idpf: refactor idpf to use libie control queues
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Pavan Kumar Linga, anthony.l.nguyen, larysa.zaremba,
przemyslaw.kitszel, aleksander.lobakin, sridhar.samudrala,
michal.swiatkowski, maciej.fijalkowski, emil.s.tantilov,
joshua.a.hay, jacob.e.keller, jayaprakash.shanmugam, jiri, horms,
corbet, richardcochran, skhan, linux-doc, Aleksandr Loktionov,
Samuel Salin
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Support to initialize and configure controlqs, and manage their
transactions was introduced in libie. As part of it, most of the existing
controlq structures are renamed and modified. Use those APIs in idpf and
make all the necessary changes.
Previously for the send and receive virtchnl messages, there used to be a
memcpy involved in controlq code to copy the buffer info passed by the send
function into the controlq specific buffers. There was no restriction to
use automatic memory in that case. The new implementation in libie removed
copying of the send buffer info and introduced DMA mapping of the send
buffer itself. To accommodate it, use dynamic memory for the larger send
buffers. For smaller ones (<= 128 bytes) libie still can copy them into the
pre-allocated message memory. Those changes result in a pretty big diff,
but the changes are fairly trivial and localized.
In case of receive, idpf receives a page pool buffer allocated by the libie
and care should be taken to release it after use in the idpf.
This refactoring introduces roughly additional 40KB of module storage used
for systems that only run idpf, so idpf + libie_cp + libie_pci takes about
7% more storage than just idpf before refactoring.
We now pre-allocate small TX buffers, so that does increase the memory
usage, but reduces the need to allocate. This results in additional 256 *
128B of memory permanently used, increasing the worst-case memory usage by
32KB but our ctlq RX buffers need to be of size 4096B anyway (not changed
by the patchset), so this is hardly noticeable.
As for the timings, the fact that we are mostly limited by the HW response
time which is far from instant, is not changed by this refactor.
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Tested-by: Samuel Salin <Samuel.salin@intel.com>
Co-developed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/idpf/Makefile | 2 -
drivers/net/ethernet/intel/idpf/idpf.h | 28 +-
.../net/ethernet/intel/idpf/idpf_controlq.c | 631 -------
.../net/ethernet/intel/idpf/idpf_controlq.h | 142 --
.../ethernet/intel/idpf/idpf_controlq_api.h | 177 --
.../ethernet/intel/idpf/idpf_controlq_setup.c | 169 --
drivers/net/ethernet/intel/idpf/idpf_dev.c | 60 +-
.../net/ethernet/intel/idpf/idpf_ethtool.c | 28 +-
drivers/net/ethernet/intel/idpf/idpf_lib.c | 51 +-
drivers/net/ethernet/intel/idpf/idpf_main.c | 3 -
drivers/net/ethernet/intel/idpf/idpf_mem.h | 20 -
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 2 +-
drivers/net/ethernet/intel/idpf/idpf_vf_dev.c | 61 +-
.../net/ethernet/intel/idpf/idpf_virtchnl.c | 1575 +++++++----------
.../net/ethernet/intel/idpf/idpf_virtchnl.h | 95 +-
.../ethernet/intel/idpf/idpf_virtchnl_ptp.c | 254 ++-
16 files changed, 821 insertions(+), 2477 deletions(-)
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq.c
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq.h
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq_api.h
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_mem.h
diff --git a/drivers/net/ethernet/intel/idpf/Makefile b/drivers/net/ethernet/intel/idpf/Makefile
index 651ddee942bd..4aaafa175ec3 100644
--- a/drivers/net/ethernet/intel/idpf/Makefile
+++ b/drivers/net/ethernet/intel/idpf/Makefile
@@ -6,8 +6,6 @@
obj-$(CONFIG_IDPF) += idpf.o
idpf-y := \
- idpf_controlq.o \
- idpf_controlq_setup.o \
idpf_dev.o \
idpf_ethtool.o \
idpf_idc.o \
diff --git a/drivers/net/ethernet/intel/idpf/idpf.h b/drivers/net/ethernet/intel/idpf/idpf.h
index 92a120aadfcd..d7d751e2a781 100644
--- a/drivers/net/ethernet/intel/idpf/idpf.h
+++ b/drivers/net/ethernet/intel/idpf/idpf.h
@@ -27,7 +27,6 @@ struct idpf_rss_data;
#include <linux/net/intel/virtchnl2.h>
#include "idpf_txrx.h"
-#include "idpf_controlq.h"
#define GETMAXVAL(num_bits) GENMASK((num_bits) - 1, 0)
@@ -37,11 +36,10 @@ struct idpf_rss_data;
#define IDPF_NUM_FILTERS_PER_MSG 20
#define IDPF_NUM_DFLT_MBX_Q 2 /* includes both TX and RX */
#define IDPF_DFLT_MBX_Q_LEN 64
-#define IDPF_DFLT_MBX_ID -1
/* maximum number of times to try before resetting mailbox */
#define IDPF_MB_MAX_ERR 20
#define IDPF_NUM_CHUNKS_PER_MSG(struct_sz, chunk_sz) \
- ((IDPF_CTLQ_MAX_BUF_LEN - (struct_sz)) / (chunk_sz))
+ ((LIBIE_CTLQ_MAX_BUF_LEN - (struct_sz)) / (chunk_sz))
#define IDPF_WAIT_FOR_MARKER_TIMEO 500
#define IDPF_MAX_WAIT 500
@@ -202,8 +200,8 @@ struct idpf_vport_max_q {
* @ptp_reg_init: PTP register initialization
*/
struct idpf_reg_ops {
- void (*ctlq_reg_init)(struct idpf_adapter *adapter,
- struct idpf_ctlq_create_info *cq);
+ void (*ctlq_reg_init)(struct libie_mmio_info *mmio,
+ struct libie_ctlq_create_info *cctlq_info);
int (*intr_reg_init)(struct idpf_vport *vport,
struct idpf_q_vec_rsrc *rsrc);
void (*mb_intr_reg_init)(struct idpf_adapter *adapter);
@@ -606,8 +604,6 @@ struct idpf_vport_config {
DECLARE_BITMAP(flags, IDPF_VPORT_CONFIG_FLAGS_NBITS);
};
-struct idpf_vc_xn_manager;
-
#define idpf_for_each_vport(adapter, iter) \
for (struct idpf_vport **__##iter = &(adapter)->vports[0], \
*iter = (adapter)->max_vports ? *__##iter : NULL; \
@@ -625,8 +621,10 @@ struct idpf_vc_xn_manager;
* @state: Init state machine
* @flags: See enum idpf_flags
* @reset_reg: See struct idpf_reset_reg
- * @hw: Device access data
* @ctlq_ctx: controlq context
+ * @asq: Send control queue info
+ * @arq: Receive control queue info
+ * @xnm: Xn transaction manager
* @num_avail_msix: Available number of MSIX vectors
* @num_msix_entries: Number of entries in MSIX table
* @msix_entries: MSIX table
@@ -659,7 +657,6 @@ struct idpf_vc_xn_manager;
* @stats_task: Periodic statistics retrieval task
* @stats_wq: Workqueue for statistics task
* @caps: Negotiated capabilities with device
- * @vcxn_mngr: Virtchnl transaction manager
* @dev_ops: See idpf_dev_ops
* @cdev_info: IDC core device info pointer
* @num_vfs: Number of allocated VFs through sysfs. PF does not directly talk
@@ -683,8 +680,10 @@ struct idpf_adapter {
enum idpf_state state;
DECLARE_BITMAP(flags, IDPF_FLAGS_NBITS);
struct idpf_reset_reg reset_reg;
- struct idpf_hw hw;
struct libie_ctlq_ctx ctlq_ctx;
+ struct libie_ctlq_info *asq;
+ struct libie_ctlq_info *arq;
+ struct libie_ctlq_xn_manager *xnm;
u16 num_avail_msix;
u16 num_msix_entries;
struct msix_entry *msix_entries;
@@ -721,7 +720,6 @@ struct idpf_adapter {
struct delayed_work stats_task;
struct workqueue_struct *stats_wq;
struct virtchnl2_get_capabilities caps;
- struct idpf_vc_xn_manager *vcxn_mngr;
struct idpf_dev_ops dev_ops;
struct iidc_rdma_core_dev_info *cdev_info;
@@ -881,12 +879,12 @@ static inline u8 idpf_get_min_tx_pkt_len(struct idpf_adapter *adapter)
*/
static inline bool idpf_is_reset_detected(struct idpf_adapter *adapter)
{
- if (!adapter->hw.arq)
+ struct libie_ctlq_info *arq = adapter->arq;
+
+ if (!arq)
return true;
- return !(readl(libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info,
- adapter->hw.arq->reg.len)) &
- adapter->hw.arq->reg.len_mask);
+ return !(readl(arq->reg.len) & arq->reg.len_mask);
}
/**
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq.c b/drivers/net/ethernet/intel/idpf/idpf_controlq.c
deleted file mode 100644
index 020b08367e18..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq.c
+++ /dev/null
@@ -1,631 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-only
-/* Copyright (C) 2023 Intel Corporation */
-
-#include "idpf.h"
-
-/**
- * idpf_ctlq_setup_regs - initialize control queue registers
- * @cq: pointer to the specific control queue
- * @q_create_info: structs containing info for each queue to be initialized
- */
-static void idpf_ctlq_setup_regs(struct idpf_ctlq_info *cq,
- struct idpf_ctlq_create_info *q_create_info)
-{
- /* set control queue registers in our local struct */
- cq->reg.head = q_create_info->reg.head;
- cq->reg.tail = q_create_info->reg.tail;
- cq->reg.len = q_create_info->reg.len;
- cq->reg.bah = q_create_info->reg.bah;
- cq->reg.bal = q_create_info->reg.bal;
- cq->reg.len_mask = q_create_info->reg.len_mask;
- cq->reg.len_ena_mask = q_create_info->reg.len_ena_mask;
- cq->reg.head_mask = q_create_info->reg.head_mask;
-}
-
-/**
- * idpf_ctlq_init_regs - Initialize control queue registers
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- * @is_rxq: true if receive control queue, false otherwise
- *
- * Initialize registers. The caller is expected to have already initialized the
- * descriptor ring memory and buffer memory
- */
-static void idpf_ctlq_init_regs(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
- bool is_rxq)
-{
- struct libie_mmio_info *mmio = &hw->back->ctlq_ctx.mmio_info;
-
- /* Update tail to post pre-allocated buffers for rx queues */
- if (is_rxq)
- writel((u32)(cq->ring_size - 1),
- libie_pci_get_mmio_addr(mmio, cq->reg.tail));
-
- /* For non-Mailbox control queues only TAIL need to be set */
- if (cq->q_id != -1)
- return;
-
- /* Clear Head for both send or receive */
- writel(0, libie_pci_get_mmio_addr(mmio, cq->reg.head));
-
- /* set starting point */
- writel(lower_32_bits(cq->desc_ring.pa),
- libie_pci_get_mmio_addr(mmio, cq->reg.bal));
- writel(upper_32_bits(cq->desc_ring.pa),
- libie_pci_get_mmio_addr(mmio, cq->reg.bah));
- writel((cq->ring_size | cq->reg.len_ena_mask),
- libie_pci_get_mmio_addr(mmio, cq->reg.len));
-}
-
-/**
- * idpf_ctlq_init_rxq_bufs - populate receive queue descriptors with buf
- * @cq: pointer to the specific Control queue
- *
- * Record the address of the receive queue DMA buffers in the descriptors.
- * The buffers must have been previously allocated.
- */
-static void idpf_ctlq_init_rxq_bufs(struct idpf_ctlq_info *cq)
-{
- int i;
-
- for (i = 0; i < cq->ring_size; i++) {
- struct idpf_ctlq_desc *desc = IDPF_CTLQ_DESC(cq, i);
- struct idpf_dma_mem *bi = cq->bi.rx_buff[i];
-
- /* No buffer to post to descriptor, continue */
- if (!bi)
- continue;
-
- desc->flags =
- cpu_to_le16(IDPF_CTLQ_FLAG_BUF | IDPF_CTLQ_FLAG_RD);
- desc->opcode = 0;
- desc->datalen = cpu_to_le16(bi->size);
- desc->ret_val = 0;
- desc->v_opcode_dtype = 0;
- desc->v_retval = 0;
- desc->params.indirect.addr_high =
- cpu_to_le32(upper_32_bits(bi->pa));
- desc->params.indirect.addr_low =
- cpu_to_le32(lower_32_bits(bi->pa));
- desc->params.indirect.param0 = 0;
- desc->params.indirect.sw_cookie = 0;
- desc->params.indirect.v_flags = 0;
- }
-}
-
-/**
- * idpf_ctlq_shutdown - shutdown the CQ
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * The main shutdown routine for any controq queue
- */
-static void idpf_ctlq_shutdown(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- spin_lock(&cq->cq_lock);
-
- /* free ring buffers and the ring itself */
- idpf_ctlq_dealloc_ring_res(hw, cq);
-
- /* Set ring_size to 0 to indicate uninitialized queue */
- cq->ring_size = 0;
-
- spin_unlock(&cq->cq_lock);
-}
-
-/**
- * idpf_ctlq_add - add one control queue
- * @hw: pointer to hardware struct
- * @qinfo: info for queue to be created
- * @cq_out: (output) double pointer to control queue to be created
- *
- * Allocate and initialize a control queue and add it to the control queue list.
- * The cq parameter will be allocated/initialized and passed back to the caller
- * if no errors occur.
- *
- * Note: idpf_ctlq_init must be called prior to any calls to idpf_ctlq_add
- */
-int idpf_ctlq_add(struct idpf_hw *hw,
- struct idpf_ctlq_create_info *qinfo,
- struct idpf_ctlq_info **cq_out)
-{
- struct idpf_ctlq_info *cq;
- bool is_rxq = false;
- int err;
-
- cq = kzalloc_obj(*cq);
- if (!cq)
- return -ENOMEM;
-
- cq->cq_type = qinfo->type;
- cq->q_id = qinfo->id;
- cq->buf_size = qinfo->buf_size;
- cq->ring_size = qinfo->len;
-
- cq->next_to_use = 0;
- cq->next_to_clean = 0;
- cq->next_to_post = cq->ring_size - 1;
-
- switch (qinfo->type) {
- case IDPF_CTLQ_TYPE_MAILBOX_RX:
- is_rxq = true;
- fallthrough;
- case IDPF_CTLQ_TYPE_MAILBOX_TX:
- err = idpf_ctlq_alloc_ring_res(hw, cq);
- break;
- default:
- err = -EBADR;
- break;
- }
-
- if (err)
- goto init_free_q;
-
- if (is_rxq) {
- idpf_ctlq_init_rxq_bufs(cq);
- } else {
- /* Allocate the array of msg pointers for TX queues */
- cq->bi.tx_msg = kzalloc_objs(struct idpf_ctlq_msg *, qinfo->len);
- if (!cq->bi.tx_msg) {
- err = -ENOMEM;
- goto init_dealloc_q_mem;
- }
- }
-
- idpf_ctlq_setup_regs(cq, qinfo);
-
- idpf_ctlq_init_regs(hw, cq, is_rxq);
-
- spin_lock_init(&cq->cq_lock);
-
- list_add(&cq->cq_list, &hw->cq_list_head);
-
- *cq_out = cq;
-
- return 0;
-
-init_dealloc_q_mem:
- /* free ring buffers and the ring itself */
- idpf_ctlq_dealloc_ring_res(hw, cq);
-init_free_q:
- kfree(cq);
-
- return err;
-}
-
-/**
- * idpf_ctlq_remove - deallocate and remove specified control queue
- * @hw: pointer to hardware struct
- * @cq: pointer to control queue to be removed
- */
-void idpf_ctlq_remove(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- list_del(&cq->cq_list);
- idpf_ctlq_shutdown(hw, cq);
- kfree(cq);
-}
-
-/**
- * idpf_ctlq_init - main initialization routine for all control queues
- * @hw: pointer to hardware struct
- * @num_q: number of queues to initialize
- * @q_info: array of structs containing info for each queue to be initialized
- *
- * This initializes any number and any type of control queues. This is an all
- * or nothing routine; if one fails, all previously allocated queues will be
- * destroyed. This must be called prior to using the individual add/remove
- * APIs.
- */
-int idpf_ctlq_init(struct idpf_hw *hw, u8 num_q,
- struct idpf_ctlq_create_info *q_info)
-{
- struct idpf_ctlq_info *cq, *tmp;
- int err;
- int i;
-
- INIT_LIST_HEAD(&hw->cq_list_head);
-
- for (i = 0; i < num_q; i++) {
- struct idpf_ctlq_create_info *qinfo = q_info + i;
-
- err = idpf_ctlq_add(hw, qinfo, &cq);
- if (err)
- goto init_destroy_qs;
- }
-
- return 0;
-
-init_destroy_qs:
- list_for_each_entry_safe(cq, tmp, &hw->cq_list_head, cq_list)
- idpf_ctlq_remove(hw, cq);
-
- return err;
-}
-
-/**
- * idpf_ctlq_deinit - destroy all control queues
- * @hw: pointer to hw struct
- */
-void idpf_ctlq_deinit(struct idpf_hw *hw)
-{
- struct idpf_ctlq_info *cq, *tmp;
-
- list_for_each_entry_safe(cq, tmp, &hw->cq_list_head, cq_list)
- idpf_ctlq_remove(hw, cq);
-}
-
-/**
- * idpf_ctlq_send - send command to Control Queue (CTQ)
- * @hw: pointer to hw struct
- * @cq: handle to control queue struct to send on
- * @num_q_msg: number of messages to send on control queue
- * @q_msg: pointer to array of queue messages to be sent
- *
- * The caller is expected to allocate DMAable buffers and pass them to the
- * send routine via the q_msg struct / control queue specific data struct.
- * The control queue will hold a reference to each send message until
- * the completion for that message has been cleaned.
- */
-int idpf_ctlq_send(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
- u16 num_q_msg, struct idpf_ctlq_msg q_msg[])
-{
- struct idpf_ctlq_desc *desc;
- int num_desc_avail;
- int err = 0;
- int i;
-
- spin_lock(&cq->cq_lock);
-
- /* Ensure there are enough descriptors to send all messages */
- num_desc_avail = IDPF_CTLQ_DESC_UNUSED(cq);
- if (num_desc_avail == 0 || num_desc_avail < num_q_msg) {
- err = -ENOSPC;
- goto err_unlock;
- }
-
- for (i = 0; i < num_q_msg; i++) {
- struct idpf_ctlq_msg *msg = &q_msg[i];
-
- desc = IDPF_CTLQ_DESC(cq, cq->next_to_use);
-
- desc->opcode = cpu_to_le16(msg->opcode);
- desc->pfid_vfid = cpu_to_le16(msg->func_id);
-
- desc->v_opcode_dtype = cpu_to_le32(msg->cookie.mbx.chnl_opcode);
- desc->v_retval = cpu_to_le32(msg->cookie.mbx.chnl_retval);
-
- desc->flags = cpu_to_le16((msg->host_id & IDPF_HOST_ID_MASK) <<
- IDPF_CTLQ_FLAG_HOST_ID_S);
- if (msg->data_len) {
- struct idpf_dma_mem *buff = msg->ctx.indirect.payload;
-
- desc->datalen |= cpu_to_le16(msg->data_len);
- desc->flags |= cpu_to_le16(IDPF_CTLQ_FLAG_BUF);
- desc->flags |= cpu_to_le16(IDPF_CTLQ_FLAG_RD);
-
- /* Update the address values in the desc with the pa
- * value for respective buffer
- */
- desc->params.indirect.addr_high =
- cpu_to_le32(upper_32_bits(buff->pa));
- desc->params.indirect.addr_low =
- cpu_to_le32(lower_32_bits(buff->pa));
-
- memcpy(&desc->params, msg->ctx.indirect.context,
- IDPF_INDIRECT_CTX_SIZE);
- } else {
- memcpy(&desc->params, msg->ctx.direct,
- IDPF_DIRECT_CTX_SIZE);
- }
-
- /* Store buffer info */
- cq->bi.tx_msg[cq->next_to_use] = msg;
-
- (cq->next_to_use)++;
- if (cq->next_to_use == cq->ring_size)
- cq->next_to_use = 0;
- }
-
- /* Force memory write to complete before letting hardware
- * know that there are new descriptors to fetch.
- */
- dma_wmb();
-
- writel(cq->next_to_use,
- libie_pci_get_mmio_addr(&hw->back->ctlq_ctx.mmio_info,
- cq->reg.tail));
-
-err_unlock:
- spin_unlock(&cq->cq_lock);
-
- return err;
-}
-
-/**
- * idpf_ctlq_clean_sq - reclaim send descriptors on HW write back for the
- * requested queue
- * @cq: pointer to the specific Control queue
- * @clean_count: (input|output) number of descriptors to clean as input, and
- * number of descriptors actually cleaned as output
- * @msg_status: (output) pointer to msg pointer array to be populated; needs
- * to be allocated by caller
- *
- * Returns an array of message pointers associated with the cleaned
- * descriptors. The pointers are to the original ctlq_msgs sent on the cleaned
- * descriptors. The status will be returned for each; any messages that failed
- * to send will have a non-zero status. The caller is expected to free original
- * ctlq_msgs and free or reuse the DMA buffers.
- */
-int idpf_ctlq_clean_sq(struct idpf_ctlq_info *cq, u16 *clean_count,
- struct idpf_ctlq_msg *msg_status[])
-{
- struct idpf_ctlq_desc *desc;
- u16 i, num_to_clean;
- u16 ntc, desc_err;
-
- if (*clean_count == 0)
- return 0;
- if (*clean_count > cq->ring_size)
- return -EBADR;
-
- spin_lock(&cq->cq_lock);
-
- ntc = cq->next_to_clean;
-
- num_to_clean = *clean_count;
-
- for (i = 0; i < num_to_clean; i++) {
- /* Fetch next descriptor and check if marked as done */
- desc = IDPF_CTLQ_DESC(cq, ntc);
- if (!(le16_to_cpu(desc->flags) & IDPF_CTLQ_FLAG_DD))
- break;
-
- /* Ensure no other fields are read until DD flag is checked */
- dma_rmb();
-
- /* strip off FW internal code */
- desc_err = le16_to_cpu(desc->ret_val) & 0xff;
-
- msg_status[i] = cq->bi.tx_msg[ntc];
- msg_status[i]->status = desc_err;
-
- cq->bi.tx_msg[ntc] = NULL;
-
- /* Zero out any stale data */
- memset(desc, 0, sizeof(*desc));
-
- ntc++;
- if (ntc == cq->ring_size)
- ntc = 0;
- }
-
- cq->next_to_clean = ntc;
-
- spin_unlock(&cq->cq_lock);
-
- /* Return number of descriptors actually cleaned */
- *clean_count = i;
-
- return 0;
-}
-
-/**
- * idpf_ctlq_post_rx_buffs - post buffers to descriptor ring
- * @hw: pointer to hw struct
- * @cq: pointer to control queue handle
- * @buff_count: (input|output) input is number of buffers caller is trying to
- * return; output is number of buffers that were not posted
- * @buffs: array of pointers to dma mem structs to be given to hardware
- *
- * Caller uses this function to return DMA buffers to the descriptor ring after
- * consuming them; buff_count will be the number of buffers.
- *
- * Note: this function needs to be called after a receive call even
- * if there are no DMA buffers to be returned, i.e. buff_count = 0,
- * buffs = NULL to support direct commands
- */
-int idpf_ctlq_post_rx_buffs(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
- u16 *buff_count, struct idpf_dma_mem **buffs)
-{
- struct idpf_ctlq_desc *desc;
- u16 ntp = cq->next_to_post;
- bool buffs_avail = false;
- u16 tbp = ntp + 1;
- int i = 0;
-
- if (*buff_count > cq->ring_size)
- return -EBADR;
-
- if (*buff_count > 0)
- buffs_avail = true;
-
- spin_lock(&cq->cq_lock);
-
- if (tbp >= cq->ring_size)
- tbp = 0;
-
- if (tbp == cq->next_to_clean)
- /* Nothing to do */
- goto post_buffs_out;
-
- /* Post buffers for as many as provided or up until the last one used */
- while (ntp != cq->next_to_clean) {
- desc = IDPF_CTLQ_DESC(cq, ntp);
-
- if (cq->bi.rx_buff[ntp])
- goto fill_desc;
- if (!buffs_avail) {
- /* If the caller hasn't given us any buffers or
- * there are none left, search the ring itself
- * for an available buffer to move to this
- * entry starting at the next entry in the ring
- */
- tbp = ntp + 1;
-
- /* Wrap ring if necessary */
- if (tbp >= cq->ring_size)
- tbp = 0;
-
- while (tbp != cq->next_to_clean) {
- if (cq->bi.rx_buff[tbp]) {
- cq->bi.rx_buff[ntp] =
- cq->bi.rx_buff[tbp];
- cq->bi.rx_buff[tbp] = NULL;
-
- /* Found a buffer, no need to
- * search anymore
- */
- break;
- }
-
- /* Wrap ring if necessary */
- tbp++;
- if (tbp >= cq->ring_size)
- tbp = 0;
- }
-
- if (tbp == cq->next_to_clean)
- goto post_buffs_out;
- } else {
- /* Give back pointer to DMA buffer */
- cq->bi.rx_buff[ntp] = buffs[i];
- i++;
-
- if (i >= *buff_count)
- buffs_avail = false;
- }
-
-fill_desc:
- desc->flags =
- cpu_to_le16(IDPF_CTLQ_FLAG_BUF | IDPF_CTLQ_FLAG_RD);
-
- /* Post buffers to descriptor */
- desc->datalen = cpu_to_le16(cq->bi.rx_buff[ntp]->size);
- desc->params.indirect.addr_high =
- cpu_to_le32(upper_32_bits(cq->bi.rx_buff[ntp]->pa));
- desc->params.indirect.addr_low =
- cpu_to_le32(lower_32_bits(cq->bi.rx_buff[ntp]->pa));
-
- ntp++;
- if (ntp == cq->ring_size)
- ntp = 0;
- }
-
-post_buffs_out:
- /* Only update tail if buffers were actually posted */
- if (cq->next_to_post != ntp) {
- if (ntp)
- /* Update next_to_post to ntp - 1 since current ntp
- * will not have a buffer
- */
- cq->next_to_post = ntp - 1;
- else
- /* Wrap to end of end ring since current ntp is 0 */
- cq->next_to_post = cq->ring_size - 1;
-
- dma_wmb();
-
- writel(cq->next_to_post,
- libie_pci_get_mmio_addr(&hw->back->ctlq_ctx.mmio_info,
- cq->reg.tail));
- }
-
- spin_unlock(&cq->cq_lock);
-
- /* return the number of buffers that were not posted */
- *buff_count = *buff_count - i;
-
- return 0;
-}
-
-/**
- * idpf_ctlq_recv - receive control queue message call back
- * @cq: pointer to control queue handle to receive on
- * @num_q_msg: (input|output) input number of messages that should be received;
- * output number of messages actually received
- * @q_msg: (output) array of received control queue messages on this q;
- * needs to be pre-allocated by caller for as many messages as requested
- *
- * Called by interrupt handler or polling mechanism. Caller is expected
- * to free buffers
- */
-int idpf_ctlq_recv(struct idpf_ctlq_info *cq, u16 *num_q_msg,
- struct idpf_ctlq_msg *q_msg)
-{
- u16 num_to_clean, ntc, flags;
- struct idpf_ctlq_desc *desc;
- int err = 0;
- u16 i;
-
- /* take the lock before we start messing with the ring */
- spin_lock(&cq->cq_lock);
-
- ntc = cq->next_to_clean;
-
- num_to_clean = *num_q_msg;
-
- for (i = 0; i < num_to_clean; i++) {
- /* Fetch next descriptor and check if marked as done */
- desc = IDPF_CTLQ_DESC(cq, ntc);
- flags = le16_to_cpu(desc->flags);
-
- if (!(flags & IDPF_CTLQ_FLAG_DD))
- break;
-
- /* Ensure no other fields are read until DD flag is checked */
- dma_rmb();
-
- q_msg[i].vmvf_type = (flags &
- (IDPF_CTLQ_FLAG_FTYPE_VM |
- IDPF_CTLQ_FLAG_FTYPE_PF)) >>
- IDPF_CTLQ_FLAG_FTYPE_S;
-
- if (flags & IDPF_CTLQ_FLAG_ERR)
- err = -EBADMSG;
-
- q_msg[i].cookie.mbx.chnl_opcode =
- le32_to_cpu(desc->v_opcode_dtype);
- q_msg[i].cookie.mbx.chnl_retval =
- le32_to_cpu(desc->v_retval);
-
- q_msg[i].opcode = le16_to_cpu(desc->opcode);
- q_msg[i].data_len = le16_to_cpu(desc->datalen);
- q_msg[i].status = le16_to_cpu(desc->ret_val);
-
- if (desc->datalen) {
- memcpy(q_msg[i].ctx.indirect.context,
- &desc->params.indirect, IDPF_INDIRECT_CTX_SIZE);
-
- /* Assign pointer to dma buffer to ctlq_msg array
- * to be given to upper layer
- */
- q_msg[i].ctx.indirect.payload = cq->bi.rx_buff[ntc];
-
- /* Zero out pointer to DMA buffer info;
- * will be repopulated by post buffers API
- */
- cq->bi.rx_buff[ntc] = NULL;
- } else {
- memcpy(q_msg[i].ctx.direct, desc->params.raw,
- IDPF_DIRECT_CTX_SIZE);
- }
-
- /* Zero out stale data in descriptor */
- memset(desc, 0, sizeof(struct idpf_ctlq_desc));
-
- ntc++;
- if (ntc == cq->ring_size)
- ntc = 0;
- }
-
- cq->next_to_clean = ntc;
-
- spin_unlock(&cq->cq_lock);
-
- *num_q_msg = i;
- if (*num_q_msg == 0)
- err = -ENOMSG;
-
- return err;
-}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq.h b/drivers/net/ethernet/intel/idpf/idpf_controlq.h
deleted file mode 100644
index acf595e9265f..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq.h
+++ /dev/null
@@ -1,142 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/* Copyright (C) 2023 Intel Corporation */
-
-#ifndef _IDPF_CONTROLQ_H_
-#define _IDPF_CONTROLQ_H_
-
-#include <linux/slab.h>
-
-#include "idpf_controlq_api.h"
-
-/* Maximum buffer length for all control queue types */
-#define IDPF_CTLQ_MAX_BUF_LEN 4096
-
-#define IDPF_CTLQ_DESC(R, i) \
- (&(((struct idpf_ctlq_desc *)((R)->desc_ring.va))[i]))
-
-#define IDPF_CTLQ_DESC_UNUSED(R) \
- ((u16)((((R)->next_to_clean > (R)->next_to_use) ? 0 : (R)->ring_size) + \
- (R)->next_to_clean - (R)->next_to_use - 1))
-
-/* Control Queue default settings */
-#define IDPF_CTRL_SQ_CMD_TIMEOUT 250 /* msecs */
-
-struct idpf_ctlq_desc {
- /* Control queue descriptor flags */
- __le16 flags;
- /* Control queue message opcode */
- __le16 opcode;
- __le16 datalen; /* 0 for direct commands */
- union {
- __le16 ret_val;
- __le16 pfid_vfid;
-#define IDPF_CTLQ_DESC_VF_ID_S 0
-#define IDPF_CTLQ_DESC_VF_ID_M (0x7FF << IDPF_CTLQ_DESC_VF_ID_S)
-#define IDPF_CTLQ_DESC_PF_ID_S 11
-#define IDPF_CTLQ_DESC_PF_ID_M (0x1F << IDPF_CTLQ_DESC_PF_ID_S)
- };
-
- /* Virtchnl message opcode and virtchnl descriptor type
- * v_opcode=[27:0], v_dtype=[31:28]
- */
- __le32 v_opcode_dtype;
- /* Virtchnl return value */
- __le32 v_retval;
- union {
- struct {
- __le32 param0;
- __le32 param1;
- __le32 param2;
- __le32 param3;
- } direct;
- struct {
- __le32 param0;
- __le16 sw_cookie;
- /* Virtchnl flags */
- __le16 v_flags;
- __le32 addr_high;
- __le32 addr_low;
- } indirect;
- u8 raw[16];
- } params;
-};
-
-/* Flags sub-structure
- * |0 |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |
- * |DD |CMP|ERR| * RSV * |FTYPE | *RSV* |RD |VFC|BUF| HOST_ID |
- */
-/* command flags and offsets */
-#define IDPF_CTLQ_FLAG_DD_S 0
-#define IDPF_CTLQ_FLAG_CMP_S 1
-#define IDPF_CTLQ_FLAG_ERR_S 2
-#define IDPF_CTLQ_FLAG_FTYPE_S 6
-#define IDPF_CTLQ_FLAG_RD_S 10
-#define IDPF_CTLQ_FLAG_VFC_S 11
-#define IDPF_CTLQ_FLAG_BUF_S 12
-#define IDPF_CTLQ_FLAG_HOST_ID_S 13
-
-#define IDPF_CTLQ_FLAG_DD BIT(IDPF_CTLQ_FLAG_DD_S) /* 0x1 */
-#define IDPF_CTLQ_FLAG_CMP BIT(IDPF_CTLQ_FLAG_CMP_S) /* 0x2 */
-#define IDPF_CTLQ_FLAG_ERR BIT(IDPF_CTLQ_FLAG_ERR_S) /* 0x4 */
-#define IDPF_CTLQ_FLAG_FTYPE_VM BIT(IDPF_CTLQ_FLAG_FTYPE_S) /* 0x40 */
-#define IDPF_CTLQ_FLAG_FTYPE_PF BIT(IDPF_CTLQ_FLAG_FTYPE_S + 1) /* 0x80 */
-#define IDPF_CTLQ_FLAG_RD BIT(IDPF_CTLQ_FLAG_RD_S) /* 0x400 */
-#define IDPF_CTLQ_FLAG_VFC BIT(IDPF_CTLQ_FLAG_VFC_S) /* 0x800 */
-#define IDPF_CTLQ_FLAG_BUF BIT(IDPF_CTLQ_FLAG_BUF_S) /* 0x1000 */
-
-/* Host ID is a special field that has 3b and not a 1b flag */
-#define IDPF_CTLQ_FLAG_HOST_ID_M MAKE_MASK(0x7000UL, IDPF_CTLQ_FLAG_HOST_ID_S)
-
-struct idpf_mbxq_desc {
- u8 pad[8]; /* CTLQ flags/opcode/len/retval fields */
- u32 chnl_opcode; /* avoid confusion with desc->opcode */
- u32 chnl_retval; /* ditto for desc->retval */
- u32 pf_vf_id; /* used by CP when sending to PF */
-};
-
-/* Max number of MMIO regions not including the mailbox and rstat regions in
- * the fallback case when the whole bar is mapped.
- */
-#define IDPF_MMIO_MAP_FALLBACK_MAX_REMAINING 3
-
-struct idpf_mmio_reg {
- void __iomem *vaddr;
- resource_size_t addr_start;
- resource_size_t addr_len;
-};
-
-/* Define the driver hardware struct to replace other control structs as needed
- * Align to ctlq_hw_info
- */
-struct idpf_hw {
- /* Array of remaining LAN BAR regions */
- int num_lan_regs;
- struct idpf_mmio_reg *lan_regs;
-
- struct idpf_adapter *back;
-
- /* control queue - send and receive */
- struct idpf_ctlq_info *asq;
- struct idpf_ctlq_info *arq;
-
- /* pci info */
- u16 device_id;
- u16 vendor_id;
- u16 subsystem_device_id;
- u16 subsystem_vendor_id;
- u8 revision_id;
- bool adapter_stopped;
-
- struct list_head cq_list_head;
-};
-
-int idpf_ctlq_alloc_ring_res(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq);
-
-void idpf_ctlq_dealloc_ring_res(struct idpf_hw *hw, struct idpf_ctlq_info *cq);
-
-/* prototype for functions used for dynamic memory allocation */
-void *idpf_alloc_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem,
- u64 size);
-void idpf_free_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem);
-#endif /* _IDPF_CONTROLQ_H_ */
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq_api.h b/drivers/net/ethernet/intel/idpf/idpf_controlq_api.h
deleted file mode 100644
index 3414c5f9a831..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq_api.h
+++ /dev/null
@@ -1,177 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/* Copyright (C) 2023 Intel Corporation */
-
-#ifndef _IDPF_CONTROLQ_API_H_
-#define _IDPF_CONTROLQ_API_H_
-
-#include "idpf_mem.h"
-
-struct idpf_hw;
-
-/* Used for queue init, response and events */
-enum idpf_ctlq_type {
- IDPF_CTLQ_TYPE_MAILBOX_TX = 0,
- IDPF_CTLQ_TYPE_MAILBOX_RX = 1,
- IDPF_CTLQ_TYPE_CONFIG_TX = 2,
- IDPF_CTLQ_TYPE_CONFIG_RX = 3,
- IDPF_CTLQ_TYPE_EVENT_RX = 4,
- IDPF_CTLQ_TYPE_RDMA_TX = 5,
- IDPF_CTLQ_TYPE_RDMA_RX = 6,
- IDPF_CTLQ_TYPE_RDMA_COMPL = 7
-};
-
-/* Generic Control Queue Structures */
-struct idpf_ctlq_reg {
- /* used for queue tracking */
- u32 head;
- u32 tail;
- /* Below applies only to default mb (if present) */
- u32 len;
- u32 bah;
- u32 bal;
- u32 len_mask;
- u32 len_ena_mask;
- u32 head_mask;
-};
-
-/* Generic queue msg structure */
-struct idpf_ctlq_msg {
- u8 vmvf_type; /* represents the source of the message on recv */
-#define IDPF_VMVF_TYPE_VF 0
-#define IDPF_VMVF_TYPE_VM 1
-#define IDPF_VMVF_TYPE_PF 2
- u8 host_id;
- /* 3b field used only when sending a message to CP - to be used in
- * combination with target func_id to route the message
- */
-#define IDPF_HOST_ID_MASK 0x7
-
- u16 opcode;
- u16 data_len; /* data_len = 0 when no payload is attached */
- union {
- u16 func_id; /* when sending a message */
- u16 status; /* when receiving a message */
- };
- union {
- struct {
- u32 chnl_opcode;
- u32 chnl_retval;
- } mbx;
- } cookie;
- union {
-#define IDPF_DIRECT_CTX_SIZE 16
-#define IDPF_INDIRECT_CTX_SIZE 8
- /* 16 bytes of context can be provided or 8 bytes of context
- * plus the address of a DMA buffer
- */
- u8 direct[IDPF_DIRECT_CTX_SIZE];
- struct {
- u8 context[IDPF_INDIRECT_CTX_SIZE];
- struct idpf_dma_mem *payload;
- } indirect;
- struct {
- u32 rsvd;
- u16 data;
- u16 flags;
- } sw_cookie;
- } ctx;
-};
-
-/* Generic queue info structures */
-/* MB, CONFIG and EVENT q do not have extended info */
-struct idpf_ctlq_create_info {
- enum idpf_ctlq_type type;
- int id; /* absolute queue offset passed as input
- * -1 for default mailbox if present
- */
- u16 len; /* Queue length passed as input */
- u16 buf_size; /* buffer size passed as input */
- u64 base_address; /* output, HPA of the Queue start */
- struct idpf_ctlq_reg reg; /* registers accessed by ctlqs */
-
- int ext_info_size;
- void *ext_info; /* Specific to q type */
-};
-
-/* Control Queue information */
-struct idpf_ctlq_info {
- struct list_head cq_list;
-
- enum idpf_ctlq_type cq_type;
- int q_id;
- spinlock_t cq_lock; /* control queue lock */
- /* used for interrupt processing */
- u16 next_to_use;
- u16 next_to_clean;
- u16 next_to_post; /* starting descriptor to post buffers
- * to after recev
- */
-
- struct idpf_dma_mem desc_ring; /* descriptor ring memory
- * idpf_dma_mem is defined in OSdep.h
- */
- union {
- struct idpf_dma_mem **rx_buff;
- struct idpf_ctlq_msg **tx_msg;
- } bi;
-
- u16 buf_size; /* queue buffer size */
- u16 ring_size; /* Number of descriptors */
- struct idpf_ctlq_reg reg; /* registers accessed by ctlqs */
-};
-
-/**
- * enum idpf_mbx_opc - PF/VF mailbox commands
- * @idpf_mbq_opc_send_msg_to_cp: used by PF or VF to send a message to its CP
- * @idpf_mbq_opc_send_msg_to_peer_drv: used by PF or VF to send a message to
- * any peer driver
- */
-enum idpf_mbx_opc {
- idpf_mbq_opc_send_msg_to_cp = 0x0801,
- idpf_mbq_opc_send_msg_to_peer_drv = 0x0804,
-};
-
-/* API supported for control queue management */
-/* Will init all required q including default mb. "q_info" is an array of
- * create_info structs equal to the number of control queues to be created.
- */
-int idpf_ctlq_init(struct idpf_hw *hw, u8 num_q,
- struct idpf_ctlq_create_info *q_info);
-
-/* Allocate and initialize a single control queue, which will be added to the
- * control queue list; returns a handle to the created control queue
- */
-int idpf_ctlq_add(struct idpf_hw *hw,
- struct idpf_ctlq_create_info *qinfo,
- struct idpf_ctlq_info **cq);
-
-/* Deinitialize and deallocate a single control queue */
-void idpf_ctlq_remove(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq);
-
-/* Sends messages to HW and will also free the buffer*/
-int idpf_ctlq_send(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq,
- u16 num_q_msg,
- struct idpf_ctlq_msg q_msg[]);
-
-/* Receives messages and called by interrupt handler/polling
- * initiated by app/process. Also caller is supposed to free the buffers
- */
-int idpf_ctlq_recv(struct idpf_ctlq_info *cq, u16 *num_q_msg,
- struct idpf_ctlq_msg *q_msg);
-
-/* Reclaims send descriptors on HW write back */
-int idpf_ctlq_clean_sq(struct idpf_ctlq_info *cq, u16 *clean_count,
- struct idpf_ctlq_msg *msg_status[]);
-
-/* Indicate RX buffers are done being processed */
-int idpf_ctlq_post_rx_buffs(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq,
- u16 *buff_count,
- struct idpf_dma_mem **buffs);
-
-/* Will destroy all q including the default mb */
-void idpf_ctlq_deinit(struct idpf_hw *hw);
-
-#endif /* _IDPF_CONTROLQ_API_H_ */
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c b/drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c
deleted file mode 100644
index d4d488c7cfd6..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c
+++ /dev/null
@@ -1,169 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-only
-/* Copyright (C) 2023 Intel Corporation */
-
-#include "idpf_controlq.h"
-
-/**
- * idpf_ctlq_alloc_desc_ring - Allocate Control Queue (CQ) rings
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- */
-static int idpf_ctlq_alloc_desc_ring(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- size_t size = cq->ring_size * sizeof(struct idpf_ctlq_desc);
-
- cq->desc_ring.va = idpf_alloc_dma_mem(hw, &cq->desc_ring, size);
- if (!cq->desc_ring.va)
- return -ENOMEM;
-
- return 0;
-}
-
-/**
- * idpf_ctlq_alloc_bufs - Allocate Control Queue (CQ) buffers
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * Allocate the buffer head for all control queues, and if it's a receive
- * queue, allocate DMA buffers
- */
-static int idpf_ctlq_alloc_bufs(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- int i;
-
- /* Do not allocate DMA buffers for transmit queues */
- if (cq->cq_type == IDPF_CTLQ_TYPE_MAILBOX_TX)
- return 0;
-
- /* We'll be allocating the buffer info memory first, then we can
- * allocate the mapped buffers for the event processing
- */
- cq->bi.rx_buff = kzalloc_objs(struct idpf_dma_mem *, cq->ring_size);
- if (!cq->bi.rx_buff)
- return -ENOMEM;
-
- /* allocate the mapped buffers (except for the last one) */
- for (i = 0; i < cq->ring_size - 1; i++) {
- struct idpf_dma_mem *bi;
- int num = 1; /* number of idpf_dma_mem to be allocated */
-
- cq->bi.rx_buff[i] = kzalloc_objs(struct idpf_dma_mem, num);
- if (!cq->bi.rx_buff[i])
- goto unwind_alloc_cq_bufs;
-
- bi = cq->bi.rx_buff[i];
-
- bi->va = idpf_alloc_dma_mem(hw, bi, cq->buf_size);
- if (!bi->va) {
- /* unwind will not free the failed entry */
- kfree(cq->bi.rx_buff[i]);
- goto unwind_alloc_cq_bufs;
- }
- }
-
- return 0;
-
-unwind_alloc_cq_bufs:
- /* don't try to free the one that failed... */
- i--;
- for (; i >= 0; i--) {
- idpf_free_dma_mem(hw, cq->bi.rx_buff[i]);
- kfree(cq->bi.rx_buff[i]);
- }
- kfree(cq->bi.rx_buff);
-
- return -ENOMEM;
-}
-
-/**
- * idpf_ctlq_free_desc_ring - Free Control Queue (CQ) rings
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * This assumes the posted send buffers have already been cleaned
- * and de-allocated
- */
-static void idpf_ctlq_free_desc_ring(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- idpf_free_dma_mem(hw, &cq->desc_ring);
-}
-
-/**
- * idpf_ctlq_free_bufs - Free CQ buffer info elements
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * Free the DMA buffers for RX queues, and DMA buffer header for both RX and TX
- * queues. The upper layers are expected to manage freeing of TX DMA buffers
- */
-static void idpf_ctlq_free_bufs(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- void *bi;
-
- if (cq->cq_type == IDPF_CTLQ_TYPE_MAILBOX_RX) {
- int i;
-
- /* free DMA buffers for rx queues*/
- for (i = 0; i < cq->ring_size; i++) {
- if (cq->bi.rx_buff[i]) {
- idpf_free_dma_mem(hw, cq->bi.rx_buff[i]);
- kfree(cq->bi.rx_buff[i]);
- }
- }
-
- bi = (void *)cq->bi.rx_buff;
- } else {
- bi = (void *)cq->bi.tx_msg;
- }
-
- /* free the buffer header */
- kfree(bi);
-}
-
-/**
- * idpf_ctlq_dealloc_ring_res - Free memory allocated for control queue
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * Free the memory used by the ring, buffers and other related structures
- */
-void idpf_ctlq_dealloc_ring_res(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- /* free ring buffers and the ring itself */
- idpf_ctlq_free_bufs(hw, cq);
- idpf_ctlq_free_desc_ring(hw, cq);
-}
-
-/**
- * idpf_ctlq_alloc_ring_res - allocate memory for descriptor ring and bufs
- * @hw: pointer to hw struct
- * @cq: pointer to control queue struct
- *
- * Do *NOT* hold cq_lock when calling this as the memory allocation routines
- * called are not going to be atomic context safe
- */
-int idpf_ctlq_alloc_ring_res(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- int err;
-
- /* allocate the ring memory */
- err = idpf_ctlq_alloc_desc_ring(hw, cq);
- if (err)
- return err;
-
- /* allocate buffers in the rings */
- err = idpf_ctlq_alloc_bufs(hw, cq);
- if (err)
- goto idpf_init_cq_free_ring;
-
- /* success! */
- return 0;
-
-idpf_init_cq_free_ring:
- idpf_free_dma_mem(hw, &cq->desc_ring);
-
- return err;
-}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_dev.c b/drivers/net/ethernet/intel/idpf/idpf_dev.c
index e36b0017186f..3a357d5dea20 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_dev.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_dev.c
@@ -10,44 +10,32 @@
/**
* idpf_ctlq_reg_init - initialize default mailbox registers
- * @adapter: adapter structure
- * @cq: pointer to the array of create control queues
+ * @mmio: struct that contains MMIO region info
+ * @cci: struct where the register offset pointer to be copied to
*/
-static void idpf_ctlq_reg_init(struct idpf_adapter *adapter,
- struct idpf_ctlq_create_info *cq)
+static void idpf_ctlq_reg_init(struct libie_mmio_info *mmio,
+ struct libie_ctlq_create_info *cci)
{
- int i;
-
- for (i = 0; i < IDPF_NUM_DFLT_MBX_Q; i++) {
- struct idpf_ctlq_create_info *ccq = cq + i;
-
- switch (ccq->type) {
- case IDPF_CTLQ_TYPE_MAILBOX_TX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = PF_FW_ATQH;
- ccq->reg.tail = PF_FW_ATQT;
- ccq->reg.len = PF_FW_ATQLEN;
- ccq->reg.bah = PF_FW_ATQBAH;
- ccq->reg.bal = PF_FW_ATQBAL;
- ccq->reg.len_mask = PF_FW_ATQLEN_ATQLEN_M;
- ccq->reg.len_ena_mask = PF_FW_ATQLEN_ATQENABLE_M;
- ccq->reg.head_mask = PF_FW_ATQH_ATQH_M;
- break;
- case IDPF_CTLQ_TYPE_MAILBOX_RX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = PF_FW_ARQH;
- ccq->reg.tail = PF_FW_ARQT;
- ccq->reg.len = PF_FW_ARQLEN;
- ccq->reg.bah = PF_FW_ARQBAH;
- ccq->reg.bal = PF_FW_ARQBAL;
- ccq->reg.len_mask = PF_FW_ARQLEN_ARQLEN_M;
- ccq->reg.len_ena_mask = PF_FW_ARQLEN_ARQENABLE_M;
- ccq->reg.head_mask = PF_FW_ARQH_ARQH_M;
- break;
- default:
- break;
- }
- }
+ struct libie_ctlq_reg *tx_reg = &cci[LIBIE_CTLQ_TYPE_TX].reg;
+ struct libie_ctlq_reg *rx_reg = &cci[LIBIE_CTLQ_TYPE_RX].reg;
+
+ tx_reg->head = libie_pci_get_mmio_addr(mmio, PF_FW_ATQH);
+ tx_reg->tail = libie_pci_get_mmio_addr(mmio, PF_FW_ATQT);
+ tx_reg->len = libie_pci_get_mmio_addr(mmio, PF_FW_ATQLEN);
+ tx_reg->addr_high = libie_pci_get_mmio_addr(mmio, PF_FW_ATQBAH);
+ tx_reg->addr_low = libie_pci_get_mmio_addr(mmio, PF_FW_ATQBAL);
+ tx_reg->len_mask = PF_FW_ATQLEN_ATQLEN_M;
+ tx_reg->len_ena_mask = PF_FW_ATQLEN_ATQENABLE_M;
+ tx_reg->head_mask = PF_FW_ATQH_ATQH_M;
+
+ rx_reg->head = libie_pci_get_mmio_addr(mmio, PF_FW_ARQH);
+ rx_reg->tail = libie_pci_get_mmio_addr(mmio, PF_FW_ARQT);
+ rx_reg->len = libie_pci_get_mmio_addr(mmio, PF_FW_ARQLEN);
+ rx_reg->addr_high = libie_pci_get_mmio_addr(mmio, PF_FW_ARQBAH);
+ rx_reg->addr_low = libie_pci_get_mmio_addr(mmio, PF_FW_ARQBAL);
+ rx_reg->len_mask = PF_FW_ARQLEN_ARQLEN_M;
+ rx_reg->len_ena_mask = PF_FW_ARQLEN_ARQENABLE_M;
+ rx_reg->head_mask = PF_FW_ARQH_ARQH_M;
}
/**
diff --git a/drivers/net/ethernet/intel/idpf/idpf_ethtool.c b/drivers/net/ethernet/intel/idpf/idpf_ethtool.c
index bb99d9e7c65d..95c45f12b0f9 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_ethtool.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_ethtool.c
@@ -225,7 +225,7 @@ static int idpf_add_flow_steer(struct net_device *netdev,
spin_unlock_bh(&vport_config->flow_steer_list_lock);
if (err)
- goto out;
+ goto out_free_fltr;
rule->vport_id = cpu_to_le32(vport->vport_id);
rule->count = cpu_to_le32(1);
@@ -252,17 +252,15 @@ static int idpf_add_flow_steer(struct net_device *netdev,
break;
default:
err = -EINVAL;
- goto out;
+ goto out_free_fltr;
}
err = idpf_add_del_fsteer_filters(vport->adapter, rule,
VIRTCHNL2_OP_ADD_FLOW_RULE);
- if (err)
- goto out;
-
- if (info->status != cpu_to_le32(VIRTCHNL2_FLOW_RULE_SUCCESS)) {
- err = -EIO;
- goto out;
+ if (err) {
+ /* virtchnl2 rule is already consumed */
+ kfree(fltr);
+ return err;
}
/* Save a copy of the user's flow spec so ethtool can later retrieve it */
@@ -274,9 +272,10 @@ static int idpf_add_flow_steer(struct net_device *netdev,
user_config->num_fsteer_fltrs++;
spin_unlock_bh(&vport_config->flow_steer_list_lock);
- goto out_free_rule;
-out:
+ return 0;
+
+out_free_fltr:
kfree(fltr);
out_free_rule:
kfree(rule);
@@ -319,12 +318,7 @@ static int idpf_del_flow_steer(struct net_device *netdev,
err = idpf_add_del_fsteer_filters(vport->adapter, rule,
VIRTCHNL2_OP_DEL_FLOW_RULE);
if (err)
- goto out;
-
- if (info->status != cpu_to_le32(VIRTCHNL2_FLOW_RULE_SUCCESS)) {
- err = -EIO;
- goto out;
- }
+ return err;
spin_lock_bh(&vport_config->flow_steer_list_lock);
list_for_each_entry_safe(f, iter,
@@ -340,8 +334,6 @@ static int idpf_del_flow_steer(struct net_device *netdev,
out_unlock:
spin_unlock_bh(&vport_config->flow_steer_list_lock);
-out:
- kfree(rule);
return err;
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_lib.c b/drivers/net/ethernet/intel/idpf/idpf_lib.c
index 875472ae77fd..6751534e022b 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_lib.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_lib.c
@@ -1371,6 +1371,7 @@ void idpf_statistics_task(struct work_struct *work)
*/
void idpf_mbx_task(struct work_struct *work)
{
+ struct libie_ctlq_xn_recv_params xn_params;
struct idpf_adapter *adapter;
adapter = container_of(work, struct idpf_adapter, mbx_task.work);
@@ -1381,7 +1382,14 @@ void idpf_mbx_task(struct work_struct *work)
queue_delayed_work(adapter->mbx_wq, &adapter->mbx_task,
usecs_to_jiffies(300));
- idpf_recv_mb_msg(adapter, adapter->hw.arq);
+ xn_params = (struct libie_ctlq_xn_recv_params) {
+ .xnm = adapter->xnm,
+ .ctlq = adapter->arq,
+ .ctlq_msg_handler = idpf_recv_event_msg,
+ .budget = LIBIE_CTLQ_MAX_XN_ENTRIES,
+ };
+
+ libie_ctlq_xn_recv(&xn_params);
}
/**
@@ -1984,7 +1992,8 @@ void idpf_vc_event_task(struct work_struct *work)
return;
func_reset:
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ if (adapter->xnm)
+ libie_ctlq_xn_shutdown(adapter->xnm);
drv_load:
set_bit(IDPF_HR_RESET_IN_PROG, adapter->flags);
idpf_init_hard_reset(adapter);
@@ -2567,44 +2576,6 @@ static int idpf_set_mac(struct net_device *netdev, void *p)
return err;
}
-/**
- * idpf_alloc_dma_mem - Allocate dma memory
- * @hw: pointer to hw struct
- * @mem: pointer to dma_mem struct
- * @size: size of the memory to allocate
- */
-void *idpf_alloc_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem, u64 size)
-{
- struct idpf_adapter *adapter = hw->back;
- size_t sz = ALIGN(size, 4096);
-
- /* The control queue resources are freed under a spinlock, contiguous
- * pages will avoid IOMMU remapping and the use vmap (and vunmap in
- * dma_free_*() path.
- */
- mem->va = dma_alloc_attrs(&adapter->pdev->dev, sz, &mem->pa,
- GFP_KERNEL, DMA_ATTR_FORCE_CONTIGUOUS);
- mem->size = sz;
-
- return mem->va;
-}
-
-/**
- * idpf_free_dma_mem - Free the allocated dma memory
- * @hw: pointer to hw struct
- * @mem: pointer to dma_mem struct
- */
-void idpf_free_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem)
-{
- struct idpf_adapter *adapter = hw->back;
-
- dma_free_attrs(&adapter->pdev->dev, mem->size,
- mem->va, mem->pa, DMA_ATTR_FORCE_CONTIGUOUS);
- mem->size = 0;
- mem->va = NULL;
- mem->pa = 0;
-}
-
static int idpf_hwtstamp_set(struct net_device *netdev,
struct kernel_hwtstamp_config *config,
struct netlink_ext_ack *extack)
diff --git a/drivers/net/ethernet/intel/idpf/idpf_main.c b/drivers/net/ethernet/intel/idpf/idpf_main.c
index 28dc6fabab3e..0bf673d84960 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_main.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_main.c
@@ -170,8 +170,6 @@ static void idpf_remove(struct pci_dev *pdev)
adapter->vport_config = NULL;
kfree(adapter->netdevs);
adapter->netdevs = NULL;
- kfree(adapter->vcxn_mngr);
- adapter->vcxn_mngr = NULL;
mutex_destroy(&adapter->vport_ctrl_lock);
mutex_destroy(&adapter->vector_lock);
@@ -239,7 +237,6 @@ static int idpf_cfg_device(struct idpf_adapter *adapter)
pci_dbg(pdev, "PCIe PTM is not supported by PCIe bus/controller\n");
pci_set_drvdata(pdev, adapter);
- adapter->hw.back = adapter;
return 0;
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_mem.h b/drivers/net/ethernet/intel/idpf/idpf_mem.h
deleted file mode 100644
index 2aaabdc02dd2..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_mem.h
+++ /dev/null
@@ -1,20 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/* Copyright (C) 2023 Intel Corporation */
-
-#ifndef _IDPF_MEM_H_
-#define _IDPF_MEM_H_
-
-#include <linux/io.h>
-
-struct idpf_dma_mem {
- void *va;
- dma_addr_t pa;
- size_t size;
-};
-
-#define idpf_mbx_wr32(a, reg, value) writel((value), ((a)->mbx.vaddr + (reg)))
-#define idpf_mbx_rd32(a, reg) readl((a)->mbx.vaddr + (reg))
-#define idpf_mbx_wr64(a, reg, value) writeq((value), ((a)->mbx.vaddr + (reg)))
-#define idpf_mbx_rd64(a, reg) readq((a)->mbx.vaddr + (reg))
-
-#endif /* _IDPF_MEM_H_ */
diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
index 08b048664029..068c432ee833 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
+++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
@@ -236,7 +236,7 @@ enum idpf_tx_ctx_desc_eipt_offload {
(sizeof(u16) * IDPF_RX_MAX_PTYPE_PROTO_IDS))
#define IDPF_RX_PTYPE_HDR_SZ sizeof(struct virtchnl2_get_ptype_info)
#define IDPF_RX_MAX_PTYPES_PER_BUF \
- DIV_ROUND_DOWN_ULL((IDPF_CTLQ_MAX_BUF_LEN - IDPF_RX_PTYPE_HDR_SZ), \
+ DIV_ROUND_DOWN_ULL(LIBIE_CTLQ_MAX_BUF_LEN - IDPF_RX_PTYPE_HDR_SZ, \
IDPF_RX_MAX_PTYPE_SZ)
#define IDPF_GET_PTYPE_SIZE(p) struct_size((p), proto_id, (p)->proto_id_count)
diff --git a/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c b/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
index 98b8f678bd9a..152ea03f54b9 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
@@ -9,42 +9,32 @@
/**
* idpf_vf_ctlq_reg_init - initialize default mailbox registers
- * @adapter: adapter structure
- * @cq: pointer to the array of create control queues
+ * @mmio: struct that contains MMIO region info
+ * @cci: struct where the register offset pointer to be copied to
*/
-static void idpf_vf_ctlq_reg_init(struct idpf_adapter *adapter,
- struct idpf_ctlq_create_info *cq)
+static void idpf_vf_ctlq_reg_init(struct libie_mmio_info *mmio,
+ struct libie_ctlq_create_info *cci)
{
- for (int i = 0; i < IDPF_NUM_DFLT_MBX_Q; i++) {
- struct idpf_ctlq_create_info *ccq = cq + i;
-
- switch (ccq->type) {
- case IDPF_CTLQ_TYPE_MAILBOX_TX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = VF_ATQH;
- ccq->reg.tail = VF_ATQT;
- ccq->reg.len = VF_ATQLEN;
- ccq->reg.bah = VF_ATQBAH;
- ccq->reg.bal = VF_ATQBAL;
- ccq->reg.len_mask = VF_ATQLEN_ATQLEN_M;
- ccq->reg.len_ena_mask = VF_ATQLEN_ATQENABLE_M;
- ccq->reg.head_mask = VF_ATQH_ATQH_M;
- break;
- case IDPF_CTLQ_TYPE_MAILBOX_RX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = VF_ARQH;
- ccq->reg.tail = VF_ARQT;
- ccq->reg.len = VF_ARQLEN;
- ccq->reg.bah = VF_ARQBAH;
- ccq->reg.bal = VF_ARQBAL;
- ccq->reg.len_mask = VF_ARQLEN_ARQLEN_M;
- ccq->reg.len_ena_mask = VF_ARQLEN_ARQENABLE_M;
- ccq->reg.head_mask = VF_ARQH_ARQH_M;
- break;
- default:
- break;
- }
- }
+ struct libie_ctlq_reg *tx_reg = &cci[LIBIE_CTLQ_TYPE_TX].reg;
+ struct libie_ctlq_reg *rx_reg = &cci[LIBIE_CTLQ_TYPE_RX].reg;
+
+ tx_reg->head = libie_pci_get_mmio_addr(mmio, VF_ATQH);
+ tx_reg->tail = libie_pci_get_mmio_addr(mmio, VF_ATQT);
+ tx_reg->len = libie_pci_get_mmio_addr(mmio, VF_ATQLEN);
+ tx_reg->addr_high = libie_pci_get_mmio_addr(mmio, VF_ATQBAH);
+ tx_reg->addr_low = libie_pci_get_mmio_addr(mmio, VF_ATQBAL);
+ tx_reg->len_mask = VF_ATQLEN_ATQLEN_M;
+ tx_reg->len_ena_mask = VF_ATQLEN_ATQENABLE_M;
+ tx_reg->head_mask = VF_ATQH_ATQH_M;
+
+ rx_reg->head = libie_pci_get_mmio_addr(mmio, VF_ARQH);
+ rx_reg->tail = libie_pci_get_mmio_addr(mmio, VF_ARQT);
+ rx_reg->len = libie_pci_get_mmio_addr(mmio, VF_ARQLEN);
+ rx_reg->addr_high = libie_pci_get_mmio_addr(mmio, VF_ARQBAH);
+ rx_reg->addr_low = libie_pci_get_mmio_addr(mmio, VF_ARQBAL);
+ rx_reg->len_mask = VF_ARQLEN_ARQLEN_M;
+ rx_reg->len_ena_mask = VF_ARQLEN_ARQENABLE_M;
+ rx_reg->head_mask = VF_ARQH_ARQH_M;
}
/**
@@ -160,8 +150,7 @@ static void idpf_vf_trigger_reset(struct idpf_adapter *adapter,
/* Do not send VIRTCHNL2_OP_RESET_VF message on driver unload */
if (trig_cause == IDPF_HR_FUNC_RESET &&
!test_bit(IDPF_REMOVE_IN_PROG, adapter->flags))
- idpf_send_mb_msg(adapter, adapter->hw.asq,
- VIRTCHNL2_OP_RESET_VF, 0, NULL, 0);
+ idpf_send_vf_reset_msg(adapter);
}
/**
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
index 357ff3e86b2e..84155d683773 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
@@ -9,20 +9,6 @@
#include "idpf_virtchnl.h"
#include "idpf_ptp.h"
-/**
- * struct idpf_vc_xn_manager - Manager for tracking transactions
- * @ring: backing and lookup for transactions
- * @free_xn_bm: bitmap for free transactions
- * @xn_bm_lock: make bitmap access synchronous where necessary
- * @salt: used to make cookie unique every message
- */
-struct idpf_vc_xn_manager {
- struct idpf_vc_xn ring[IDPF_VC_XN_RING_LEN];
- DECLARE_BITMAP(free_xn_bm, IDPF_VC_XN_RING_LEN);
- spinlock_t xn_bm_lock;
- u8 salt;
-};
-
/**
* idpf_vid_to_vport - Translate vport id to vport pointer
* @adapter: private data struct
@@ -83,79 +69,64 @@ static void idpf_handle_event_link(struct idpf_adapter *adapter,
/**
* idpf_recv_event_msg - Receive virtchnl event message
- * @adapter: Driver specific private structure
+ * @ctx: control queue context
* @ctlq_msg: message to copy from
*
* Receive virtchnl event message
*/
-static void idpf_recv_event_msg(struct idpf_adapter *adapter,
- struct idpf_ctlq_msg *ctlq_msg)
+void idpf_recv_event_msg(struct libie_ctlq_ctx *ctx,
+ struct libie_ctlq_msg *ctlq_msg)
{
- int payload_size = ctlq_msg->ctx.indirect.payload->size;
+ struct kvec *buff = &ctlq_msg->recv_mem;
+ int payload_size = buff->iov_len;
+ struct idpf_adapter *adapter;
struct virtchnl2_event *v2e;
u32 event;
+ adapter = container_of(ctx, struct idpf_adapter, ctlq_ctx);
if (payload_size < sizeof(*v2e)) {
dev_err_ratelimited(&adapter->pdev->dev, "Failed to receive valid payload for event msg (op %d len %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode,
+ ctlq_msg->chnl_opcode,
payload_size);
- return;
+ goto free_rx_buf;
}
- v2e = (struct virtchnl2_event *)ctlq_msg->ctx.indirect.payload->va;
+ v2e = (struct virtchnl2_event *)buff->iov_base;
event = le32_to_cpu(v2e->event);
switch (event) {
case VIRTCHNL2_EVENT_LINK_CHANGE:
idpf_handle_event_link(adapter, v2e);
- return;
+ break;
default:
dev_err(&adapter->pdev->dev,
"Unknown event %d from PF\n", event);
break;
}
+
+free_rx_buf:
+ libie_ctlq_release_rx_buf(buff);
}
/**
* idpf_mb_clean - Reclaim the send mailbox queue entries
* @adapter: driver specific private structure
* @asq: send control queue info
+ * @deinit: release all buffers before destroying the queue
*
- * Reclaim the send mailbox queue entries to be used to send further messages
- *
- * Return: 0 on success, negative on failure
+ * This is a helper function to clean the send mailbox queue entries.
*/
-static int idpf_mb_clean(struct idpf_adapter *adapter,
- struct idpf_ctlq_info *asq)
+static void idpf_mb_clean(struct idpf_adapter *adapter,
+ struct libie_ctlq_info *asq, bool deinit)
{
- u16 i, num_q_msg = IDPF_DFLT_MBX_Q_LEN;
- struct idpf_ctlq_msg **q_msg;
- struct idpf_dma_mem *dma_mem;
- int err;
-
- q_msg = kzalloc_objs(struct idpf_ctlq_msg *, num_q_msg, GFP_ATOMIC);
- if (!q_msg)
- return -ENOMEM;
-
- err = idpf_ctlq_clean_sq(asq, &num_q_msg, q_msg);
- if (err)
- goto err_kfree;
-
- for (i = 0; i < num_q_msg; i++) {
- if (!q_msg[i])
- continue;
- dma_mem = q_msg[i]->ctx.indirect.payload;
- if (dma_mem)
- dma_free_coherent(&adapter->pdev->dev, dma_mem->size,
- dma_mem->va, dma_mem->pa);
- kfree(q_msg[i]);
- kfree(dma_mem);
- }
-
-err_kfree:
- kfree(q_msg);
+ struct libie_ctlq_xn_clean_params clean_params = {
+ .ctlq = asq,
+ .rel_tx_buf = kfree,
+ .num_msgs = IDPF_DFLT_MBX_Q_LEN,
+ .force = deinit,
+ };
- return err;
+ libie_ctlq_xn_send_clean(&clean_params);
}
#if IS_ENABLED(CONFIG_PTP_1588_CLOCK)
@@ -189,7 +160,7 @@ static bool idpf_ptp_is_mb_msg(u32 op)
* @ctlq_msg: Corresponding control queue message
*/
static void idpf_prepare_ptp_mb_msg(struct idpf_adapter *adapter, u32 op,
- struct idpf_ctlq_msg *ctlq_msg)
+ struct libie_ctlq_msg *ctlq_msg)
{
/* If the message is PTP-related and the secondary mailbox is available,
* send the message through the secondary mailbox.
@@ -197,530 +168,106 @@ static void idpf_prepare_ptp_mb_msg(struct idpf_adapter *adapter, u32 op,
if (!idpf_ptp_is_mb_msg(op) || !adapter->ptp->secondary_mbx.valid)
return;
- ctlq_msg->opcode = idpf_mbq_opc_send_msg_to_peer_drv;
+ ctlq_msg->opcode = LIBIE_CTLQ_SEND_MSG_TO_PEER;
ctlq_msg->func_id = adapter->ptp->secondary_mbx.peer_mbx_q_id;
- ctlq_msg->host_id = adapter->ptp->secondary_mbx.peer_id;
+ ctlq_msg->flags = FIELD_PREP(LIBIE_CTLQ_DESC_FLAG_HOST_ID,
+ adapter->ptp->secondary_mbx.peer_id);
}
#else /* !CONFIG_PTP_1588_CLOCK */
static void idpf_prepare_ptp_mb_msg(struct idpf_adapter *adapter, u32 op,
- struct idpf_ctlq_msg *ctlq_msg)
+ struct libie_ctlq_msg *ctlq_msg)
{ }
#endif /* CONFIG_PTP_1588_CLOCK */
/**
- * idpf_send_mb_msg - Send message over mailbox
+ * idpf_send_mb_msg - send mailbox message to the device control plane
* @adapter: driver specific private structure
- * @asq: control queue to send message to
- * @op: virtchnl opcode
- * @msg_size: size of the payload
- * @msg: pointer to buffer holding the payload
- * @cookie: unique SW generated cookie per message
+ * @xn_params: Xn send parameters to fill
+ * @send_buf: buffer to send
+ * @send_buf_size: size of the send buffer
*
- * Will prepare the control queue message and initiates the send api
+ * Fill the Xn parameters with the required info to send a virtchnl message.
+ * The send buffer is DMA mapped in the libie to avoid memcpy.
*
- * Return: 0 on success, negative on failure
- */
-int idpf_send_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *asq,
- u32 op, u16 msg_size, u8 *msg, u16 cookie)
-{
- struct idpf_ctlq_msg *ctlq_msg;
- struct idpf_dma_mem *dma_mem;
- int err;
-
- /* If we are here and a reset is detected nothing much can be
- * done. This thread should silently abort and expected to
- * be corrected with a new run either by user or driver
- * flows after reset
- */
- if (idpf_is_reset_detected(adapter))
- return 0;
-
- err = idpf_mb_clean(adapter, asq);
- if (err)
- return err;
-
- ctlq_msg = kzalloc_obj(*ctlq_msg, GFP_ATOMIC);
- if (!ctlq_msg)
- return -ENOMEM;
-
- dma_mem = kzalloc_obj(*dma_mem, GFP_ATOMIC);
- if (!dma_mem) {
- err = -ENOMEM;
- goto dma_mem_error;
- }
-
- ctlq_msg->opcode = idpf_mbq_opc_send_msg_to_cp;
- ctlq_msg->func_id = 0;
-
- idpf_prepare_ptp_mb_msg(adapter, op, ctlq_msg);
-
- ctlq_msg->data_len = msg_size;
- ctlq_msg->cookie.mbx.chnl_opcode = op;
- ctlq_msg->cookie.mbx.chnl_retval = 0;
- dma_mem->size = IDPF_CTLQ_MAX_BUF_LEN;
- dma_mem->va = dma_alloc_coherent(&adapter->pdev->dev, dma_mem->size,
- &dma_mem->pa, GFP_ATOMIC);
- if (!dma_mem->va) {
- err = -ENOMEM;
- goto dma_alloc_error;
- }
-
- /* It's possible we're just sending an opcode but no buffer */
- if (msg && msg_size)
- memcpy(dma_mem->va, msg, msg_size);
- ctlq_msg->ctx.indirect.payload = dma_mem;
- ctlq_msg->ctx.sw_cookie.data = cookie;
-
- err = idpf_ctlq_send(&adapter->hw, asq, 1, ctlq_msg);
- if (err)
- goto send_error;
-
- return 0;
-
-send_error:
- dma_free_coherent(&adapter->pdev->dev, dma_mem->size, dma_mem->va,
- dma_mem->pa);
-dma_alloc_error:
- kfree(dma_mem);
-dma_mem_error:
- kfree(ctlq_msg);
-
- return err;
-}
-
-/* API for virtchnl "transaction" support ("xn" for short). */
-
-/**
- * idpf_vc_xn_lock - Request exclusive access to vc transaction
- * @xn: struct idpf_vc_xn* to access
- */
-#define idpf_vc_xn_lock(xn) \
- spin_lock(&(xn)->lock)
-
-/**
- * idpf_vc_xn_unlock - Release exclusive access to vc transaction
- * @xn: struct idpf_vc_xn* to access
- */
-#define idpf_vc_xn_unlock(xn) \
- spin_unlock(&(xn)->lock)
-
-/**
- * idpf_vc_xn_release_bufs - Release reference to reply buffer(s) and
- * reset the transaction state.
- * @xn: struct idpf_vc_xn to update
- */
-static void idpf_vc_xn_release_bufs(struct idpf_vc_xn *xn)
-{
- xn->reply.iov_base = NULL;
- xn->reply.iov_len = 0;
-
- if (xn->state != IDPF_VC_XN_SHUTDOWN)
- xn->state = IDPF_VC_XN_IDLE;
-}
-
-/**
- * idpf_vc_xn_init - Initialize virtchnl transaction object
- * @vcxn_mngr: pointer to vc transaction manager struct
- */
-static void idpf_vc_xn_init(struct idpf_vc_xn_manager *vcxn_mngr)
-{
- int i;
-
- spin_lock_init(&vcxn_mngr->xn_bm_lock);
-
- for (i = 0; i < ARRAY_SIZE(vcxn_mngr->ring); i++) {
- struct idpf_vc_xn *xn = &vcxn_mngr->ring[i];
-
- xn->state = IDPF_VC_XN_IDLE;
- xn->idx = i;
- idpf_vc_xn_release_bufs(xn);
- spin_lock_init(&xn->lock);
- init_completion(&xn->completed);
- }
-
- bitmap_fill(vcxn_mngr->free_xn_bm, IDPF_VC_XN_RING_LEN);
-}
-
-/**
- * idpf_vc_xn_shutdown - Uninitialize virtchnl transaction object
- * @vcxn_mngr: pointer to vc transaction manager struct
+ * Cleanup the mailbox queue entries of the previously sent message to
+ * unmap and release the buffer.
*
- * All waiting threads will be woken-up and their transaction aborted. Further
- * operations on that object will fail.
+ * Return: 0 if the request was successful, -%EBUSY if reset is detected
+ * or Tx control queue is full, other negative error code on failure.
*/
-void idpf_vc_xn_shutdown(struct idpf_vc_xn_manager *vcxn_mngr)
+int idpf_send_mb_msg(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size)
{
- int i;
-
- spin_lock_bh(&vcxn_mngr->xn_bm_lock);
- bitmap_zero(vcxn_mngr->free_xn_bm, IDPF_VC_XN_RING_LEN);
- spin_unlock_bh(&vcxn_mngr->xn_bm_lock);
+ struct libie_ctlq_msg ctlq_msg = {};
- for (i = 0; i < ARRAY_SIZE(vcxn_mngr->ring); i++) {
- struct idpf_vc_xn *xn = &vcxn_mngr->ring[i];
+ if (idpf_is_reset_detected(adapter)) {
+ if (!libie_cp_can_send_onstack(send_buf_size))
+ kfree(send_buf);
- idpf_vc_xn_lock(xn);
- xn->state = IDPF_VC_XN_SHUTDOWN;
- idpf_vc_xn_release_bufs(xn);
- idpf_vc_xn_unlock(xn);
- complete_all(&xn->completed);
+ return -EBUSY;
}
-}
-/**
- * idpf_vc_xn_pop_free - Pop a free transaction from free list
- * @vcxn_mngr: transaction manager to pop from
- *
- * Returns NULL if no free transactions
- */
-static
-struct idpf_vc_xn *idpf_vc_xn_pop_free(struct idpf_vc_xn_manager *vcxn_mngr)
-{
- struct idpf_vc_xn *xn = NULL;
- unsigned long free_idx;
-
- spin_lock_bh(&vcxn_mngr->xn_bm_lock);
- free_idx = find_first_bit(vcxn_mngr->free_xn_bm, IDPF_VC_XN_RING_LEN);
- if (free_idx == IDPF_VC_XN_RING_LEN)
- goto do_unlock;
+ idpf_prepare_ptp_mb_msg(adapter, xn_params->chnl_opcode, &ctlq_msg);
+ xn_params->ctlq_msg = ctlq_msg.opcode ? &ctlq_msg : NULL;
- clear_bit(free_idx, vcxn_mngr->free_xn_bm);
- xn = &vcxn_mngr->ring[free_idx];
- xn->salt = vcxn_mngr->salt++;
+ xn_params->send_buf.iov_base = send_buf;
+ xn_params->send_buf.iov_len = send_buf_size;
+ xn_params->xnm = adapter->xnm;
+ xn_params->ctlq = xn_params->ctlq ? xn_params->ctlq : adapter->asq;
+ xn_params->rel_tx_buf = kfree;
-do_unlock:
- spin_unlock_bh(&vcxn_mngr->xn_bm_lock);
+ idpf_mb_clean(adapter, xn_params->ctlq, false);
- return xn;
+ return libie_ctlq_xn_send(xn_params);
}
/**
- * idpf_vc_xn_push_free - Push a free transaction to free list
- * @vcxn_mngr: transaction manager to push to
- * @xn: transaction to push
- */
-static void idpf_vc_xn_push_free(struct idpf_vc_xn_manager *vcxn_mngr,
- struct idpf_vc_xn *xn)
-{
- idpf_vc_xn_release_bufs(xn);
- spin_lock_bh(&vcxn_mngr->xn_bm_lock);
- set_bit(xn->idx, vcxn_mngr->free_xn_bm);
- spin_unlock_bh(&vcxn_mngr->xn_bm_lock);
-}
-
-/**
- * idpf_vc_xn_exec - Perform a send/recv virtchnl transaction
- * @adapter: driver specific private structure with vcxn_mngr
- * @params: parameters for this particular transaction including
- * -vc_op: virtchannel operation to send
- * -send_buf: kvec iov for send buf and len
- * -recv_buf: kvec iov for recv buf and len (ignored if NULL)
- * -timeout_ms: timeout waiting for a reply (milliseconds)
- * -async: don't wait for message reply, will lose caller context
- * -async_handler: callback to handle async replies
+ * idpf_send_mb_msg_kfree - send mailbox message and free the send buffer
+ * @adapter: driver specific private structure
+ * @xn_params: Xn send parameters to fill
+ * @send_buf: buffer to send, can be released with kfree()
+ * @send_buf_size: size of the send buffer
*
- * @returns >= 0 for success, the size of the initial reply (may or may not be
- * >= @recv_buf.iov_len, but we never overflow @@recv_buf_iov_base). < 0 for
- * error.
- */
-ssize_t idpf_vc_xn_exec(struct idpf_adapter *adapter,
- const struct idpf_vc_xn_params *params)
-{
- const struct kvec *send_buf = ¶ms->send_buf;
- struct idpf_vc_xn *xn;
- ssize_t retval;
- u16 cookie;
-
- xn = idpf_vc_xn_pop_free(adapter->vcxn_mngr);
- /* no free transactions available */
- if (!xn)
- return -ENOSPC;
-
- idpf_vc_xn_lock(xn);
- if (xn->state == IDPF_VC_XN_SHUTDOWN) {
- retval = -ENXIO;
- goto only_unlock;
- } else if (xn->state != IDPF_VC_XN_IDLE) {
- /* We're just going to clobber this transaction even though
- * it's not IDLE. If we don't reuse it we could theoretically
- * eventually leak all the free transactions and not be able to
- * send any messages. At least this way we make an attempt to
- * remain functional even though something really bad is
- * happening that's corrupting what was supposed to be free
- * transactions.
- */
- WARN_ONCE(1, "There should only be idle transactions in free list (idx %d op %d)\n",
- xn->idx, xn->vc_op);
- }
-
- xn->reply = params->recv_buf;
- xn->reply_sz = 0;
- xn->state = params->async ? IDPF_VC_XN_ASYNC : IDPF_VC_XN_WAITING;
- xn->vc_op = params->vc_op;
- xn->async_handler = params->async_handler;
- idpf_vc_xn_unlock(xn);
-
- if (!params->async)
- reinit_completion(&xn->completed);
- cookie = FIELD_PREP(IDPF_VC_XN_SALT_M, xn->salt) |
- FIELD_PREP(IDPF_VC_XN_IDX_M, xn->idx);
-
- retval = idpf_send_mb_msg(adapter, adapter->hw.asq, params->vc_op,
- send_buf->iov_len, send_buf->iov_base,
- cookie);
- if (retval) {
- idpf_vc_xn_lock(xn);
- goto release_and_unlock;
- }
-
- if (params->async)
- return 0;
-
- wait_for_completion_timeout(&xn->completed,
- msecs_to_jiffies(params->timeout_ms));
-
- /* No need to check the return value; we check the final state of the
- * transaction below. It's possible the transaction actually gets more
- * timeout than specified if we get preempted here but after
- * wait_for_completion_timeout returns. This should be non-issue
- * however.
- */
- idpf_vc_xn_lock(xn);
- switch (xn->state) {
- case IDPF_VC_XN_SHUTDOWN:
- retval = -ENXIO;
- goto only_unlock;
- case IDPF_VC_XN_WAITING:
- dev_notice_ratelimited(&adapter->pdev->dev,
- "Transaction timed-out (op:%d cookie:%04x vc_op:%d salt:%02x timeout:%dms)\n",
- params->vc_op, cookie, xn->vc_op,
- xn->salt, params->timeout_ms);
- retval = -ETIME;
- break;
- case IDPF_VC_XN_COMPLETED_SUCCESS:
- retval = xn->reply_sz;
- break;
- case IDPF_VC_XN_COMPLETED_FAILED:
- dev_notice_ratelimited(&adapter->pdev->dev, "Transaction failed (op %d)\n",
- params->vc_op);
- retval = -EIO;
- break;
- default:
- /* Invalid state. */
- WARN_ON_ONCE(1);
- retval = -EIO;
- break;
- }
-
-release_and_unlock:
- idpf_vc_xn_push_free(adapter->vcxn_mngr, xn);
- /* If we receive a VC reply after here, it will be dropped. */
-only_unlock:
- idpf_vc_xn_unlock(xn);
-
- return retval;
-}
-
-/**
- * idpf_vc_xn_forward_async - Handle async reply receives
- * @adapter: private data struct
- * @xn: transaction to handle
- * @ctlq_msg: corresponding ctlq_msg
+ * libie_cp functions consume only buffers above certain size,
+ * smaller buffers are assumed to be on the stack. However, for some
+ * commands with variable message size it makes sense to always use kzalloc(),
+ * which means we have to free smaller buffers ourselves.
*
- * For async sends we're going to lose the caller's context so, if an
- * async_handler was provided, it can deal with the reply, otherwise we'll just
- * check and report if there is an error.
+ * Return: 0 if no unexpected errors were encountered,
+ * negative error code otherwise.
*/
-static int
-idpf_vc_xn_forward_async(struct idpf_adapter *adapter, struct idpf_vc_xn *xn,
- const struct idpf_ctlq_msg *ctlq_msg)
+int idpf_send_mb_msg_kfree(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size)
{
- int err = 0;
-
- if (ctlq_msg->cookie.mbx.chnl_opcode != xn->vc_op) {
- dev_err_ratelimited(&adapter->pdev->dev, "Async message opcode does not match transaction opcode (msg: %d) (xn: %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode, xn->vc_op);
- xn->reply_sz = 0;
- err = -EINVAL;
- goto release_bufs;
- }
-
- if (xn->async_handler) {
- err = xn->async_handler(adapter, xn, ctlq_msg);
- goto release_bufs;
- }
-
- if (ctlq_msg->cookie.mbx.chnl_retval) {
- xn->reply_sz = 0;
- dev_err_ratelimited(&adapter->pdev->dev, "Async message failure (op %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode);
- err = -EINVAL;
- }
-
-release_bufs:
- idpf_vc_xn_push_free(adapter->vcxn_mngr, xn);
-
- return err;
-}
-
-/**
- * idpf_vc_xn_forward_reply - copy a reply back to receiving thread
- * @adapter: driver specific private structure with vcxn_mngr
- * @ctlq_msg: controlq message to send back to receiving thread
- */
-static int
-idpf_vc_xn_forward_reply(struct idpf_adapter *adapter,
- const struct idpf_ctlq_msg *ctlq_msg)
-{
- const void *payload = NULL;
- size_t payload_size = 0;
- struct idpf_vc_xn *xn;
- u16 msg_info;
- int err = 0;
- u16 xn_idx;
- u16 salt;
-
- msg_info = ctlq_msg->ctx.sw_cookie.data;
- xn_idx = FIELD_GET(IDPF_VC_XN_IDX_M, msg_info);
- if (xn_idx >= ARRAY_SIZE(adapter->vcxn_mngr->ring)) {
- dev_err_ratelimited(&adapter->pdev->dev, "Out of bounds cookie received: %02x\n",
- xn_idx);
- return -EINVAL;
- }
- xn = &adapter->vcxn_mngr->ring[xn_idx];
- idpf_vc_xn_lock(xn);
- salt = FIELD_GET(IDPF_VC_XN_SALT_M, msg_info);
- if (xn->salt != salt) {
- dev_err_ratelimited(&adapter->pdev->dev, "Transaction salt does not match (exp:%d@%02x(%d) != got:%d@%02x)\n",
- xn->vc_op, xn->salt, xn->state,
- ctlq_msg->cookie.mbx.chnl_opcode, salt);
- idpf_vc_xn_unlock(xn);
- return -EINVAL;
- }
-
- switch (xn->state) {
- case IDPF_VC_XN_WAITING:
- /* success */
- break;
- case IDPF_VC_XN_IDLE:
- dev_err_ratelimited(&adapter->pdev->dev, "Unexpected or belated VC reply (op %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode);
- err = -EINVAL;
- goto out_unlock;
- case IDPF_VC_XN_SHUTDOWN:
- /* ENXIO is a bit special here as the recv msg loop uses that
- * know if it should stop trying to clean the ring if we lost
- * the virtchnl. We need to stop playing with registers and
- * yield.
- */
- err = -ENXIO;
- goto out_unlock;
- case IDPF_VC_XN_ASYNC:
- /* Set reply_sz from the actual payload so that async_handler
- * can evaluate the response.
- */
- xn->reply_sz = ctlq_msg->data_len;
- err = idpf_vc_xn_forward_async(adapter, xn, ctlq_msg);
- idpf_vc_xn_unlock(xn);
- return err;
- default:
- dev_err_ratelimited(&adapter->pdev->dev, "Overwriting VC reply (op %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode);
- err = -EBUSY;
- goto out_unlock;
- }
-
- if (ctlq_msg->cookie.mbx.chnl_opcode != xn->vc_op) {
- dev_err_ratelimited(&adapter->pdev->dev, "Message opcode does not match transaction opcode (msg: %d) (xn: %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode, xn->vc_op);
- xn->reply_sz = 0;
- xn->state = IDPF_VC_XN_COMPLETED_FAILED;
- err = -EINVAL;
- goto out_unlock;
- }
-
- if (ctlq_msg->cookie.mbx.chnl_retval) {
- xn->reply_sz = 0;
- xn->state = IDPF_VC_XN_COMPLETED_FAILED;
- err = -EINVAL;
- goto out_unlock;
- }
-
- if (ctlq_msg->data_len) {
- payload = ctlq_msg->ctx.indirect.payload->va;
- payload_size = ctlq_msg->data_len;
- }
-
- xn->reply_sz = payload_size;
- xn->state = IDPF_VC_XN_COMPLETED_SUCCESS;
+ int err = idpf_send_mb_msg(adapter, xn_params, send_buf, send_buf_size);
- if (xn->reply.iov_base && xn->reply.iov_len && payload_size)
- memcpy(xn->reply.iov_base, payload,
- min_t(size_t, xn->reply.iov_len, payload_size));
-
-out_unlock:
- idpf_vc_xn_unlock(xn);
- /* we _cannot_ hold lock while calling complete */
- complete(&xn->completed);
+ if (libie_cp_can_send_onstack(send_buf_size))
+ kfree(send_buf);
return err;
}
/**
- * idpf_recv_mb_msg - Receive message over mailbox
+ * idpf_send_vf_reset_msg - send one way VF reset message
* @adapter: driver specific private structure
- * @arq: control queue to receive message from
- *
- * Will receive control queue message and posts the receive buffer.
- *
- * Return: 0 on success and negative on failure.
*/
-int idpf_recv_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *arq)
+void idpf_send_vf_reset_msg(struct idpf_adapter *adapter)
{
- struct idpf_ctlq_msg ctlq_msg;
- struct idpf_dma_mem *dma_mem;
- int post_err, err;
- u16 num_recv;
-
- while (1) {
- /* This will get <= num_recv messages and output how many
- * actually received on num_recv.
- */
- num_recv = 1;
- err = idpf_ctlq_recv(arq, &num_recv, &ctlq_msg);
- if (err || !num_recv)
- break;
-
- if (ctlq_msg.data_len) {
- dma_mem = ctlq_msg.ctx.indirect.payload;
- } else {
- dma_mem = NULL;
- num_recv = 0;
- }
-
- if (ctlq_msg.cookie.mbx.chnl_opcode == VIRTCHNL2_OP_EVENT)
- idpf_recv_event_msg(adapter, &ctlq_msg);
- else
- err = idpf_vc_xn_forward_reply(adapter, &ctlq_msg);
+ struct libie_ctlq_info *ctlq = adapter->asq;
- post_err = idpf_ctlq_post_rx_buffs(&adapter->hw, arq,
- &num_recv, &dma_mem);
+ /* Forcefully claim send queue slot */
+ idpf_mb_clean(adapter, ctlq, true);
- /* If post failed clear the only buffer we supplied */
- if (post_err) {
- if (dma_mem)
- dma_free_coherent(&adapter->pdev->dev,
- dma_mem->size, dma_mem->va,
- dma_mem->pa);
- break;
- }
+ scoped_guard(spinlock, &ctlq->lock) {
+ *ctlq->tx_msg[ctlq->next_to_use] = (struct libie_ctlq_msg) {
+ .opcode = LIBIE_CTLQ_SEND_MSG_TO_CP,
+ .chnl_opcode = VIRTCHNL2_OP_RESET_VF,
+ };
- /* virtchnl trying to shutdown, stop cleaning */
- if (err == -ENXIO)
- break;
+ libie_ctlq_send(adapter->asq, 1);
}
-
- return err;
}
struct idpf_chunked_msg_params {
@@ -768,45 +315,43 @@ struct idpf_queue_set *idpf_alloc_queue_set(struct idpf_adapter *adapter,
static int idpf_send_chunked_msg(struct idpf_adapter *adapter,
const struct idpf_chunked_msg_params *params)
{
- struct idpf_vc_xn_params xn_params = {
- .vc_op = params->vc_op,
+ struct libie_ctlq_xn_send_params xn_params = {
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = params->vc_op,
};
const void *pos = params->chunks;
- u32 num_chunks, num_msgs, buf_sz;
- void *buf __free(kfree) = NULL;
u32 totqs = params->num_chunks;
u32 vid = params->vport_id;
+ u32 num_chunks, num_msgs;
- num_chunks = min(IDPF_NUM_CHUNKS_PER_MSG(params->config_sz,
- params->chunk_sz), totqs);
+ num_chunks = IDPF_NUM_CHUNKS_PER_MSG(params->config_sz,
+ params->chunk_sz);
num_msgs = DIV_ROUND_UP(totqs, num_chunks);
- buf_sz = params->config_sz + num_chunks * params->chunk_sz;
- buf = kzalloc(buf_sz, GFP_KERNEL);
- if (!buf)
- return -ENOMEM;
-
- xn_params.send_buf.iov_base = buf;
-
for (u32 i = 0; i < num_msgs; i++) {
- ssize_t reply_sz;
+ u32 buf_sz;
+ void *buf;
+ int err;
- memset(buf, 0, buf_sz);
- xn_params.send_buf.iov_len = buf_sz;
+ num_chunks = min(num_chunks, totqs);
+ buf_sz = params->config_sz + num_chunks * params->chunk_sz;
+ buf = kzalloc(buf_sz, GFP_KERNEL);
+ if (!buf)
+ return -ENOMEM;
- if (params->prepare_msg(vid, buf, pos, num_chunks) != buf_sz)
+ if (params->prepare_msg(vid, buf, pos, num_chunks) != buf_sz) {
+ kfree(buf);
return -EINVAL;
+ }
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, buf, buf_sz);
+ if (err)
+ return err;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ xn_params.recv_mem = (struct kvec) {};
pos += num_chunks * params->chunk_sz;
totqs -= num_chunks;
-
- num_chunks = min(num_chunks, totqs);
- buf_sz = params->config_sz + num_chunks * params->chunk_sz;
}
return 0;
@@ -881,11 +426,14 @@ static int idpf_wait_for_marker_event(struct idpf_vport *vport)
*/
static int idpf_send_ver_msg(struct idpf_adapter *adapter)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_VERSION,
+ };
+ struct virtchnl2_version_info *vvi_recv;
struct virtchnl2_version_info vvi;
- ssize_t reply_sz;
u32 major, minor;
- int err = 0;
+ int err;
if (adapter->virt_ver_maj) {
vvi.major = cpu_to_le32(adapter->virt_ver_maj);
@@ -895,24 +443,23 @@ static int idpf_send_ver_msg(struct idpf_adapter *adapter)
vvi.minor = cpu_to_le32(IDPF_VIRTCHNL_VERSION_MINOR);
}
- xn_params.vc_op = VIRTCHNL2_OP_VERSION;
- xn_params.send_buf.iov_base = &vvi;
- xn_params.send_buf.iov_len = sizeof(vvi);
- xn_params.recv_buf = xn_params.send_buf;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ err = idpf_send_mb_msg(adapter, &xn_params, &vvi, sizeof(vvi));
+ if (err)
+ return err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz < sizeof(vvi))
- return -EIO;
+ if (xn_params.recv_mem.iov_len < sizeof(*vvi_recv)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
- major = le32_to_cpu(vvi.major);
- minor = le32_to_cpu(vvi.minor);
+ vvi_recv = xn_params.recv_mem.iov_base;
+ major = le32_to_cpu(vvi_recv->major);
+ minor = le32_to_cpu(vvi_recv->minor);
if (major > IDPF_VIRTCHNL_VERSION_MAJOR) {
dev_warn(&adapter->pdev->dev, "Virtchnl major version greater than supported\n");
- return -EINVAL;
+ err = -EINVAL;
+ goto free_rx_buf;
}
if (major == IDPF_VIRTCHNL_VERSION_MAJOR &&
@@ -930,6 +477,9 @@ static int idpf_send_ver_msg(struct idpf_adapter *adapter)
adapter->virt_ver_maj = major;
adapter->virt_ver_min = minor;
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
return err;
}
@@ -942,9 +492,12 @@ static int idpf_send_ver_msg(struct idpf_adapter *adapter)
*/
static int idpf_send_get_caps_msg(struct idpf_adapter *adapter)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_GET_CAPS,
+ };
struct virtchnl2_get_capabilities caps = {};
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ int err;
caps.csum_caps =
cpu_to_le32(VIRTCHNL2_CAP_TX_CSUM_L3_IPV4 |
@@ -1004,20 +557,22 @@ static int idpf_send_get_caps_msg(struct idpf_adapter *adapter)
VIRTCHNL2_CAP_LOOPBACK |
VIRTCHNL2_CAP_PTP);
- xn_params.vc_op = VIRTCHNL2_OP_GET_CAPS;
- xn_params.send_buf.iov_base = ∩︀
- xn_params.send_buf.iov_len = sizeof(caps);
- xn_params.recv_buf.iov_base = &adapter->caps;
- xn_params.recv_buf.iov_len = sizeof(adapter->caps);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ err = idpf_send_mb_msg(adapter, &xn_params, &caps, sizeof(caps));
+ if (err)
+ return err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz < sizeof(adapter->caps))
- return -EIO;
+ if (xn_params.recv_mem.iov_len < sizeof(adapter->caps)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
- return 0;
+ memcpy(&adapter->caps, xn_params.recv_mem.iov_base,
+ sizeof(adapter->caps));
+
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return err;
}
/**
@@ -1062,37 +617,39 @@ static void idpf_decfg_lan_memory_regions(struct idpf_adapter *adapter)
*/
static int idpf_cfg_lan_memory_regions(struct idpf_adapter *adapter)
{
- struct virtchnl2_get_lan_memory_regions *rcvd_regions __free(kfree);
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_GET_LAN_MEMORY_REGIONS,
- .recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN,
- .send_buf.iov_len =
- sizeof(struct virtchnl2_get_lan_memory_regions) +
- sizeof(struct virtchnl2_mem_region),
+ struct virtchnl2_get_lan_memory_regions *send_regions, *rcvd_regions;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_GET_LAN_MEMORY_REGIONS,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int num_regions, size;
- ssize_t reply_sz;
+ size_t send_sz, reply_sz, size;
+ int num_regions;
int err = 0;
- rcvd_regions = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!rcvd_regions)
+ send_sz = sizeof(struct virtchnl2_get_lan_memory_regions) +
+ sizeof(struct virtchnl2_mem_region);
+ send_regions = kzalloc(send_sz, GFP_KERNEL);
+ if (!send_regions)
return -ENOMEM;
- xn_params.recv_buf.iov_base = rcvd_regions;
- rcvd_regions->num_memory_regions = cpu_to_le16(1);
- xn_params.send_buf.iov_base = rcvd_regions;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ send_regions->num_memory_regions = cpu_to_le16(1);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, send_regions,
+ send_sz);
+ if (err)
+ return err;
+ rcvd_regions = xn_params.recv_mem.iov_base;
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz < sizeof(*rcvd_regions)) {
+ err = -EIO;
+ goto rel_rx_buf;
+ }
num_regions = le16_to_cpu(rcvd_regions->num_memory_regions);
size = struct_size(rcvd_regions, mem_reg, num_regions);
- if (reply_sz < size)
- return -EIO;
-
- if (size > IDPF_CTLQ_MAX_BUF_LEN)
- return -EINVAL;
+ if (reply_sz < size) {
+ err = -EIO;
+ goto rel_rx_buf;
+ }
for (int i = 0; i < num_regions; i++) {
struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
@@ -1102,10 +659,14 @@ static int idpf_cfg_lan_memory_regions(struct idpf_adapter *adapter)
len = le64_to_cpu(rcvd_regions->mem_reg[i].size);
if (len && !libie_pci_map_mmio_region(mmio, offset, len)) {
idpf_decfg_lan_memory_regions(adapter);
- return -EIO;
+ err = -EIO;
+ goto rel_rx_buf;
}
}
+rel_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
return err;
}
@@ -1164,24 +725,43 @@ int idpf_add_del_fsteer_filters(struct idpf_adapter *adapter,
struct virtchnl2_flow_rule_add_del *rule,
enum virtchnl2_op opcode)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = opcode,
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ };
+ struct virtchnl2_flow_rule_add_del *rx_rule;
int rule_count = le32_to_cpu(rule->count);
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ size_t send_sz;
+ int err;
if (opcode != VIRTCHNL2_OP_ADD_FLOW_RULE &&
- opcode != VIRTCHNL2_OP_DEL_FLOW_RULE)
+ opcode != VIRTCHNL2_OP_DEL_FLOW_RULE) {
+ kfree(rule);
return -EINVAL;
+ }
+
+ send_sz = struct_size(rule, rule_info, rule_count);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, rule, send_sz);
+ if (err)
+ return err;
+
+ if (xn_params.recv_mem.iov_len < send_sz) {
+ err = -EIO;
+ goto rel_rx;
+ }
- xn_params.vc_op = opcode;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.async = false;
- xn_params.send_buf.iov_base = rule;
- xn_params.send_buf.iov_len = struct_size(rule, rule_info, rule_count);
- xn_params.recv_buf.iov_base = rule;
- xn_params.recv_buf.iov_len = struct_size(rule, rule_info, rule_count);
+ rx_rule = xn_params.recv_mem.iov_base;
+ for (int i = 0; i < rule_count; i++) {
+ if (rx_rule->rule_info[i].status !=
+ cpu_to_le32(VIRTCHNL2_FLOW_RULE_SUCCESS)) {
+ err = -EIO;
+ goto rel_rx;
+ }
+ }
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- return reply_sz < 0 ? reply_sz : 0;
+rel_rx:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -1555,11 +1135,13 @@ int idpf_queue_reg_init(struct idpf_vport *vport,
int idpf_send_create_vport_msg(struct idpf_adapter *adapter,
struct idpf_vport_max_q *max_q)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_CREATE_VPORT,
+ };
struct virtchnl2_create_vport *vport_msg;
- struct idpf_vc_xn_params xn_params = {};
u16 idx = adapter->next_vport;
int err, buf_size;
- ssize_t reply_sz;
buf_size = sizeof(struct virtchnl2_create_vport);
vport_msg = kzalloc(buf_size, GFP_KERNEL);
@@ -1586,33 +1168,29 @@ int idpf_send_create_vport_msg(struct idpf_adapter *adapter,
}
if (!adapter->vport_params_recvd[idx]) {
- adapter->vport_params_recvd[idx] = kzalloc(IDPF_CTLQ_MAX_BUF_LEN,
- GFP_KERNEL);
+ adapter->vport_params_recvd[idx] =
+ kzalloc(LIBIE_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
if (!adapter->vport_params_recvd[idx]) {
err = -ENOMEM;
goto rel_buf;
}
}
- xn_params.vc_op = VIRTCHNL2_OP_CREATE_VPORT;
- xn_params.send_buf.iov_base = vport_msg;
- xn_params.send_buf.iov_len = buf_size;
- xn_params.recv_buf.iov_base = adapter->vport_params_recvd[idx];
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0) {
- err = reply_sz;
- goto free_vport_params;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, vport_msg,
+ sizeof(*vport_msg));
+ if (err) {
+ kfree(adapter->vport_params_recvd[idx]);
+ adapter->vport_params_recvd[idx] = NULL;
+ return err;
}
- kfree(vport_msg);
+ memcpy(adapter->vport_params_recvd[idx], xn_params.recv_mem.iov_base,
+ xn_params.recv_mem.iov_len);
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
return 0;
-free_vport_params:
- kfree(adapter->vport_params_recvd[idx]);
- adapter->vport_params_recvd[idx] = NULL;
rel_buf:
kfree(vport_msg);
@@ -1674,19 +1252,22 @@ int idpf_check_supported_desc_ids(struct idpf_vport *vport)
*/
int idpf_send_destroy_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DESTROY_VPORT,
+ };
struct virtchnl2_vport v_id;
- ssize_t reply_sz;
+ int err;
v_id.vport_id = cpu_to_le32(vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_DESTROY_VPORT;
- xn_params.send_buf.iov_base = &v_id;
- xn_params.send_buf.iov_len = sizeof(v_id);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg(adapter, &xn_params, &v_id, sizeof(v_id));
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return reply_sz < 0 ? reply_sz : 0;
+ return 0;
}
/**
@@ -1698,19 +1279,22 @@ int idpf_send_destroy_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
*/
int idpf_send_enable_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_ENABLE_VPORT,
+ };
struct virtchnl2_vport v_id;
- ssize_t reply_sz;
+ int err;
v_id.vport_id = cpu_to_le32(vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_ENABLE_VPORT;
- xn_params.send_buf.iov_base = &v_id;
- xn_params.send_buf.iov_len = sizeof(v_id);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg(adapter, &xn_params, &v_id, sizeof(v_id));
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return reply_sz < 0 ? reply_sz : 0;
+ return 0;
}
/**
@@ -1722,19 +1306,22 @@ int idpf_send_enable_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
*/
int idpf_send_disable_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DISABLE_VPORT,
+ };
struct virtchnl2_vport v_id;
- ssize_t reply_sz;
+ int err;
v_id.vport_id = cpu_to_le32(vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_DISABLE_VPORT;
- xn_params.send_buf.iov_base = &v_id;
- xn_params.send_buf.iov_len = sizeof(v_id);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg(adapter, &xn_params, &v_id, sizeof(v_id));
+ if (err)
+ return err;
- return reply_sz < 0 ? reply_sz : 0;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return 0;
}
/**
@@ -2573,11 +2160,14 @@ int idpf_send_delete_queues_msg(struct idpf_adapter *adapter,
struct idpf_queue_id_reg_info *chunks,
u32 vport_id)
{
- struct virtchnl2_del_ena_dis_queues *eq __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DEL_QUEUES,
+ };
+ struct virtchnl2_del_ena_dis_queues *eq;
+ ssize_t buf_size;
u16 num_chunks;
- int buf_size;
+ int err;
num_chunks = chunks->num_chunks;
buf_size = struct_size(eq, chunks.chunks, num_chunks);
@@ -2592,13 +2182,13 @@ int idpf_send_delete_queues_msg(struct idpf_adapter *adapter,
idpf_convert_reg_to_queue_chunks(eq->chunks.chunks, chunks->queue_chunks,
num_chunks);
- xn_params.vc_op = VIRTCHNL2_OP_DEL_QUEUES;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = eq;
- xn_params.send_buf.iov_len = buf_size;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, eq, buf_size);
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return reply_sz < 0 ? reply_sz : 0;
+ return 0;
}
/**
@@ -2636,15 +2226,14 @@ int idpf_send_add_queues_msg(struct idpf_adapter *adapter,
struct idpf_q_vec_rsrc *rsrc,
u32 vport_id)
{
- struct virtchnl2_add_queues *vc_msg __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_ADD_QUEUES,
+ };
+ struct virtchnl2_add_queues *vc_msg;
struct virtchnl2_add_queues aq = {};
- ssize_t reply_sz;
- int size;
-
- vc_msg = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!vc_msg)
- return -ENOMEM;
+ size_t size;
+ int err;
aq.vport_id = cpu_to_le32(vport_id);
aq.num_tx_q = cpu_to_le16(rsrc->num_txq);
@@ -2652,29 +2241,38 @@ int idpf_send_add_queues_msg(struct idpf_adapter *adapter,
aq.num_rx_q = cpu_to_le16(rsrc->num_rxq);
aq.num_rx_bufq = cpu_to_le16(rsrc->num_bufq);
- xn_params.vc_op = VIRTCHNL2_OP_ADD_QUEUES;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &aq;
- xn_params.send_buf.iov_len = sizeof(aq);
- xn_params.recv_buf.iov_base = vc_msg;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg(adapter, &xn_params, &aq, sizeof(aq));
+ if (err)
+ return err;
+
+ vc_msg = xn_params.recv_mem.iov_base;
+ if (xn_params.recv_mem.iov_len < sizeof(*vc_msg)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
/* compare vc_msg num queues with vport num queues */
if (le16_to_cpu(vc_msg->num_tx_q) != rsrc->num_txq ||
le16_to_cpu(vc_msg->num_rx_q) != rsrc->num_rxq ||
le16_to_cpu(vc_msg->num_tx_complq) != rsrc->num_complq ||
- le16_to_cpu(vc_msg->num_rx_bufq) != rsrc->num_bufq)
- return -EINVAL;
+ le16_to_cpu(vc_msg->num_rx_bufq) != rsrc->num_bufq) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
size = struct_size(vc_msg, chunks.chunks,
le16_to_cpu(vc_msg->chunks.num_chunks));
- if (reply_sz < size)
- return -EIO;
+ if (xn_params.recv_mem.iov_len < size) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
+
+ err = idpf_vport_init_queue_reg_chunks(vport_config, &vc_msg->chunks);
+
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return idpf_vport_init_queue_reg_chunks(vport_config, &vc_msg->chunks);
+ return err;
}
/**
@@ -2686,49 +2284,51 @@ int idpf_send_add_queues_msg(struct idpf_adapter *adapter,
*/
int idpf_send_alloc_vectors_msg(struct idpf_adapter *adapter, u16 num_vectors)
{
- struct virtchnl2_alloc_vectors *rcvd_vec __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_ALLOC_VECTORS,
+ };
+ struct virtchnl2_alloc_vectors *rcvd_vec;
struct virtchnl2_alloc_vectors ac = {};
- ssize_t reply_sz;
u16 num_vchunks;
- int size;
+ int size, err;
ac.num_vectors = cpu_to_le16(num_vectors);
- rcvd_vec = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!rcvd_vec)
- return -ENOMEM;
+ err = idpf_send_mb_msg(adapter, &xn_params, &ac, sizeof(ac));
+ if (err)
+ return err;
- xn_params.vc_op = VIRTCHNL2_OP_ALLOC_VECTORS;
- xn_params.send_buf.iov_base = ∾
- xn_params.send_buf.iov_len = sizeof(ac);
- xn_params.recv_buf.iov_base = rcvd_vec;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ rcvd_vec = xn_params.recv_mem.iov_base;
+ if (xn_params.recv_mem.iov_len < sizeof(*rcvd_vec)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
num_vchunks = le16_to_cpu(rcvd_vec->vchunks.num_vchunks);
size = struct_size(rcvd_vec, vchunks.vchunks, num_vchunks);
- if (reply_sz < size)
- return -EIO;
-
- if (size > IDPF_CTLQ_MAX_BUF_LEN)
- return -EINVAL;
+ if (xn_params.recv_mem.iov_len < size) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
kfree(adapter->req_vec_chunks);
adapter->req_vec_chunks = kmemdup(rcvd_vec, size, GFP_KERNEL);
- if (!adapter->req_vec_chunks)
- return -ENOMEM;
+ if (!adapter->req_vec_chunks) {
+ err = -ENOMEM;
+ goto free_rx_buf;
+ }
if (le16_to_cpu(adapter->req_vec_chunks->num_vectors) < num_vectors) {
kfree(adapter->req_vec_chunks);
adapter->req_vec_chunks = NULL;
- return -EINVAL;
+ err = -EINVAL;
}
- return 0;
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return err;
}
/**
@@ -2740,24 +2340,28 @@ int idpf_send_alloc_vectors_msg(struct idpf_adapter *adapter, u16 num_vectors)
int idpf_send_dealloc_vectors_msg(struct idpf_adapter *adapter)
{
struct virtchnl2_alloc_vectors *ac = adapter->req_vec_chunks;
- struct virtchnl2_vector_chunks *vcs = &ac->vchunks;
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
- int buf_size;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DEALLOC_VECTORS,
+ };
+ struct virtchnl2_vector_chunks *vcs;
+ int buf_size, err;
- buf_size = struct_size(vcs, vchunks, le16_to_cpu(vcs->num_vchunks));
+ buf_size = struct_size(&ac->vchunks, vchunks,
+ le16_to_cpu(ac->vchunks.num_vchunks));
+ vcs = kmemdup(&ac->vchunks, buf_size, GFP_KERNEL);
+ if (!vcs)
+ return -ENOMEM;
- xn_params.vc_op = VIRTCHNL2_OP_DEALLOC_VECTORS;
- xn_params.send_buf.iov_base = vcs;
- xn_params.send_buf.iov_len = buf_size;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, vcs, buf_size);
+ if (err)
+ return err;
kfree(adapter->req_vec_chunks);
adapter->req_vec_chunks = NULL;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
return 0;
}
@@ -2781,18 +2385,22 @@ static int idpf_get_max_vfs(struct idpf_adapter *adapter)
*/
int idpf_send_set_sriov_vfs_msg(struct idpf_adapter *adapter, u16 num_vfs)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_SET_SRIOV_VFS,
+ };
struct virtchnl2_sriov_vfs_info svi = {};
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ int err;
svi.num_vfs = cpu_to_le16(num_vfs);
- xn_params.vc_op = VIRTCHNL2_OP_SET_SRIOV_VFS;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &svi;
- xn_params.send_buf.iov_len = sizeof(svi);
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- return reply_sz < 0 ? reply_sz : 0;
+ err = idpf_send_mb_msg(adapter, &xn_params, &svi, sizeof(svi));
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return 0;
}
/**
@@ -2805,10 +2413,14 @@ int idpf_send_set_sriov_vfs_msg(struct idpf_adapter *adapter, u16 num_vfs)
int idpf_send_get_stats_msg(struct idpf_netdev_priv *np,
struct idpf_port_stats *port_stats)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_GET_STATS,
+ };
struct rtnl_link_stats64 *netstats = &np->netstats;
+ struct virtchnl2_vport_stats *stats_recv;
struct virtchnl2_vport_stats stats_msg = {};
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ int err;
/* Don't send get_stats message if the link is down */
@@ -2817,38 +2429,41 @@ int idpf_send_get_stats_msg(struct idpf_netdev_priv *np,
stats_msg.vport_id = cpu_to_le32(np->vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_GET_STATS;
- xn_params.send_buf.iov_base = &stats_msg;
- xn_params.send_buf.iov_len = sizeof(stats_msg);
- xn_params.recv_buf = xn_params.send_buf;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ err = idpf_send_mb_msg(np->adapter, &xn_params, &stats_msg,
+ sizeof(stats_msg));
+ if (err)
+ return err;
- reply_sz = idpf_vc_xn_exec(np->adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz < sizeof(stats_msg))
- return -EIO;
+ if (xn_params.recv_mem.iov_len < sizeof(*stats_recv)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
+
+ stats_recv = xn_params.recv_mem.iov_base;
spin_lock_bh(&np->stats_lock);
- netstats->rx_packets = le64_to_cpu(stats_msg.rx_unicast) +
- le64_to_cpu(stats_msg.rx_multicast) +
- le64_to_cpu(stats_msg.rx_broadcast);
- netstats->tx_packets = le64_to_cpu(stats_msg.tx_unicast) +
- le64_to_cpu(stats_msg.tx_multicast) +
- le64_to_cpu(stats_msg.tx_broadcast);
- netstats->rx_bytes = le64_to_cpu(stats_msg.rx_bytes);
- netstats->tx_bytes = le64_to_cpu(stats_msg.tx_bytes);
- netstats->rx_errors = le64_to_cpu(stats_msg.rx_errors);
- netstats->tx_errors = le64_to_cpu(stats_msg.tx_errors);
- netstats->rx_dropped = le64_to_cpu(stats_msg.rx_discards);
- netstats->tx_dropped = le64_to_cpu(stats_msg.tx_discards);
-
- port_stats->vport_stats = stats_msg;
+ netstats->rx_packets = le64_to_cpu(stats_recv->rx_unicast) +
+ le64_to_cpu(stats_recv->rx_multicast) +
+ le64_to_cpu(stats_recv->rx_broadcast);
+ netstats->tx_packets = le64_to_cpu(stats_recv->tx_unicast) +
+ le64_to_cpu(stats_recv->tx_multicast) +
+ le64_to_cpu(stats_recv->tx_broadcast);
+ netstats->rx_bytes = le64_to_cpu(stats_recv->rx_bytes);
+ netstats->tx_bytes = le64_to_cpu(stats_recv->tx_bytes);
+ netstats->rx_errors = le64_to_cpu(stats_recv->rx_errors);
+ netstats->tx_errors = le64_to_cpu(stats_recv->tx_errors);
+ netstats->rx_dropped = le64_to_cpu(stats_recv->rx_discards);
+ netstats->tx_dropped = le64_to_cpu(stats_recv->tx_discards);
+
+ port_stats->vport_stats = *stats_recv;
spin_unlock_bh(&np->stats_lock);
- return 0;
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return err;
}
/**
@@ -2866,13 +2481,14 @@ int idpf_send_get_stats_msg(struct idpf_netdev_priv *np,
int idpf_send_set_rss_lut_msg(struct idpf_adapter *adapter,
struct idpf_rss_data *rss_data, u32 vport_id)
{
- struct virtchnl2_rss_lut *rl __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_SET_RSS_LUT,
+ };
+ struct virtchnl2_rss_lut *rl;
struct idpf_vport *vport;
- ssize_t reply_sz;
+ int buf_size, i, err;
bool rxhash_ena;
- int buf_size;
- int i;
vport = idpf_vid_to_vport(adapter, vport_id);
if (!vport)
@@ -2886,21 +2502,17 @@ int idpf_send_set_rss_lut_msg(struct idpf_adapter *adapter,
return -ENOMEM;
rl->vport_id = cpu_to_le32(vport_id);
-
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = rl;
- xn_params.send_buf.iov_len = buf_size;
- xn_params.vc_op = VIRTCHNL2_OP_SET_RSS_LUT;
-
rl->lut_entries = cpu_to_le16(rss_data->rss_lut_size);
for (i = 0; i < rss_data->rss_lut_size; i++)
rl->lut[i] = rxhash_ena ? cpu_to_le32(rss_data->rss_lut[i]) : 0;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, rl, buf_size);
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -2914,10 +2526,12 @@ int idpf_send_set_rss_lut_msg(struct idpf_adapter *adapter,
int idpf_send_set_rss_key_msg(struct idpf_adapter *adapter,
struct idpf_rss_data *rss_data, u32 vport_id)
{
- struct virtchnl2_rss_key *rk __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
- int i, buf_size;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_SET_RSS_KEY,
+ };
+ struct virtchnl2_rss_key *rk;
+ int i, buf_size, err;
buf_size = struct_size(rk, key_flex, rss_data->rss_key_size);
rk = kzalloc(buf_size, GFP_KERNEL);
@@ -2925,20 +2539,17 @@ int idpf_send_set_rss_key_msg(struct idpf_adapter *adapter,
return -ENOMEM;
rk->vport_id = cpu_to_le32(vport_id);
- xn_params.send_buf.iov_base = rk;
- xn_params.send_buf.iov_len = buf_size;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.vc_op = VIRTCHNL2_OP_SET_RSS_KEY;
-
rk->key_len = cpu_to_le16(rss_data->rss_key_size);
for (i = 0; i < rss_data->rss_key_size; i++)
rk->key_flex[i] = rss_data->rss_key[i];
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, rk, buf_size);
+ if (err)
+ return err;
- return 0;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return err;
}
/**
@@ -3115,15 +2726,18 @@ static void idpf_parse_protocol_ids(struct virtchnl2_ptype *ptype,
*/
static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
{
- struct virtchnl2_get_ptype_info *get_ptype_info __free(kfree) = NULL;
- struct virtchnl2_get_ptype_info *ptype_info __free(kfree) = NULL;
struct libeth_rx_pt *singleq_pt_lkup __free(kfree) = NULL;
struct libeth_rx_pt *splitq_pt_lkup __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_GET_PTYPE_INFO,
+ };
+ struct virtchnl2_get_ptype_info *get_ptype_info;
+ struct virtchnl2_get_ptype_info *ptype_info;
+ int err = 0, max_ptype = IDPF_RX_MAX_PTYPE;
+ int buf_size = sizeof(*get_ptype_info);
int ptypes_recvd = 0, ptype_offset;
- u32 max_ptype = IDPF_RX_MAX_PTYPE;
u16 next_ptype_id = 0;
- ssize_t reply_sz;
singleq_pt_lkup = kzalloc_objs(*singleq_pt_lkup, IDPF_RX_MAX_BASE_PTYPE);
if (!singleq_pt_lkup)
@@ -3133,42 +2747,38 @@ static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
if (!splitq_pt_lkup)
return -ENOMEM;
- get_ptype_info = kzalloc_obj(*get_ptype_info);
- if (!get_ptype_info)
- return -ENOMEM;
-
- ptype_info = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!ptype_info)
- return -ENOMEM;
+ while (next_ptype_id < max_ptype) {
+ u16 num_ptypes;
- xn_params.vc_op = VIRTCHNL2_OP_GET_PTYPE_INFO;
- xn_params.send_buf.iov_base = get_ptype_info;
- xn_params.send_buf.iov_len = sizeof(*get_ptype_info);
- xn_params.recv_buf.iov_base = ptype_info;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ get_ptype_info = kzalloc(buf_size, GFP_KERNEL);
+ if (!get_ptype_info)
+ return -ENOMEM;
- while (next_ptype_id < max_ptype) {
get_ptype_info->start_ptype_id = cpu_to_le16(next_ptype_id);
if ((next_ptype_id + IDPF_RX_MAX_PTYPES_PER_BUF) > max_ptype)
- get_ptype_info->num_ptypes =
- cpu_to_le16(max_ptype - next_ptype_id);
+ num_ptypes = max_ptype - next_ptype_id;
else
- get_ptype_info->num_ptypes =
- cpu_to_le16(IDPF_RX_MAX_PTYPES_PER_BUF);
+ num_ptypes = IDPF_RX_MAX_PTYPES_PER_BUF;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ get_ptype_info->num_ptypes = cpu_to_le16(num_ptypes);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params,
+ get_ptype_info, buf_size);
+ if (err)
+ return err;
+ ptype_info = xn_params.recv_mem.iov_base;
+ if (xn_params.recv_mem.iov_len < sizeof(*ptype_info)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
ptypes_recvd += le16_to_cpu(ptype_info->num_ptypes);
- if (ptypes_recvd > max_ptype)
- return -EINVAL;
-
- next_ptype_id = le16_to_cpu(get_ptype_info->start_ptype_id) +
- le16_to_cpu(get_ptype_info->num_ptypes);
+ if (ptypes_recvd > max_ptype) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
+ next_ptype_id = next_ptype_id + num_ptypes;
ptype_offset = IDPF_RX_PTYPE_HDR_SZ;
for (u16 i = 0; i < le16_to_cpu(ptype_info->num_ptypes); i++) {
@@ -3178,19 +2788,28 @@ static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
ptype = (struct virtchnl2_ptype *)
((u8 *)ptype_info + ptype_offset);
+ if (xn_params.recv_mem.iov_len <
+ ptype_offset + sizeof(struct virtchnl2_ptype)) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
pt_10 = le16_to_cpu(ptype->ptype_id_10);
pt_8 = ptype->ptype_id_8;
ptype_offset += IDPF_GET_PTYPE_SIZE(ptype);
- if (ptype_offset > IDPF_CTLQ_MAX_BUF_LEN)
- return -EINVAL;
+ if (xn_params.recv_mem.iov_len < ptype_offset) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
/* 0xFFFF indicates end of ptypes */
if (pt_10 == IDPF_INVALID_PTYPE_ID)
goto out;
- if (pt_10 >= max_ptype)
- return -EINVAL;
+ if (pt_10 >= max_ptype) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
idpf_parse_protocol_ids(ptype, &rx_pt);
idpf_finalize_ptype_lookup(&rx_pt);
@@ -3204,13 +2823,18 @@ static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
if (!singleq_pt_lkup[pt_8].outer_ip)
singleq_pt_lkup[pt_8] = rx_pt;
}
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ xn_params.recv_mem = (struct kvec) {};
}
out:
adapter->splitq_pt_lkup = no_free_ptr(splitq_pt_lkup);
adapter->singleq_pt_lkup = no_free_ptr(singleq_pt_lkup);
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -3238,40 +2862,24 @@ static void idpf_rel_rx_pt_lkup(struct idpf_adapter *adapter)
int idpf_send_ena_dis_loopback_msg(struct idpf_adapter *adapter, u32 vport_id,
bool loopback_ena)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_LOOPBACK,
+ };
struct virtchnl2_loopback loopback;
- ssize_t reply_sz;
+ int err;
loopback.vport_id = cpu_to_le32(vport_id);
loopback.enable = loopback_ena;
- xn_params.vc_op = VIRTCHNL2_OP_LOOPBACK;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &loopback;
- xn_params.send_buf.iov_len = sizeof(loopback);
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
-
- return reply_sz < 0 ? reply_sz : 0;
-}
-
-/**
- * idpf_find_ctlq - Given a type and id, find ctlq info
- * @hw: hardware struct
- * @type: type of ctrlq to find
- * @id: ctlq id to find
- *
- * Returns pointer to found ctlq info struct, NULL otherwise.
- */
-static struct idpf_ctlq_info *idpf_find_ctlq(struct idpf_hw *hw,
- enum idpf_ctlq_type type, int id)
-{
- struct idpf_ctlq_info *cq, *tmp;
+ err = idpf_send_mb_msg(adapter, &xn_params, &loopback,
+ sizeof(loopback));
+ if (err)
+ return err;
- list_for_each_entry_safe(cq, tmp, &hw->cq_list_head, cq_list)
- if (cq->q_id == id && cq->cq_type == type)
- return cq;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return NULL;
+ return 0;
}
/**
@@ -3282,40 +2890,45 @@ static struct idpf_ctlq_info *idpf_find_ctlq(struct idpf_hw *hw,
*/
int idpf_init_dflt_mbx(struct idpf_adapter *adapter)
{
- struct idpf_ctlq_create_info ctlq_info[] = {
+ struct libie_ctlq_ctx *ctx = &adapter->ctlq_ctx;
+ struct libie_ctlq_create_info ctlq_info[] = {
{
- .type = IDPF_CTLQ_TYPE_MAILBOX_TX,
- .id = IDPF_DFLT_MBX_ID,
+ .type = LIBIE_CTLQ_TYPE_TX,
+ .id = LIBIE_CTLQ_MBX_ID,
.len = IDPF_DFLT_MBX_Q_LEN,
- .buf_size = IDPF_CTLQ_MAX_BUF_LEN
},
{
- .type = IDPF_CTLQ_TYPE_MAILBOX_RX,
- .id = IDPF_DFLT_MBX_ID,
+ .type = LIBIE_CTLQ_TYPE_RX,
+ .id = LIBIE_CTLQ_MBX_ID,
.len = IDPF_DFLT_MBX_Q_LEN,
- .buf_size = IDPF_CTLQ_MAX_BUF_LEN
}
};
- struct idpf_hw *hw = &adapter->hw;
+ struct libie_ctlq_xn_init_params params = {
+ .num_qs = IDPF_NUM_DFLT_MBX_Q,
+ .cctlq_info = ctlq_info,
+ .ctx = ctx,
+ };
int err;
- adapter->dev_ops.reg_ops.ctlq_reg_init(adapter, ctlq_info);
+ adapter->dev_ops.reg_ops.ctlq_reg_init(&ctx->mmio_info,
+ params.cctlq_info);
- err = idpf_ctlq_init(hw, IDPF_NUM_DFLT_MBX_Q, ctlq_info);
+ err = libie_ctlq_xn_init(¶ms);
if (err)
return err;
- hw->asq = idpf_find_ctlq(hw, IDPF_CTLQ_TYPE_MAILBOX_TX,
- IDPF_DFLT_MBX_ID);
- hw->arq = idpf_find_ctlq(hw, IDPF_CTLQ_TYPE_MAILBOX_RX,
- IDPF_DFLT_MBX_ID);
-
- if (!hw->asq || !hw->arq) {
- idpf_ctlq_deinit(hw);
-
+ adapter->asq = libie_find_ctlq(ctx, LIBIE_CTLQ_TYPE_TX,
+ LIBIE_CTLQ_MBX_ID);
+ adapter->arq = libie_find_ctlq(ctx, LIBIE_CTLQ_TYPE_RX,
+ LIBIE_CTLQ_MBX_ID);
+ if (!adapter->asq || !adapter->arq) {
+ adapter->asq = NULL;
+ adapter->arq = NULL;
+ libie_ctlq_xn_deinit(params.xnm, ctx);
return -ENOENT;
}
+ adapter->xnm = params.xnm;
adapter->state = __IDPF_VER_CHECK;
return 0;
@@ -3327,12 +2940,14 @@ int idpf_init_dflt_mbx(struct idpf_adapter *adapter)
*/
void idpf_deinit_dflt_mbx(struct idpf_adapter *adapter)
{
- if (adapter->hw.arq && adapter->hw.asq) {
- idpf_mb_clean(adapter, adapter->hw.asq);
- idpf_ctlq_deinit(&adapter->hw);
+ if (adapter->xnm) {
+ idpf_mb_clean(adapter, adapter->asq, true);
+ libie_ctlq_xn_deinit(adapter->xnm, &adapter->ctlq_ctx);
}
- adapter->hw.arq = NULL;
- adapter->hw.asq = NULL;
+
+ adapter->arq = NULL;
+ adapter->asq = NULL;
+ adapter->xnm = NULL;
}
/**
@@ -3403,15 +3018,6 @@ int idpf_vc_core_init(struct idpf_adapter *adapter)
u16 num_max_vports;
int err = 0;
- if (!adapter->vcxn_mngr) {
- adapter->vcxn_mngr = kzalloc_obj(*adapter->vcxn_mngr);
- if (!adapter->vcxn_mngr) {
- err = -ENOMEM;
- goto init_failed;
- }
- }
- idpf_vc_xn_init(adapter->vcxn_mngr);
-
while (adapter->state != __IDPF_INIT_SW) {
switch (adapter->state) {
case __IDPF_VER_CHECK:
@@ -3558,8 +3164,7 @@ int idpf_vc_core_init(struct idpf_adapter *adapter)
* the mailbox again
*/
adapter->state = __IDPF_VER_CHECK;
- if (adapter->vcxn_mngr)
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ libie_ctlq_xn_shutdown(adapter->xnm);
set_bit(IDPF_HR_DRV_LOAD, adapter->flags);
queue_delayed_work(adapter->vc_event_wq, &adapter->vc_event_task,
msecs_to_jiffies(task_delay));
@@ -3582,7 +3187,7 @@ void idpf_vc_core_deinit(struct idpf_adapter *adapter)
/* Avoid transaction timeouts when called during reset */
remove_in_prog = test_bit(IDPF_REMOVE_IN_PROG, adapter->flags);
if (!remove_in_prog)
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ libie_ctlq_xn_shutdown(adapter->xnm);
idpf_ptp_release(adapter);
idpf_deinit_task(adapter);
@@ -3591,7 +3196,7 @@ void idpf_vc_core_deinit(struct idpf_adapter *adapter)
idpf_intr_rel(adapter);
if (remove_in_prog)
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ libie_ctlq_xn_shutdown(adapter->xnm);
cancel_delayed_work_sync(&adapter->serv_task);
cancel_delayed_work_sync(&adapter->mbx_task);
@@ -4128,9 +3733,9 @@ static void idpf_set_mac_type(const u8 *default_mac_addr,
/**
* idpf_mac_filter_async_handler - Async callback for mac filters
- * @adapter: private data struct
- * @xn: transaction for message
- * @ctlq_msg: received message
+ * @ctx: controlq context structure
+ * @buff: response buffer pointer and size
+ * @status: async call return value
*
* In some scenarios driver can't sleep and wait for a reply (e.g.: stack is
* holding rtnl_lock) when adding a new mac filter. It puts us in a difficult
@@ -4138,13 +3743,14 @@ static void idpf_set_mac_type(const u8 *default_mac_addr,
* ultimately do is remove it from our list of mac filters and report the
* error.
*/
-static int idpf_mac_filter_async_handler(struct idpf_adapter *adapter,
- struct idpf_vc_xn *xn,
- const struct idpf_ctlq_msg *ctlq_msg)
+static void idpf_mac_filter_async_handler(void *ctx,
+ struct kvec *buff,
+ int status)
{
struct virtchnl2_mac_addr_list *ma_list;
struct idpf_vport_config *vport_config;
struct virtchnl2_mac_addr *mac_addr;
+ struct idpf_adapter *adapter = ctx;
struct idpf_mac_filter *f, *tmp;
struct list_head *ma_list_head;
struct idpf_vport *vport;
@@ -4152,18 +3758,18 @@ static int idpf_mac_filter_async_handler(struct idpf_adapter *adapter,
int i;
/* if success we're done, we're only here if something bad happened */
- if (!ctlq_msg->cookie.mbx.chnl_retval)
- return 0;
+ if (!status || status == -ETIMEDOUT)
+ return;
+ ma_list = buff->iov_base;
/* make sure at least struct is there */
- if (xn->reply_sz < sizeof(*ma_list))
+ if (buff->iov_len < sizeof(*ma_list))
goto invalid_payload;
- ma_list = ctlq_msg->ctx.indirect.payload->va;
mac_addr = ma_list->mac_addr_list;
num_entries = le16_to_cpu(ma_list->num_mac_addr);
/* we should have received a buffer at least this big */
- if (xn->reply_sz < struct_size(ma_list, mac_addr_list, num_entries))
+ if (buff->iov_len < struct_size(ma_list, mac_addr_list, num_entries))
goto invalid_payload;
vport = idpf_vid_to_vport(adapter, le32_to_cpu(ma_list->vport_id));
@@ -4183,16 +3789,13 @@ static int idpf_mac_filter_async_handler(struct idpf_adapter *adapter,
if (ether_addr_equal(mac_addr[i].addr, f->macaddr))
list_del(&f->list);
spin_unlock_bh(&vport_config->mac_filter_list_lock);
- dev_err_ratelimited(&adapter->pdev->dev, "Received error sending MAC filter request (op %d)\n",
- xn->vc_op);
-
- return 0;
+ dev_err_ratelimited(&adapter->pdev->dev, "Received error %d on sending MAC filter request\n",
+ status);
+ return;
invalid_payload:
- dev_err_ratelimited(&adapter->pdev->dev, "Received invalid MAC filter payload (op %d) (len %zd)\n",
- xn->vc_op, xn->reply_sz);
-
- return -EINVAL;
+ dev_err_ratelimited(&adapter->pdev->dev, "Received invalid MAC filter payload (len %zd)\n",
+ buff->iov_len);
}
/**
@@ -4211,19 +3814,21 @@ int idpf_add_del_mac_filters(struct idpf_adapter *adapter,
const u8 *default_mac_addr, u32 vport_id,
bool add, bool async)
{
- struct virtchnl2_mac_addr_list *ma_list __free(kfree) = NULL;
struct virtchnl2_mac_addr *mac_addr __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = add ? VIRTCHNL2_OP_ADD_MAC_ADDR :
+ VIRTCHNL2_OP_DEL_MAC_ADDR,
+ };
+ struct virtchnl2_mac_addr_list *ma_list;
u32 num_msgs, total_filters = 0;
struct idpf_mac_filter *f;
- ssize_t reply_sz;
- int i = 0, k;
+ int i = 0;
- xn_params.vc_op = add ? VIRTCHNL2_OP_ADD_MAC_ADDR :
- VIRTCHNL2_OP_DEL_MAC_ADDR;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.async = async;
- xn_params.async_handler = idpf_mac_filter_async_handler;
+ if (async) {
+ xn_params.resp_cb = idpf_mac_filter_async_handler;
+ xn_params.send_ctx = adapter;
+ }
spin_lock_bh(&vport_config->mac_filter_list_lock);
@@ -4278,32 +3883,31 @@ int idpf_add_del_mac_filters(struct idpf_adapter *adapter,
*/
num_msgs = DIV_ROUND_UP(total_filters, IDPF_NUM_FILTERS_PER_MSG);
- for (i = 0, k = 0; i < num_msgs; i++) {
- u32 entries_size, buf_size, num_entries;
+ for (u32 i = 0, k = 0; i < num_msgs; i++) {
+ u32 entries_size, num_entries;
+ size_t buf_size;
+ int err;
num_entries = min_t(u32, total_filters,
IDPF_NUM_FILTERS_PER_MSG);
entries_size = sizeof(struct virtchnl2_mac_addr) * num_entries;
buf_size = struct_size(ma_list, mac_addr_list, num_entries);
- if (!ma_list || num_entries != IDPF_NUM_FILTERS_PER_MSG) {
- kfree(ma_list);
- ma_list = kzalloc(buf_size, GFP_ATOMIC);
- if (!ma_list)
- return -ENOMEM;
- } else {
- memset(ma_list, 0, buf_size);
- }
+ ma_list = kzalloc(buf_size, GFP_ATOMIC);
+ if (!ma_list)
+ return -ENOMEM;
ma_list->vport_id = cpu_to_le32(vport_id);
ma_list->num_mac_addr = cpu_to_le16(num_entries);
memcpy(ma_list->mac_addr_list, &mac_addr[k], entries_size);
- xn_params.send_buf.iov_base = ma_list;
- xn_params.send_buf.iov_len = buf_size;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, ma_list,
+ buf_size);
+ if (err)
+ return err;
+
+ if (!async)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
k += num_entries;
total_filters -= num_entries;
@@ -4312,6 +3916,26 @@ int idpf_add_del_mac_filters(struct idpf_adapter *adapter,
return 0;
}
+/**
+ * idpf_promiscuous_async_handler - async callback for promiscuous mode
+ * @ctx: controlq context structure
+ * @buff: response buffer pointer and size
+ * @status: async call return value
+ *
+ * Nobody is waiting for the promiscuous virtchnl message response. Print
+ * an error message if something went wrong and return.
+ */
+static void idpf_promiscuous_async_handler(void *ctx,
+ struct kvec *buff,
+ int status)
+{
+ struct idpf_adapter *adapter = ctx;
+
+ if (status)
+ dev_err_ratelimited(&adapter->pdev->dev, "Failed to set promiscuous mode: %d\n",
+ status);
+}
+
/**
* idpf_set_promiscuous - set promiscuous and send message to mailbox
* @adapter: Driver specific private structure
@@ -4326,9 +3950,13 @@ int idpf_set_promiscuous(struct idpf_adapter *adapter,
struct idpf_vport_user_config_data *config_data,
u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_CONFIG_PROMISCUOUS_MODE,
+ .resp_cb = idpf_promiscuous_async_handler,
+ .send_ctx = adapter,
+ };
struct virtchnl2_promisc_info vpi;
- ssize_t reply_sz;
u16 flags = 0;
if (test_bit(__IDPF_PROMISC_UC, config_data->user_flags))
@@ -4339,15 +3967,7 @@ int idpf_set_promiscuous(struct idpf_adapter *adapter,
vpi.vport_id = cpu_to_le32(vport_id);
vpi.flags = cpu_to_le16(flags);
- xn_params.vc_op = VIRTCHNL2_OP_CONFIG_PROMISCUOUS_MODE;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &vpi;
- xn_params.send_buf.iov_len = sizeof(vpi);
- /* setting promiscuous is only ever done asynchronously */
- xn_params.async = true;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
-
- return reply_sz < 0 ? reply_sz : 0;
+ return idpf_send_mb_msg(adapter, &xn_params, &vpi, sizeof(vpi));
}
/**
@@ -4356,7 +3976,7 @@ int idpf_set_promiscuous(struct idpf_adapter *adapter,
* @send_msg: message to send
* @msg_size: size of message to send
* @recv_msg: message to populate on reception of response
- * @recv_len: length of message copied into recv_msg or 0 on error
+ * @recv_len: length of message copied into recv_msg
*
* Return: 0 on success or error code on failure.
*/
@@ -4365,26 +3985,39 @@ int idpf_idc_rdma_vc_send_sync(struct iidc_rdma_core_dev_info *cdev_info,
u8 *recv_msg, u16 *recv_len)
{
struct idpf_adapter *adapter = pci_get_drvdata(cdev_info->pdev);
- struct idpf_vc_xn_params xn_params = { };
- ssize_t reply_sz;
- u16 recv_size;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_RDMA,
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ };
+ u8 on_stack_buf[LIBIE_CP_TX_COPYBREAK];
+ void *send_buf;
+ int err;
- if (!recv_msg || !recv_len || msg_size > IDPF_CTLQ_MAX_BUF_LEN)
+ if (!recv_msg || !recv_len || msg_size > LIBIE_CTLQ_MAX_BUF_LEN)
return -EINVAL;
- recv_size = min_t(u16, *recv_len, IDPF_CTLQ_MAX_BUF_LEN);
- *recv_len = 0;
- xn_params.vc_op = VIRTCHNL2_OP_RDMA;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = send_msg;
- xn_params.send_buf.iov_len = msg_size;
- xn_params.recv_buf.iov_base = recv_msg;
- xn_params.recv_buf.iov_len = recv_size;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- *recv_len = reply_sz;
+ if (!libie_cp_can_send_onstack(msg_size)) {
+ send_buf = kzalloc(msg_size, GFP_KERNEL);
+ if (!send_buf)
+ return -ENOMEM;
+ } else {
+ send_buf = on_stack_buf;
+ }
- return 0;
+ memcpy(send_buf, send_msg, msg_size);
+ err = idpf_send_mb_msg(adapter, &xn_params, send_buf, msg_size);
+ if (err)
+ return err;
+
+ if (xn_params.recv_mem.iov_len > *recv_len) {
+ err = -EINVAL;
+ goto rel_buf;
+ }
+
+ *recv_len = xn_params.recv_mem.iov_len;
+ memcpy(recv_msg, xn_params.recv_mem.iov_base, *recv_len);
+rel_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
EXPORT_SYMBOL_GPL(idpf_idc_rdma_vc_send_sync);
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h
index 7210dc7b426b..23d6e91ff0bf 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h
@@ -7,86 +7,6 @@
#include <linux/net/intel/virtchnl2.h>
#define IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC (60 * 1000)
-#define IDPF_VC_XN_IDX_M GENMASK(7, 0)
-#define IDPF_VC_XN_SALT_M GENMASK(15, 8)
-#define IDPF_VC_XN_RING_LEN U8_MAX
-
-/**
- * enum idpf_vc_xn_state - Virtchnl transaction status
- * @IDPF_VC_XN_IDLE: not expecting a reply, ready to be used
- * @IDPF_VC_XN_WAITING: expecting a reply, not yet received
- * @IDPF_VC_XN_COMPLETED_SUCCESS: a reply was expected and received, buffer
- * updated
- * @IDPF_VC_XN_COMPLETED_FAILED: a reply was expected and received, but there
- * was an error, buffer not updated
- * @IDPF_VC_XN_SHUTDOWN: transaction object cannot be used, VC torn down
- * @IDPF_VC_XN_ASYNC: transaction sent asynchronously and doesn't have the
- * return context; a callback may be provided to handle
- * return
- */
-enum idpf_vc_xn_state {
- IDPF_VC_XN_IDLE = 1,
- IDPF_VC_XN_WAITING,
- IDPF_VC_XN_COMPLETED_SUCCESS,
- IDPF_VC_XN_COMPLETED_FAILED,
- IDPF_VC_XN_SHUTDOWN,
- IDPF_VC_XN_ASYNC,
-};
-
-struct idpf_vc_xn;
-/* Callback for asynchronous messages */
-typedef int (*async_vc_cb) (struct idpf_adapter *, struct idpf_vc_xn *,
- const struct idpf_ctlq_msg *);
-
-/**
- * struct idpf_vc_xn - Data structure representing virtchnl transactions
- * @completed: virtchnl event loop uses that to signal when a reply is
- * available, uses kernel completion API
- * @lock: protects the transaction state fields below
- * @state: virtchnl event loop stores the data below, protected by @lock
- * @reply_sz: Original size of reply, may be > reply_buf.iov_len; it will be
- * truncated on its way to the receiver thread according to
- * reply_buf.iov_len.
- * @reply: Reference to the buffer(s) where the reply data should be written
- * to. May be 0-length (then NULL address permitted) if the reply data
- * should be ignored.
- * @async_handler: if sent asynchronously, a callback can be provided to handle
- * the reply when it's received
- * @vc_op: corresponding opcode sent with this transaction
- * @idx: index used as retrieval on reply receive, used for cookie
- * @salt: changed every message to make unique, used for cookie
- */
-struct idpf_vc_xn {
- struct completion completed;
- spinlock_t lock;
- enum idpf_vc_xn_state state;
- size_t reply_sz;
- struct kvec reply;
- async_vc_cb async_handler;
- u32 vc_op;
- u8 idx;
- u8 salt;
-};
-
-/**
- * struct idpf_vc_xn_params - Parameters for executing transaction
- * @send_buf: kvec for send buffer
- * @recv_buf: kvec for recv buffer, may be NULL, must then have zero length
- * @timeout_ms: timeout to wait for reply
- * @async: send message asynchronously, will not wait on completion
- * @async_handler: If sent asynchronously, optional callback handler. The user
- * must be careful when using async handlers as the memory for
- * the recv_buf _cannot_ be on stack if this is async.
- * @vc_op: virtchnl op to send
- */
-struct idpf_vc_xn_params {
- struct kvec send_buf;
- struct kvec recv_buf;
- int timeout_ms;
- bool async;
- async_vc_cb async_handler;
- u32 vc_op;
-};
struct idpf_adapter;
struct idpf_netdev_priv;
@@ -96,8 +16,6 @@ struct idpf_vport_max_q;
struct idpf_vport_config;
struct idpf_vport_user_config_data;
-ssize_t idpf_vc_xn_exec(struct idpf_adapter *adapter,
- const struct idpf_vc_xn_params *params);
int idpf_init_dflt_mbx(struct idpf_adapter *adapter);
void idpf_deinit_dflt_mbx(struct idpf_adapter *adapter);
int idpf_vc_core_init(struct idpf_adapter *adapter);
@@ -124,9 +42,15 @@ bool idpf_sideband_action_ena(struct idpf_vport *vport,
struct ethtool_rx_flow_spec *fsp);
unsigned int idpf_fsteer_max_rules(struct idpf_vport *vport);
-int idpf_recv_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *arq);
-int idpf_send_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *asq,
- u32 op, u16 msg_size, u8 *msg, u16 cookie);
+void idpf_recv_event_msg(struct libie_ctlq_ctx *ctx,
+ struct libie_ctlq_msg *ctlq_msg);
+int idpf_send_mb_msg(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size);
+int idpf_send_mb_msg_kfree(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size);
+void idpf_send_vf_reset_msg(struct idpf_adapter *adapter);
struct idpf_queue_ptr {
enum virtchnl2_queue_type type;
@@ -212,7 +136,6 @@ int idpf_send_set_rss_key_msg(struct idpf_adapter *adapter,
struct idpf_rss_data *rss_data, u32 vport_id);
int idpf_send_set_rss_lut_msg(struct idpf_adapter *adapter,
struct idpf_rss_data *rss_data, u32 vport_id);
-void idpf_vc_xn_shutdown(struct idpf_vc_xn_manager *vcxn_mngr);
int idpf_idc_rdma_vc_send_sync(struct iidc_rdma_core_dev_info *cdev_info,
u8 *send_msg, u16 msg_size,
u8 *recv_msg, u16 *recv_len);
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
index 8d8fb498e092..1d8dd67e299d 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
@@ -15,7 +15,6 @@
*/
int idpf_ptp_get_caps(struct idpf_adapter *adapter)
{
- struct virtchnl2_ptp_get_caps *recv_ptp_caps_msg __free(kfree) = NULL;
struct virtchnl2_ptp_get_caps send_ptp_caps_msg = {
.caps = cpu_to_le32(VIRTCHNL2_CAP_PTP_GET_DEVICE_CLK_TIME |
VIRTCHNL2_CAP_PTP_GET_DEVICE_CLK_TIME_MB |
@@ -24,34 +23,34 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
VIRTCHNL2_CAP_PTP_ADJ_DEVICE_CLK_MB |
VIRTCHNL2_CAP_PTP_TX_TSTAMPS_MB)
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_CAPS,
- .send_buf.iov_base = &send_ptp_caps_msg,
- .send_buf.iov_len = sizeof(send_ptp_caps_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_CAPS,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
struct virtchnl2_ptp_cross_time_reg_offsets cross_tstamp_offsets;
struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
struct virtchnl2_ptp_clk_adj_reg_offsets clk_adj_offsets;
struct virtchnl2_ptp_clk_reg_offsets clock_offsets;
+ struct virtchnl2_ptp_get_caps *recv_ptp_caps_msg;
struct idpf_ptp_secondary_mbx *scnd_mbx;
struct idpf_ptp *ptp = adapter->ptp;
enum idpf_ptp_access access_type;
u32 temp_offset;
- int reply_sz;
+ size_t reply_sz;
+ int err;
- recv_ptp_caps_msg = kzalloc_obj(struct virtchnl2_ptp_get_caps);
- if (!recv_ptp_caps_msg)
- return -ENOMEM;
+ err = idpf_send_mb_msg(adapter, &xn_params, &send_ptp_caps_msg,
+ sizeof(send_ptp_caps_msg));
+ if (err)
+ return err;
- xn_params.recv_buf.iov_base = recv_ptp_caps_msg;
- xn_params.recv_buf.iov_len = sizeof(*recv_ptp_caps_msg);
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz != sizeof(*recv_ptp_caps_msg)) {
+ err = -EIO;
+ goto free_resp;
+ }
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- else if (reply_sz != sizeof(*recv_ptp_caps_msg))
- return -EIO;
+ recv_ptp_caps_msg = xn_params.recv_mem.iov_base;
ptp->caps = le32_to_cpu(recv_ptp_caps_msg->caps);
ptp->base_incval = le64_to_cpu(recv_ptp_caps_msg->base_incval);
@@ -112,7 +111,7 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
discipline_clock:
access_type = ptp->adj_dev_clk_time_access;
if (access_type != IDPF_PTP_DIRECT)
- return 0;
+ goto free_resp;
clk_adj_offsets = recv_ptp_caps_msg->clk_adj_offsets;
@@ -145,7 +144,9 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
ptp->dev_clk_regs.phy_shadj_h =
libie_pci_get_mmio_addr(mmio, temp_offset);
- return 0;
+free_resp:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -160,28 +161,34 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
int idpf_ptp_get_dev_clk_time(struct idpf_adapter *adapter,
struct idpf_ptp_dev_timers *dev_clk_time)
{
+ struct virtchnl2_ptp_get_dev_clk_time *get_dev_clk_time_resp;
struct virtchnl2_ptp_get_dev_clk_time get_dev_clk_time_msg;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_DEV_CLK_TIME,
- .send_buf.iov_base = &get_dev_clk_time_msg,
- .send_buf.iov_len = sizeof(get_dev_clk_time_msg),
- .recv_buf.iov_base = &get_dev_clk_time_msg,
- .recv_buf.iov_len = sizeof(get_dev_clk_time_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_DEV_CLK_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ size_t reply_sz;
u64 dev_time;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(get_dev_clk_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &get_dev_clk_time_msg,
+ sizeof(get_dev_clk_time_msg));
+ if (err)
+ return err;
- dev_time = le64_to_cpu(get_dev_clk_time_msg.dev_time_ns);
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz != sizeof(*get_dev_clk_time_resp)) {
+ err = -EIO;
+ goto free_resp;
+ }
+
+ get_dev_clk_time_resp = xn_params.recv_mem.iov_base;
+ dev_time = le64_to_cpu(get_dev_clk_time_resp->dev_time_ns);
dev_clk_time->dev_clk_time_ns = dev_time;
- return 0;
+free_resp:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -197,27 +204,30 @@ int idpf_ptp_get_dev_clk_time(struct idpf_adapter *adapter,
int idpf_ptp_get_cross_time(struct idpf_adapter *adapter,
struct idpf_ptp_dev_timers *cross_time)
{
- struct virtchnl2_ptp_get_cross_time cross_time_msg;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_CROSS_TIME,
- .send_buf.iov_base = &cross_time_msg,
- .send_buf.iov_len = sizeof(cross_time_msg),
- .recv_buf.iov_base = &cross_time_msg,
- .recv_buf.iov_len = sizeof(cross_time_msg),
+ struct virtchnl2_ptp_get_cross_time cross_time_send, *cross_time_recv;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_CROSS_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err = 0;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(cross_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &cross_time_send,
+ sizeof(cross_time_send));
+ if (err)
+ return err;
- cross_time->dev_clk_time_ns = le64_to_cpu(cross_time_msg.dev_time_ns);
- cross_time->sys_time_ns = le64_to_cpu(cross_time_msg.sys_time_ns);
+ if (xn_params.recv_mem.iov_len != sizeof(*cross_time_recv)) {
+ err = -EIO;
+ goto free_resp;
+ }
- return 0;
+ cross_time_recv = xn_params.recv_mem.iov_base;
+ cross_time->dev_clk_time_ns = le64_to_cpu(cross_time_recv->dev_time_ns);
+ cross_time->sys_time_ns = le64_to_cpu(cross_time_recv->sys_time_ns);
+
+free_resp:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -234,23 +244,18 @@ int idpf_ptp_set_dev_clk_time(struct idpf_adapter *adapter, u64 time)
struct virtchnl2_ptp_set_dev_clk_time set_dev_clk_time_msg = {
.dev_time_ns = cpu_to_le64(time),
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_SET_DEV_CLK_TIME,
- .send_buf.iov_base = &set_dev_clk_time_msg,
- .send_buf.iov_len = sizeof(set_dev_clk_time_msg),
- .recv_buf.iov_base = &set_dev_clk_time_msg,
- .recv_buf.iov_len = sizeof(set_dev_clk_time_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_SET_DEV_CLK_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(set_dev_clk_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &set_dev_clk_time_msg,
+ sizeof(set_dev_clk_time_msg));
+ if (!err)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -267,23 +272,18 @@ int idpf_ptp_adj_dev_clk_time(struct idpf_adapter *adapter, s64 delta)
struct virtchnl2_ptp_adj_dev_clk_time adj_dev_clk_time_msg = {
.delta = cpu_to_le64(delta),
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_TIME,
- .send_buf.iov_base = &adj_dev_clk_time_msg,
- .send_buf.iov_len = sizeof(adj_dev_clk_time_msg),
- .recv_buf.iov_base = &adj_dev_clk_time_msg,
- .recv_buf.iov_len = sizeof(adj_dev_clk_time_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(adj_dev_clk_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &adj_dev_clk_time_msg,
+ sizeof(adj_dev_clk_time_msg));
+ if (!err)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -301,23 +301,18 @@ int idpf_ptp_adj_dev_clk_fine(struct idpf_adapter *adapter, u64 incval)
struct virtchnl2_ptp_adj_dev_clk_fine adj_dev_clk_fine_msg = {
.incval = cpu_to_le64(incval),
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_FINE,
- .send_buf.iov_base = &adj_dev_clk_fine_msg,
- .send_buf.iov_len = sizeof(adj_dev_clk_fine_msg),
- .recv_buf.iov_base = &adj_dev_clk_fine_msg,
- .recv_buf.iov_len = sizeof(adj_dev_clk_fine_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_FINE,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(adj_dev_clk_fine_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &adj_dev_clk_fine_msg,
+ sizeof(adj_dev_clk_fine_msg));
+ if (!err)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -336,18 +331,16 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
struct virtchnl2_ptp_tx_tstamp_latch_caps tx_tstamp_latch_caps;
struct idpf_ptp_vport_tx_tstamp_caps *tstamp_caps;
struct idpf_ptp_tx_tstamp *ptp_tx_tstamp, *tmp;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP_CAPS,
- .send_buf.iov_base = &send_tx_tstamp_caps,
- .send_buf.iov_len = sizeof(send_tx_tstamp_caps),
- .recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN,
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP_CAPS,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
enum idpf_ptp_access tstamp_access, get_dev_clk_access;
struct idpf_ptp *ptp = vport->adapter->ptp;
struct list_head *head;
- int err = 0, reply_sz;
+ size_t reply_sz;
u16 num_latches;
+ int err = 0;
u32 size;
if (!ptp)
@@ -359,19 +352,19 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
get_dev_clk_access == IDPF_PTP_NONE)
return -EOPNOTSUPP;
- rcv_tx_tstamp_caps = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!rcv_tx_tstamp_caps)
- return -ENOMEM;
-
send_tx_tstamp_caps.vport_id = cpu_to_le32(vport->vport_id);
- xn_params.recv_buf.iov_base = rcv_tx_tstamp_caps;
- reply_sz = idpf_vc_xn_exec(vport->adapter, &xn_params);
- if (reply_sz < 0) {
- err = reply_sz;
+ err = idpf_send_mb_msg(vport->adapter, &xn_params, &send_tx_tstamp_caps,
+ sizeof(send_tx_tstamp_caps));
+ if (err)
+ return err;
+
+ rcv_tx_tstamp_caps = xn_params.recv_mem.iov_base;
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz < sizeof(*rcv_tx_tstamp_caps)) {
+ err = -EIO;
goto get_tstamp_caps_out;
}
-
num_latches = le16_to_cpu(rcv_tx_tstamp_caps->num_latches);
size = struct_size(rcv_tx_tstamp_caps, tstamp_latches, num_latches);
if (reply_sz != size) {
@@ -426,7 +419,7 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
}
vport->tx_tstamp_caps = tstamp_caps;
- kfree(rcv_tx_tstamp_caps);
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
return 0;
@@ -439,7 +432,7 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
kfree(tstamp_caps);
get_tstamp_caps_out:
- kfree(rcv_tx_tstamp_caps);
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
return err;
}
@@ -536,9 +529,9 @@ idpf_ptp_get_tstamp_value(struct idpf_vport *vport,
/**
* idpf_ptp_get_tx_tstamp_async_handler - Async callback for getting Tx tstamps
- * @adapter: Driver specific private structure
- * @xn: transaction for message
- * @ctlq_msg: received message
+ * @ctx: adapter pointer
+ * @mem: address and size of the response
+ * @status: return value of the request
*
* Read the tstamps Tx tstamp values from a received message and put them
* directly to the skb. The number of timestamps to read is specified by
@@ -546,22 +539,26 @@ idpf_ptp_get_tstamp_value(struct idpf_vport *vport,
*
* Return: 0 on success, -errno otherwise.
*/
-static int
-idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
- struct idpf_vc_xn *xn,
- const struct idpf_ctlq_msg *ctlq_msg)
+static void
+idpf_ptp_get_tx_tstamp_async_handler(void *ctx, struct kvec *mem, int status)
{
struct virtchnl2_ptp_get_vport_tx_tstamp_latches *recv_tx_tstamp_msg;
struct idpf_ptp_vport_tx_tstamp_caps *tx_tstamp_caps;
struct virtchnl2_ptp_tx_tstamp_latch tstamp_latch;
struct idpf_ptp_tx_tstamp *tx_tstamp, *tmp;
struct idpf_vport *tstamp_vport = NULL;
+ struct idpf_adapter *adapter = ctx;
struct list_head *head;
u16 num_latches;
u32 vport_id;
- int err = 0;
- recv_tx_tstamp_msg = ctlq_msg->ctx.indirect.payload->va;
+ if (status)
+ return;
+
+ recv_tx_tstamp_msg = mem->iov_base;
+ if (mem->iov_len < sizeof(*recv_tx_tstamp_msg))
+ return;
+
vport_id = le32_to_cpu(recv_tx_tstamp_msg->vport_id);
idpf_for_each_vport(adapter, vport) {
@@ -575,10 +572,13 @@ idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
}
if (!tstamp_vport || !tstamp_vport->tx_tstamp_caps)
- return -EINVAL;
+ return;
tx_tstamp_caps = tstamp_vport->tx_tstamp_caps;
num_latches = le16_to_cpu(recv_tx_tstamp_msg->num_latches);
+ if (mem->iov_len < struct_size(recv_tx_tstamp_msg, tstamp_latches,
+ num_latches))
+ return;
spin_lock_bh(&tx_tstamp_caps->latches_lock);
head = &tx_tstamp_caps->latches_in_use;
@@ -589,13 +589,13 @@ idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
if (!tstamp_latch.valid)
continue;
- if (list_empty(head)) {
- err = -ENOBUFS;
+ if (list_empty(head))
goto unlock;
- }
list_for_each_entry_safe(tx_tstamp, tmp, head, list_member) {
if (tstamp_latch.index == tx_tstamp->idx) {
+ int err;
+
list_del(&tx_tstamp->list_member);
err = idpf_ptp_get_tstamp_value(tstamp_vport,
&tstamp_latch,
@@ -610,8 +610,6 @@ idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
unlock:
spin_unlock_bh(&tx_tstamp_caps->latches_lock);
-
- return err;
}
/**
@@ -627,15 +625,15 @@ int idpf_ptp_get_tx_tstamp(struct idpf_vport *vport)
{
struct virtchnl2_ptp_get_vport_tx_tstamp_latches *send_tx_tstamp_msg;
struct idpf_ptp_vport_tx_tstamp_caps *tx_tstamp_caps;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP,
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
- .async = true,
- .async_handler = idpf_ptp_get_tx_tstamp_async_handler,
+ .resp_cb = idpf_ptp_get_tx_tstamp_async_handler,
+ .send_ctx = vport->adapter,
};
struct idpf_ptp_tx_tstamp *ptp_tx_tstamp;
- int reply_sz, size, msg_size;
struct list_head *head;
+ int size, msg_size;
bool state_upd;
u16 id = 0;
@@ -668,11 +666,7 @@ int idpf_ptp_get_tx_tstamp(struct idpf_vport *vport)
msg_size = struct_size(send_tx_tstamp_msg, tstamp_latches, id);
send_tx_tstamp_msg->vport_id = cpu_to_le32(vport->vport_id);
send_tx_tstamp_msg->num_latches = cpu_to_le16(id);
- xn_params.send_buf.iov_base = send_tx_tstamp_msg;
- xn_params.send_buf.iov_len = msg_size;
-
- reply_sz = idpf_vc_xn_exec(vport->adapter, &xn_params);
- kfree(send_tx_tstamp_msg);
- return min(reply_sz, 0);
+ return idpf_send_mb_msg_kfree(vport->adapter, &xn_params,
+ send_tx_tstamp_msg, msg_size);
}
--
2.47.1
^ permalink raw reply related
* [PATCH net-next v5 08/15] idpf: refactor idpf to use libie_pci APIs
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Pavan Kumar Linga, anthony.l.nguyen, larysa.zaremba,
przemyslaw.kitszel, aleksander.lobakin, sridhar.samudrala,
michal.swiatkowski, maciej.fijalkowski, emil.s.tantilov,
joshua.a.hay, jacob.e.keller, jayaprakash.shanmugam, jiri, horms,
corbet, richardcochran, skhan, linux-doc, Samuel Salin
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Use libie_pci init and MMIO APIs where possible, struct idpf_hw cannot be
deleted for now as it also houses control queues that will be refactored
later. Use libie_cp header for libie_ctlq_ctx that contains mmio info from
the start in order to not increase the diff later.
Reviewed-by: Madhu Chittim <madhu.chittim@intel.com>
Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Tested-by: Samuel Salin <Samuel.salin@intel.com>
Co-developed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/idpf/Kconfig | 1 +
drivers/net/ethernet/intel/idpf/idpf.h | 70 +-------
.../net/ethernet/intel/idpf/idpf_controlq.c | 26 ++-
.../net/ethernet/intel/idpf/idpf_controlq.h | 2 -
drivers/net/ethernet/intel/idpf/idpf_dev.c | 61 ++++---
drivers/net/ethernet/intel/idpf/idpf_idc.c | 38 ++--
drivers/net/ethernet/intel/idpf/idpf_lib.c | 7 +-
drivers/net/ethernet/intel/idpf/idpf_main.c | 114 ++++++------
drivers/net/ethernet/intel/idpf/idpf_vf_dev.c | 57 +++---
.../net/ethernet/intel/idpf/idpf_virtchnl.c | 169 +++++++++---------
.../ethernet/intel/idpf/idpf_virtchnl_ptp.c | 58 +++---
11 files changed, 288 insertions(+), 315 deletions(-)
diff --git a/drivers/net/ethernet/intel/idpf/Kconfig b/drivers/net/ethernet/intel/idpf/Kconfig
index adab2154125b..586df3a4afe9 100644
--- a/drivers/net/ethernet/intel/idpf/Kconfig
+++ b/drivers/net/ethernet/intel/idpf/Kconfig
@@ -6,6 +6,7 @@ config IDPF
depends on PCI_MSI
depends on PTP_1588_CLOCK_OPTIONAL
select DIMLIB
+ select LIBIE_CP
select LIBETH_XDP
help
This driver supports Intel(R) Infrastructure Data Path Function
diff --git a/drivers/net/ethernet/intel/idpf/idpf.h b/drivers/net/ethernet/intel/idpf/idpf.h
index c5e47e79a641..92a120aadfcd 100644
--- a/drivers/net/ethernet/intel/idpf/idpf.h
+++ b/drivers/net/ethernet/intel/idpf/idpf.h
@@ -23,6 +23,7 @@ struct idpf_rss_data;
#include <linux/net/intel/iidc_rdma.h>
#include <linux/net/intel/iidc_rdma_idpf.h>
+#include <linux/net/intel/libie/controlq.h>
#include <linux/net/intel/virtchnl2.h>
#include "idpf_txrx.h"
@@ -625,6 +626,7 @@ struct idpf_vc_xn_manager;
* @flags: See enum idpf_flags
* @reset_reg: See struct idpf_reset_reg
* @hw: Device access data
+ * @ctlq_ctx: controlq context
* @num_avail_msix: Available number of MSIX vectors
* @num_msix_entries: Number of entries in MSIX table
* @msix_entries: MSIX table
@@ -682,6 +684,7 @@ struct idpf_adapter {
DECLARE_BITMAP(flags, IDPF_FLAGS_NBITS);
struct idpf_reset_reg reset_reg;
struct idpf_hw hw;
+ struct libie_ctlq_ctx ctlq_ctx;
u16 num_avail_msix;
u16 num_msix_entries;
struct msix_entry *msix_entries;
@@ -870,70 +873,6 @@ static inline u8 idpf_get_min_tx_pkt_len(struct idpf_adapter *adapter)
return pkt_len ? pkt_len : IDPF_TX_MIN_PKT_LEN;
}
-/**
- * idpf_get_mbx_reg_addr - Get BAR0 mailbox register address
- * @adapter: private data struct
- * @reg_offset: register offset value
- *
- * Return: BAR0 mailbox register address based on register offset.
- */
-static inline void __iomem *idpf_get_mbx_reg_addr(struct idpf_adapter *adapter,
- resource_size_t reg_offset)
-{
- return adapter->hw.mbx.vaddr + reg_offset;
-}
-
-/**
- * idpf_get_rstat_reg_addr - Get BAR0 rstat register address
- * @adapter: private data struct
- * @reg_offset: register offset value
- *
- * Return: BAR0 rstat register address based on register offset.
- */
-static inline void __iomem *idpf_get_rstat_reg_addr(struct idpf_adapter *adapter,
- resource_size_t reg_offset)
-{
- reg_offset -= adapter->dev_ops.static_reg_info[1].start;
-
- return adapter->hw.rstat.vaddr + reg_offset;
-}
-
-/**
- * idpf_get_reg_addr - Get BAR0 register address
- * @adapter: private data struct
- * @reg_offset: register offset value
- *
- * Based on the register offset, return the actual BAR0 register address
- */
-static inline void __iomem *idpf_get_reg_addr(struct idpf_adapter *adapter,
- resource_size_t reg_offset)
-{
- struct idpf_hw *hw = &adapter->hw;
-
- for (int i = 0; i < hw->num_lan_regs; i++) {
- struct idpf_mmio_reg *region = &hw->lan_regs[i];
-
- if (reg_offset >= region->addr_start &&
- reg_offset < (region->addr_start + region->addr_len)) {
- /* Convert the offset so that it is relative to the
- * start of the region. Then add the base address of
- * the region to get the final address.
- */
- reg_offset -= region->addr_start;
-
- return region->vaddr + reg_offset;
- }
- }
-
- /* It's impossible to hit this case with offsets from the CP. But if we
- * do for any other reason, the kernel will panic on that register
- * access. Might as well do it here to make it clear what's happening.
- */
- BUG();
-
- return NULL;
-}
-
/**
* idpf_is_reset_detected - check if we were reset at some point
* @adapter: driver specific private structure
@@ -945,7 +884,8 @@ static inline bool idpf_is_reset_detected(struct idpf_adapter *adapter)
if (!adapter->hw.arq)
return true;
- return !(readl(idpf_get_mbx_reg_addr(adapter, adapter->hw.arq->reg.len)) &
+ return !(readl(libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info,
+ adapter->hw.arq->reg.len)) &
adapter->hw.arq->reg.len_mask);
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq.c b/drivers/net/ethernet/intel/idpf/idpf_controlq.c
index d2dde43269e9..020b08367e18 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_controlq.c
@@ -1,7 +1,7 @@
// SPDX-License-Identifier: GPL-2.0-only
/* Copyright (C) 2023 Intel Corporation */
-#include "idpf_controlq.h"
+#include "idpf.h"
/**
* idpf_ctlq_setup_regs - initialize control queue registers
@@ -34,21 +34,27 @@ static void idpf_ctlq_setup_regs(struct idpf_ctlq_info *cq,
static void idpf_ctlq_init_regs(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
bool is_rxq)
{
+ struct libie_mmio_info *mmio = &hw->back->ctlq_ctx.mmio_info;
+
/* Update tail to post pre-allocated buffers for rx queues */
if (is_rxq)
- idpf_mbx_wr32(hw, cq->reg.tail, (u32)(cq->ring_size - 1));
+ writel((u32)(cq->ring_size - 1),
+ libie_pci_get_mmio_addr(mmio, cq->reg.tail));
/* For non-Mailbox control queues only TAIL need to be set */
if (cq->q_id != -1)
return;
/* Clear Head for both send or receive */
- idpf_mbx_wr32(hw, cq->reg.head, 0);
+ writel(0, libie_pci_get_mmio_addr(mmio, cq->reg.head));
/* set starting point */
- idpf_mbx_wr32(hw, cq->reg.bal, lower_32_bits(cq->desc_ring.pa));
- idpf_mbx_wr32(hw, cq->reg.bah, upper_32_bits(cq->desc_ring.pa));
- idpf_mbx_wr32(hw, cq->reg.len, (cq->ring_size | cq->reg.len_ena_mask));
+ writel(lower_32_bits(cq->desc_ring.pa),
+ libie_pci_get_mmio_addr(mmio, cq->reg.bal));
+ writel(upper_32_bits(cq->desc_ring.pa),
+ libie_pci_get_mmio_addr(mmio, cq->reg.bah));
+ writel((cq->ring_size | cq->reg.len_ena_mask),
+ libie_pci_get_mmio_addr(mmio, cq->reg.len));
}
/**
@@ -326,7 +332,9 @@ int idpf_ctlq_send(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
*/
dma_wmb();
- idpf_mbx_wr32(hw, cq->reg.tail, cq->next_to_use);
+ writel(cq->next_to_use,
+ libie_pci_get_mmio_addr(&hw->back->ctlq_ctx.mmio_info,
+ cq->reg.tail));
err_unlock:
spin_unlock(&cq->cq_lock);
@@ -518,7 +526,9 @@ int idpf_ctlq_post_rx_buffs(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
dma_wmb();
- idpf_mbx_wr32(hw, cq->reg.tail, cq->next_to_post);
+ writel(cq->next_to_post,
+ libie_pci_get_mmio_addr(&hw->back->ctlq_ctx.mmio_info,
+ cq->reg.tail));
}
spin_unlock(&cq->cq_lock);
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq.h b/drivers/net/ethernet/intel/idpf/idpf_controlq.h
index de4ece40c2ff..acf595e9265f 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq.h
+++ b/drivers/net/ethernet/intel/idpf/idpf_controlq.h
@@ -109,8 +109,6 @@ struct idpf_mmio_reg {
* Align to ctlq_hw_info
*/
struct idpf_hw {
- struct idpf_mmio_reg mbx;
- struct idpf_mmio_reg rstat;
/* Array of remaining LAN BAR regions */
int num_lan_regs;
struct idpf_mmio_reg *lan_regs;
diff --git a/drivers/net/ethernet/intel/idpf/idpf_dev.c b/drivers/net/ethernet/intel/idpf/idpf_dev.c
index 1a0c71c95ef1..e36b0017186f 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_dev.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_dev.c
@@ -16,7 +16,6 @@
static void idpf_ctlq_reg_init(struct idpf_adapter *adapter,
struct idpf_ctlq_create_info *cq)
{
- resource_size_t mbx_start = adapter->dev_ops.static_reg_info[0].start;
int i;
for (i = 0; i < IDPF_NUM_DFLT_MBX_Q; i++) {
@@ -25,22 +24,22 @@ static void idpf_ctlq_reg_init(struct idpf_adapter *adapter,
switch (ccq->type) {
case IDPF_CTLQ_TYPE_MAILBOX_TX:
/* set head and tail registers in our local struct */
- ccq->reg.head = PF_FW_ATQH - mbx_start;
- ccq->reg.tail = PF_FW_ATQT - mbx_start;
- ccq->reg.len = PF_FW_ATQLEN - mbx_start;
- ccq->reg.bah = PF_FW_ATQBAH - mbx_start;
- ccq->reg.bal = PF_FW_ATQBAL - mbx_start;
+ ccq->reg.head = PF_FW_ATQH;
+ ccq->reg.tail = PF_FW_ATQT;
+ ccq->reg.len = PF_FW_ATQLEN;
+ ccq->reg.bah = PF_FW_ATQBAH;
+ ccq->reg.bal = PF_FW_ATQBAL;
ccq->reg.len_mask = PF_FW_ATQLEN_ATQLEN_M;
ccq->reg.len_ena_mask = PF_FW_ATQLEN_ATQENABLE_M;
ccq->reg.head_mask = PF_FW_ATQH_ATQH_M;
break;
case IDPF_CTLQ_TYPE_MAILBOX_RX:
/* set head and tail registers in our local struct */
- ccq->reg.head = PF_FW_ARQH - mbx_start;
- ccq->reg.tail = PF_FW_ARQT - mbx_start;
- ccq->reg.len = PF_FW_ARQLEN - mbx_start;
- ccq->reg.bah = PF_FW_ARQBAH - mbx_start;
- ccq->reg.bal = PF_FW_ARQBAL - mbx_start;
+ ccq->reg.head = PF_FW_ARQH;
+ ccq->reg.tail = PF_FW_ARQT;
+ ccq->reg.len = PF_FW_ARQLEN;
+ ccq->reg.bah = PF_FW_ARQBAH;
+ ccq->reg.bal = PF_FW_ARQBAL;
ccq->reg.len_mask = PF_FW_ARQLEN_ARQLEN_M;
ccq->reg.len_ena_mask = PF_FW_ARQLEN_ARQENABLE_M;
ccq->reg.head_mask = PF_FW_ARQH_ARQH_M;
@@ -57,13 +56,14 @@ static void idpf_ctlq_reg_init(struct idpf_adapter *adapter,
*/
static void idpf_mb_intr_reg_init(struct idpf_adapter *adapter)
{
+ struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
struct idpf_intr_reg *intr = &adapter->mb_vector.intr_reg;
u32 dyn_ctl = le32_to_cpu(adapter->caps.mailbox_dyn_ctl);
- intr->dyn_ctl = idpf_get_reg_addr(adapter, dyn_ctl);
+ intr->dyn_ctl = libie_pci_get_mmio_addr(mmio, dyn_ctl);
intr->dyn_ctl_intena_m = PF_GLINT_DYN_CTL_INTENA_M;
intr->dyn_ctl_itridx_m = PF_GLINT_DYN_CTL_ITR_INDX_M;
- intr->icr_ena = idpf_get_reg_addr(adapter, PF_INT_DIR_OICR_ENA);
+ intr->icr_ena = libie_pci_get_mmio_addr(mmio, PF_INT_DIR_OICR_ENA);
intr->icr_ena_ctlq_m = PF_INT_DIR_OICR_ENA_M;
}
@@ -78,6 +78,7 @@ static int idpf_intr_reg_init(struct idpf_vport *vport,
struct idpf_adapter *adapter = vport->adapter;
u16 num_vecs = rsrc->num_q_vectors;
struct idpf_vec_regs *reg_vals;
+ struct libie_mmio_info *mmio;
int num_regs, i, err = 0;
u32 rx_itr, tx_itr, val;
u16 total_vecs;
@@ -93,14 +94,17 @@ static int idpf_intr_reg_init(struct idpf_vport *vport,
goto free_reg_vals;
}
+ mmio = &adapter->ctlq_ctx.mmio_info;
+
for (i = 0; i < num_vecs; i++) {
struct idpf_q_vector *q_vector = &rsrc->q_vectors[i];
u16 vec_id = rsrc->q_vector_idxs[i] - IDPF_MBX_Q_VEC;
struct idpf_intr_reg *intr = &q_vector->intr_reg;
+ struct idpf_vec_regs *reg = ®_vals[vec_id];
u32 spacing;
- intr->dyn_ctl = idpf_get_reg_addr(adapter,
- reg_vals[vec_id].dyn_ctl_reg);
+ intr->dyn_ctl = libie_pci_get_mmio_addr(mmio,
+ reg->dyn_ctl_reg);
intr->dyn_ctl_intena_m = PF_GLINT_DYN_CTL_INTENA_M;
intr->dyn_ctl_intena_msk_m = PF_GLINT_DYN_CTL_INTENA_MSK_M;
intr->dyn_ctl_itridx_s = PF_GLINT_DYN_CTL_ITR_INDX_S;
@@ -110,22 +114,21 @@ static int idpf_intr_reg_init(struct idpf_vport *vport,
intr->dyn_ctl_sw_itridx_ena_m =
PF_GLINT_DYN_CTL_SW_ITR_INDX_ENA_M;
- spacing = IDPF_ITR_IDX_SPACING(reg_vals[vec_id].itrn_index_spacing,
+ spacing = IDPF_ITR_IDX_SPACING(reg->itrn_index_spacing,
IDPF_PF_ITR_IDX_SPACING);
rx_itr = PF_GLINT_ITR_ADDR(VIRTCHNL2_ITR_IDX_0,
- reg_vals[vec_id].itrn_reg,
- spacing);
+ reg->itrn_reg, spacing);
tx_itr = PF_GLINT_ITR_ADDR(VIRTCHNL2_ITR_IDX_1,
- reg_vals[vec_id].itrn_reg,
- spacing);
- intr->rx_itr = idpf_get_reg_addr(adapter, rx_itr);
- intr->tx_itr = idpf_get_reg_addr(adapter, tx_itr);
+ reg->itrn_reg, spacing);
+ intr->rx_itr = libie_pci_get_mmio_addr(mmio, rx_itr);
+ intr->tx_itr = libie_pci_get_mmio_addr(mmio, tx_itr);
}
/* Data vector for NOIRQ queues */
val = reg_vals[rsrc->q_vector_idxs[i] - IDPF_MBX_Q_VEC].dyn_ctl_reg;
- rsrc->noirq_dyn_ctl = idpf_get_reg_addr(adapter, val);
+ rsrc->noirq_dyn_ctl =
+ libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info, val);
val = PF_GLINT_DYN_CTL_WB_ON_ITR_M | PF_GLINT_DYN_CTL_INTENA_MSK_M |
FIELD_PREP(PF_GLINT_DYN_CTL_ITR_INDX_M, IDPF_NO_ITR_UPDATE_IDX);
@@ -143,7 +146,9 @@ static int idpf_intr_reg_init(struct idpf_vport *vport,
*/
static void idpf_reset_reg_init(struct idpf_adapter *adapter)
{
- adapter->reset_reg.rstat = idpf_get_rstat_reg_addr(adapter, PFGEN_RSTAT);
+ adapter->reset_reg.rstat =
+ libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info,
+ PFGEN_RSTAT);
adapter->reset_reg.rstat_m = PFGEN_RSTAT_PFR_STATE_M;
}
@@ -155,11 +160,11 @@ static void idpf_reset_reg_init(struct idpf_adapter *adapter)
static void idpf_trigger_reset(struct idpf_adapter *adapter,
enum idpf_flags __always_unused trig_cause)
{
- u32 reset_reg;
+ void __iomem *addr;
- reset_reg = readl(idpf_get_rstat_reg_addr(adapter, PFGEN_CTRL));
- writel(reset_reg | PFGEN_CTRL_PFSWR,
- idpf_get_rstat_reg_addr(adapter, PFGEN_CTRL));
+ addr = libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info,
+ PFGEN_CTRL);
+ writel(readl(addr) | PFGEN_CTRL_PFSWR, addr);
}
/**
diff --git a/drivers/net/ethernet/intel/idpf/idpf_idc.c b/drivers/net/ethernet/intel/idpf/idpf_idc.c
index b7d6b08fc89e..0a7edb783758 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_idc.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_idc.c
@@ -416,9 +416,12 @@ idpf_idc_init_msix_data(struct idpf_adapter *adapter)
int idpf_idc_init_aux_core_dev(struct idpf_adapter *adapter,
enum iidc_function_type ftype)
{
+ struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
struct iidc_rdma_core_dev_info *cdev_info;
struct iidc_rdma_priv_dev_info *privd;
- int err, i;
+ struct libie_pci_mmio_region *mr;
+ size_t num_mem_regions;
+ int err, i = 0;
adapter->cdev_info = kzalloc_obj(*cdev_info);
if (!adapter->cdev_info)
@@ -436,22 +439,37 @@ int idpf_idc_init_aux_core_dev(struct idpf_adapter *adapter,
cdev_info->rdma_protocol = IIDC_RDMA_PROTOCOL_ROCEV2;
privd->ftype = ftype;
+ num_mem_regions = list_count_nodes(&mmio->mmio_list);
+ if (num_mem_regions <= IDPF_MMIO_REG_NUM_STATIC) {
+ err = -EINVAL;
+ goto err_plug_aux_dev;
+ }
+
+ num_mem_regions -= IDPF_MMIO_REG_NUM_STATIC;
privd->mapped_mem_regions =
kzalloc_objs(struct iidc_rdma_lan_mapped_mem_region,
- adapter->hw.num_lan_regs);
+ num_mem_regions);
if (!privd->mapped_mem_regions) {
err = -ENOMEM;
goto err_plug_aux_dev;
}
- privd->num_memory_regions = cpu_to_le16(adapter->hw.num_lan_regs);
- for (i = 0; i < adapter->hw.num_lan_regs; i++) {
- privd->mapped_mem_regions[i].region_addr =
- adapter->hw.lan_regs[i].vaddr;
- privd->mapped_mem_regions[i].size =
- cpu_to_le64(adapter->hw.lan_regs[i].addr_len);
- privd->mapped_mem_regions[i].start_offset =
- cpu_to_le64(adapter->hw.lan_regs[i].addr_start);
+ privd->num_memory_regions = cpu_to_le16(num_mem_regions);
+ list_for_each_entry(mr, &mmio->mmio_list, list) {
+ struct resource *static_regs = adapter->dev_ops.static_reg_info;
+ bool is_static = false;
+
+ for (uint j = 0; j < IDPF_MMIO_REG_NUM_STATIC; j++)
+ if (mr->offset == static_regs[j].start)
+ is_static = true;
+
+ if (is_static)
+ continue;
+
+ privd->mapped_mem_regions[i].region_addr = mr->addr;
+ privd->mapped_mem_regions[i].size = cpu_to_le64(mr->size);
+ privd->mapped_mem_regions[i++].start_offset =
+ cpu_to_le64(mr->offset);
}
idpf_idc_init_msix_data(adapter);
diff --git a/drivers/net/ethernet/intel/idpf/idpf_lib.c b/drivers/net/ethernet/intel/idpf/idpf_lib.c
index d88ca59edf97..875472ae77fd 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_lib.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_lib.c
@@ -1847,15 +1847,14 @@ void idpf_deinit_task(struct idpf_adapter *adapter)
/**
* idpf_check_reset_complete - check that reset is complete
- * @hw: pointer to hw struct
+ * @adapter: adapter to check
* @reset_reg: struct with reset registers
*
* Returns 0 if device is ready to use, or -EBUSY if it's in reset.
**/
-static int idpf_check_reset_complete(struct idpf_hw *hw,
+static int idpf_check_reset_complete(struct idpf_adapter *adapter,
struct idpf_reset_reg *reset_reg)
{
- struct idpf_adapter *adapter = hw->back;
int i;
for (i = 0; i < 2000; i++) {
@@ -1918,7 +1917,7 @@ static void idpf_init_hard_reset(struct idpf_adapter *adapter)
}
/* Wait for reset to complete */
- err = idpf_check_reset_complete(&adapter->hw, &adapter->reset_reg);
+ err = idpf_check_reset_complete(adapter, &adapter->reset_reg);
if (err) {
dev_err(dev, "The driver was unable to contact the device's firmware. Check that the FW is running. Driver state= 0x%x\n",
adapter->state);
diff --git a/drivers/net/ethernet/intel/idpf/idpf_main.c b/drivers/net/ethernet/intel/idpf/idpf_main.c
index ab3c409e587b..28dc6fabab3e 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_main.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_main.c
@@ -15,6 +15,8 @@
MODULE_DESCRIPTION(DRV_SUMMARY);
MODULE_IMPORT_NS("LIBETH");
+MODULE_IMPORT_NS("LIBIE_CP");
+MODULE_IMPORT_NS("LIBIE_PCI");
MODULE_IMPORT_NS("LIBETH_XDP");
MODULE_LICENSE("GPL");
@@ -56,8 +58,16 @@ static int idpf_get_device_type(struct pci_dev *pdev)
static int idpf_dev_init(struct idpf_adapter *adapter,
const struct pci_device_id *ent)
{
+ struct libie_mmio_info *mmio_info = &adapter->ctlq_ctx.mmio_info;
int ret;
+ ret = libie_pci_init_dev(adapter->pdev);
+ if (ret)
+ return ret;
+
+ mmio_info->pdev = adapter->pdev;
+ INIT_LIST_HEAD(&mmio_info->mmio_list);
+
if (ent->class == IDPF_CLASS_NETWORK_ETHERNET_PROGIF) {
ret = idpf_get_device_type(adapter->pdev);
switch (ret) {
@@ -90,6 +100,15 @@ static int idpf_dev_init(struct idpf_adapter *adapter,
return 0;
}
+/**
+ * idpf_decfg_device - deconfigure device and device specific resources
+ * @adapter: driver specific private structure
+ */
+static void idpf_decfg_device(struct idpf_adapter *adapter)
+{
+ libie_pci_unmap_all_mmio_regions(&adapter->ctlq_ctx.mmio_info);
+}
+
/**
* idpf_remove - Device removal routine
* @pdev: PCI device information struct
@@ -159,6 +178,7 @@ static void idpf_remove(struct pci_dev *pdev)
mutex_destroy(&adapter->queue_lock);
mutex_destroy(&adapter->vc_buf_lock);
+ idpf_decfg_device(adapter);
pci_set_drvdata(pdev, NULL);
kfree(adapter);
}
@@ -181,46 +201,45 @@ static void idpf_shutdown(struct pci_dev *pdev)
}
/**
- * idpf_cfg_hw - Initialize HW struct
- * @adapter: adapter to setup hw struct for
+ * idpf_cfg_device - configure device and device specific resources
+ * @adapter: driver specific private structure
*
- * Returns 0 on success, negative on failure
+ * Return: %0 on success, -%errno on failure.
*/
-static int idpf_cfg_hw(struct idpf_adapter *adapter)
+static int idpf_cfg_device(struct idpf_adapter *adapter)
{
- resource_size_t res_start, mbx_start, rstat_start;
+ struct libie_mmio_info *mmio_info = &adapter->ctlq_ctx.mmio_info;
struct pci_dev *pdev = adapter->pdev;
- struct idpf_hw *hw = &adapter->hw;
- struct device *dev = &pdev->dev;
- long len;
-
- res_start = pci_resource_start(pdev, 0);
+ struct resource *region;
+ bool mapped = false;
+ int err;
/* Map mailbox space for virtchnl communication */
- mbx_start = res_start + adapter->dev_ops.static_reg_info[0].start;
- len = resource_size(&adapter->dev_ops.static_reg_info[0]);
- hw->mbx.vaddr = devm_ioremap(dev, mbx_start, len);
- if (!hw->mbx.vaddr) {
- pci_err(pdev, "failed to allocate BAR0 mbx region\n");
-
+ region = &adapter->dev_ops.static_reg_info[0];
+ mapped = libie_pci_map_mmio_region(mmio_info, region->start,
+ resource_size(region));
+ if (!mapped) {
+ pci_err(pdev, "failed to map BAR0 mbx region\n");
return -ENOMEM;
}
- hw->mbx.addr_start = adapter->dev_ops.static_reg_info[0].start;
- hw->mbx.addr_len = len;
/* Map rstat space for resets */
- rstat_start = res_start + adapter->dev_ops.static_reg_info[1].start;
- len = resource_size(&adapter->dev_ops.static_reg_info[1]);
- hw->rstat.vaddr = devm_ioremap(dev, rstat_start, len);
- if (!hw->rstat.vaddr) {
- pci_err(pdev, "failed to allocate BAR0 rstat region\n");
+ region = &adapter->dev_ops.static_reg_info[1];
+ mapped = libie_pci_map_mmio_region(mmio_info, region->start,
+ resource_size(region));
+ if (!mapped) {
+ pci_err(pdev, "failed to map BAR0 rstat region\n");
+ libie_pci_unmap_all_mmio_regions(mmio_info);
return -ENOMEM;
}
- hw->rstat.addr_start = adapter->dev_ops.static_reg_info[1].start;
- hw->rstat.addr_len = len;
- hw->back = adapter;
+ err = pci_enable_ptm(pdev);
+ if (err)
+ pci_dbg(pdev, "PCIe PTM is not supported by PCIe bus/controller\n");
+
+ pci_set_drvdata(pdev, adapter);
+ adapter->hw.back = adapter;
return 0;
}
@@ -246,32 +265,21 @@ static int idpf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
adapter->req_rx_splitq = true;
adapter->pdev = pdev;
- err = pcim_enable_device(pdev);
- if (err)
- goto err_free;
- err = pcim_request_region(pdev, 0, pci_name(pdev));
+ err = idpf_dev_init(adapter, ent);
if (err) {
- pci_err(pdev, "pcim_request_region failed %pe\n", ERR_PTR(err));
-
+ dev_err(&pdev->dev, "Failed to initialize device (ID 0x%x): %d\n",
+ ent->device, err);
goto err_free;
}
- err = pci_enable_ptm(pdev);
- if (err)
- pci_dbg(pdev, "PCIe PTM is not supported by PCIe bus/controller\n");
-
- /* set up for high or low dma */
- err = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(64));
+ err = idpf_cfg_device(adapter);
if (err) {
- pci_err(pdev, "DMA configuration failed: %pe\n", ERR_PTR(err));
-
+ pci_err(pdev, "Failed to configure device specific resources: %pe\n",
+ ERR_PTR(err));
goto err_free;
}
- pci_set_master(pdev);
- pci_set_drvdata(pdev, adapter);
-
adapter->init_wq = alloc_workqueue("%s-%s-init",
WQ_UNBOUND | WQ_MEM_RECLAIM, 0,
dev_driver_string(dev),
@@ -279,7 +287,7 @@ static int idpf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
if (!adapter->init_wq) {
dev_err(dev, "Failed to allocate init workqueue\n");
err = -ENOMEM;
- goto err_free;
+ goto err_init_wq;
}
adapter->serv_wq = alloc_workqueue("%s-%s-service",
@@ -324,20 +332,6 @@ static int idpf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
/* setup msglvl */
adapter->msg_enable = netif_msg_init(-1, IDPF_AVAIL_NETIF_M);
- err = idpf_dev_init(adapter, ent);
- if (err) {
- dev_err(&pdev->dev, "Unexpected dev ID 0x%x in idpf probe\n",
- ent->device);
- goto destroy_vc_event_wq;
- }
-
- err = idpf_cfg_hw(adapter);
- if (err) {
- dev_err(dev, "Failed to configure HW structure for adapter: %d\n",
- err);
- goto destroy_vc_event_wq;
- }
-
mutex_init(&adapter->vport_ctrl_lock);
mutex_init(&adapter->vector_lock);
mutex_init(&adapter->queue_lock);
@@ -356,8 +350,6 @@ static int idpf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
return 0;
-destroy_vc_event_wq:
- destroy_workqueue(adapter->vc_event_wq);
err_vc_event_wq_alloc:
destroy_workqueue(adapter->stats_wq);
err_stats_wq_alloc:
@@ -366,6 +358,8 @@ static int idpf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
destroy_workqueue(adapter->serv_wq);
err_serv_wq_alloc:
destroy_workqueue(adapter->init_wq);
+err_init_wq:
+ idpf_decfg_device(adapter);
err_free:
kfree(adapter);
return err;
diff --git a/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c b/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
index a07d7e808ca9..98b8f678bd9a 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
@@ -15,31 +15,28 @@
static void idpf_vf_ctlq_reg_init(struct idpf_adapter *adapter,
struct idpf_ctlq_create_info *cq)
{
- resource_size_t mbx_start = adapter->dev_ops.static_reg_info[0].start;
- int i;
-
- for (i = 0; i < IDPF_NUM_DFLT_MBX_Q; i++) {
+ for (int i = 0; i < IDPF_NUM_DFLT_MBX_Q; i++) {
struct idpf_ctlq_create_info *ccq = cq + i;
switch (ccq->type) {
case IDPF_CTLQ_TYPE_MAILBOX_TX:
/* set head and tail registers in our local struct */
- ccq->reg.head = VF_ATQH - mbx_start;
- ccq->reg.tail = VF_ATQT - mbx_start;
- ccq->reg.len = VF_ATQLEN - mbx_start;
- ccq->reg.bah = VF_ATQBAH - mbx_start;
- ccq->reg.bal = VF_ATQBAL - mbx_start;
+ ccq->reg.head = VF_ATQH;
+ ccq->reg.tail = VF_ATQT;
+ ccq->reg.len = VF_ATQLEN;
+ ccq->reg.bah = VF_ATQBAH;
+ ccq->reg.bal = VF_ATQBAL;
ccq->reg.len_mask = VF_ATQLEN_ATQLEN_M;
ccq->reg.len_ena_mask = VF_ATQLEN_ATQENABLE_M;
ccq->reg.head_mask = VF_ATQH_ATQH_M;
break;
case IDPF_CTLQ_TYPE_MAILBOX_RX:
/* set head and tail registers in our local struct */
- ccq->reg.head = VF_ARQH - mbx_start;
- ccq->reg.tail = VF_ARQT - mbx_start;
- ccq->reg.len = VF_ARQLEN - mbx_start;
- ccq->reg.bah = VF_ARQBAH - mbx_start;
- ccq->reg.bal = VF_ARQBAL - mbx_start;
+ ccq->reg.head = VF_ARQH;
+ ccq->reg.tail = VF_ARQT;
+ ccq->reg.len = VF_ARQLEN;
+ ccq->reg.bah = VF_ARQBAH;
+ ccq->reg.bal = VF_ARQBAL;
ccq->reg.len_mask = VF_ARQLEN_ARQLEN_M;
ccq->reg.len_ena_mask = VF_ARQLEN_ARQENABLE_M;
ccq->reg.head_mask = VF_ARQH_ARQH_M;
@@ -56,13 +53,14 @@ static void idpf_vf_ctlq_reg_init(struct idpf_adapter *adapter,
*/
static void idpf_vf_mb_intr_reg_init(struct idpf_adapter *adapter)
{
+ struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
struct idpf_intr_reg *intr = &adapter->mb_vector.intr_reg;
u32 dyn_ctl = le32_to_cpu(adapter->caps.mailbox_dyn_ctl);
- intr->dyn_ctl = idpf_get_reg_addr(adapter, dyn_ctl);
+ intr->dyn_ctl = libie_pci_get_mmio_addr(mmio, dyn_ctl);
intr->dyn_ctl_intena_m = VF_INT_DYN_CTL0_INTENA_M;
intr->dyn_ctl_itridx_m = VF_INT_DYN_CTL0_ITR_INDX_M;
- intr->icr_ena = idpf_get_reg_addr(adapter, VF_INT_ICR0_ENA1);
+ intr->icr_ena = libie_pci_get_mmio_addr(mmio, VF_INT_ICR0_ENA1);
intr->icr_ena_ctlq_m = VF_INT_ICR0_ENA1_ADMINQ_M;
}
@@ -77,6 +75,7 @@ static int idpf_vf_intr_reg_init(struct idpf_vport *vport,
struct idpf_adapter *adapter = vport->adapter;
u16 num_vecs = rsrc->num_q_vectors;
struct idpf_vec_regs *reg_vals;
+ struct libie_mmio_info *mmio;
int num_regs, i, err = 0;
u32 rx_itr, tx_itr, val;
u16 total_vecs;
@@ -92,14 +91,17 @@ static int idpf_vf_intr_reg_init(struct idpf_vport *vport,
goto free_reg_vals;
}
+ mmio = &adapter->ctlq_ctx.mmio_info;
+
for (i = 0; i < num_vecs; i++) {
struct idpf_q_vector *q_vector = &rsrc->q_vectors[i];
u16 vec_id = rsrc->q_vector_idxs[i] - IDPF_MBX_Q_VEC;
struct idpf_intr_reg *intr = &q_vector->intr_reg;
+ struct idpf_vec_regs *reg = ®_vals[vec_id];
u32 spacing;
- intr->dyn_ctl = idpf_get_reg_addr(adapter,
- reg_vals[vec_id].dyn_ctl_reg);
+ intr->dyn_ctl = libie_pci_get_mmio_addr(mmio,
+ reg->dyn_ctl_reg);
intr->dyn_ctl_intena_m = VF_INT_DYN_CTLN_INTENA_M;
intr->dyn_ctl_intena_msk_m = VF_INT_DYN_CTLN_INTENA_MSK_M;
intr->dyn_ctl_itridx_s = VF_INT_DYN_CTLN_ITR_INDX_S;
@@ -109,22 +111,21 @@ static int idpf_vf_intr_reg_init(struct idpf_vport *vport,
intr->dyn_ctl_sw_itridx_ena_m =
VF_INT_DYN_CTLN_SW_ITR_INDX_ENA_M;
- spacing = IDPF_ITR_IDX_SPACING(reg_vals[vec_id].itrn_index_spacing,
+ spacing = IDPF_ITR_IDX_SPACING(reg->itrn_index_spacing,
IDPF_VF_ITR_IDX_SPACING);
rx_itr = VF_INT_ITRN_ADDR(VIRTCHNL2_ITR_IDX_0,
- reg_vals[vec_id].itrn_reg,
- spacing);
+ reg->itrn_reg, spacing);
tx_itr = VF_INT_ITRN_ADDR(VIRTCHNL2_ITR_IDX_1,
- reg_vals[vec_id].itrn_reg,
- spacing);
- intr->rx_itr = idpf_get_reg_addr(adapter, rx_itr);
- intr->tx_itr = idpf_get_reg_addr(adapter, tx_itr);
+ reg->itrn_reg, spacing);
+ intr->rx_itr = libie_pci_get_mmio_addr(mmio, rx_itr);
+ intr->tx_itr = libie_pci_get_mmio_addr(mmio, tx_itr);
}
/* Data vector for NOIRQ queues */
val = reg_vals[rsrc->q_vector_idxs[i] - IDPF_MBX_Q_VEC].dyn_ctl_reg;
- rsrc->noirq_dyn_ctl = idpf_get_reg_addr(adapter, val);
+ rsrc->noirq_dyn_ctl =
+ libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info, val);
val = VF_INT_DYN_CTLN_WB_ON_ITR_M | VF_INT_DYN_CTLN_INTENA_MSK_M |
FIELD_PREP(VF_INT_DYN_CTLN_ITR_INDX_M, IDPF_NO_ITR_UPDATE_IDX);
@@ -142,7 +143,9 @@ static int idpf_vf_intr_reg_init(struct idpf_vport *vport,
*/
static void idpf_vf_reset_reg_init(struct idpf_adapter *adapter)
{
- adapter->reset_reg.rstat = idpf_get_rstat_reg_addr(adapter, VFGEN_RSTAT);
+ adapter->reset_reg.rstat =
+ libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info,
+ VFGEN_RSTAT);
adapter->reset_reg.rstat_m = VFGEN_RSTAT_VFR_STATE_M;
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
index a3c17f0e14f3..357ff3e86b2e 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
@@ -2,6 +2,7 @@
/* Copyright (C) 2023 Intel Corporation */
#include <linux/export.h>
+#include <linux/net/intel/libie/pci.h>
#include <net/libeth/rx.h>
#include "idpf.h"
@@ -1020,12 +1021,46 @@ static int idpf_send_get_caps_msg(struct idpf_adapter *adapter)
}
/**
- * idpf_send_get_lan_memory_regions - Send virtchnl get LAN memory regions msg
+ * idpf_mmio_region_non_static - Check if region is not static
+ * @mmio_info: PCI resources info
+ * @reg: region to check
+ *
+ * Return: %true if region can be received though virtchnl command,
+ * %false if region is related to mailbox or resetting
+ */
+static bool idpf_mmio_region_non_static(struct libie_mmio_info *mmio_info,
+ struct libie_pci_mmio_region *reg)
+{
+ struct idpf_adapter *adapter =
+ container_of(mmio_info, struct idpf_adapter,
+ ctlq_ctx.mmio_info);
+
+ for (uint i = 0; i < IDPF_MMIO_REG_NUM_STATIC; i++) {
+ if (reg->bar_idx == 0 &&
+ reg->offset == adapter->dev_ops.static_reg_info[i].start)
+ return false;
+ }
+
+ return true;
+}
+
+/**
+ * idpf_decfg_lan_memory_regions - Unmap non-static memory regions
+ * @adapter: Driver specific private structure
+ */
+static void idpf_decfg_lan_memory_regions(struct idpf_adapter *adapter)
+{
+ libie_pci_unmap_fltr_regs(&adapter->ctlq_ctx.mmio_info,
+ idpf_mmio_region_non_static);
+}
+
+/**
+ * idpf_cfg_lan_memory_regions - Send virtchnl get LAN memory regions msg
* @adapter: Driver specific private struct
*
* Return: 0 on success or error code on failure.
*/
-static int idpf_send_get_lan_memory_regions(struct idpf_adapter *adapter)
+static int idpf_cfg_lan_memory_regions(struct idpf_adapter *adapter)
{
struct virtchnl2_get_lan_memory_regions *rcvd_regions __free(kfree);
struct idpf_vc_xn_params xn_params = {
@@ -1037,7 +1072,6 @@ static int idpf_send_get_lan_memory_regions(struct idpf_adapter *adapter)
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
int num_regions, size;
- struct idpf_hw *hw;
ssize_t reply_sz;
int err = 0;
@@ -1060,86 +1094,56 @@ static int idpf_send_get_lan_memory_regions(struct idpf_adapter *adapter)
if (size > IDPF_CTLQ_MAX_BUF_LEN)
return -EINVAL;
- hw = &adapter->hw;
- hw->lan_regs = kzalloc_objs(*hw->lan_regs, num_regions);
- if (!hw->lan_regs)
- return -ENOMEM;
-
for (int i = 0; i < num_regions; i++) {
- hw->lan_regs[i].addr_len =
- le64_to_cpu(rcvd_regions->mem_reg[i].size);
- hw->lan_regs[i].addr_start =
- le64_to_cpu(rcvd_regions->mem_reg[i].start_offset);
+ struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
+ resource_size_t offset, len;
+
+ offset = le64_to_cpu(rcvd_regions->mem_reg[i].start_offset);
+ len = le64_to_cpu(rcvd_regions->mem_reg[i].size);
+ if (len && !libie_pci_map_mmio_region(mmio, offset, len)) {
+ idpf_decfg_lan_memory_regions(adapter);
+ return -EIO;
+ }
}
- hw->num_lan_regs = num_regions;
return err;
}
/**
- * idpf_calc_remaining_mmio_regs - calculate MMIO regions outside mbx and rstat
+ * idpf_map_remaining_mmio_regs - map MMIO regions outside mbx and rstat
* @adapter: Driver specific private structure
*
- * Called when idpf_send_get_lan_memory_regions is not supported. This will
+ * Called when idpf_cfg_lan_memory_regions is not supported. This will
* calculate the offsets and sizes for the regions before, in between, and
* after the mailbox and rstat MMIO mappings.
*
* Return: 0 on success or error code on failure.
*/
-static int idpf_calc_remaining_mmio_regs(struct idpf_adapter *adapter)
+static int idpf_map_remaining_mmio_regs(struct idpf_adapter *adapter)
{
struct resource *rstat_reg = &adapter->dev_ops.static_reg_info[1];
struct resource *mbx_reg = &adapter->dev_ops.static_reg_info[0];
- struct idpf_hw *hw = &adapter->hw;
-
- hw->num_lan_regs = IDPF_MMIO_MAP_FALLBACK_MAX_REMAINING;
- hw->lan_regs = kzalloc_objs(*hw->lan_regs, hw->num_lan_regs);
- if (!hw->lan_regs)
- return -ENOMEM;
+ struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
+ resource_size_t reg_start, size;
+ bool ok = true;
/* Region preceding mailbox */
- hw->lan_regs[0].addr_start = 0;
- hw->lan_regs[0].addr_len = mbx_reg->start;
- /* Region between mailbox and rstat */
- hw->lan_regs[1].addr_start = mbx_reg->end + 1;
- hw->lan_regs[1].addr_len = rstat_reg->start -
- hw->lan_regs[1].addr_start;
- /* Region after rstat */
- hw->lan_regs[2].addr_start = rstat_reg->end + 1;
- hw->lan_regs[2].addr_len = pci_resource_len(adapter->pdev, 0) -
- hw->lan_regs[2].addr_start;
-
- return 0;
-}
+ size = mbx_reg->start;
+ ok &= !size || libie_pci_map_mmio_region(mmio, 0, size);
-/**
- * idpf_map_lan_mmio_regs - map remaining LAN BAR regions
- * @adapter: Driver specific private structure
- *
- * Return: 0 on success or error code on failure.
- */
-static int idpf_map_lan_mmio_regs(struct idpf_adapter *adapter)
-{
- struct pci_dev *pdev = adapter->pdev;
- struct idpf_hw *hw = &adapter->hw;
- resource_size_t res_start;
-
- res_start = pci_resource_start(pdev, 0);
-
- for (int i = 0; i < hw->num_lan_regs; i++) {
- resource_size_t start;
- long len;
+ /* Region between mailbox and rstat */
+ reg_start = mbx_reg->end + 1;
+ size = rstat_reg->start - reg_start;
+ ok &= !size || libie_pci_map_mmio_region(mmio, reg_start, size);
- len = hw->lan_regs[i].addr_len;
- if (!len)
- continue;
- start = hw->lan_regs[i].addr_start + res_start;
+ /* Region after rstat */
+ reg_start = rstat_reg->end + 1;
+ size = pci_resource_len(adapter->pdev, 0) - reg_start;
+ ok &= !size || libie_pci_map_mmio_region(mmio, reg_start, size);
- hw->lan_regs[i].vaddr = devm_ioremap(&pdev->dev, start, len);
- if (!hw->lan_regs[i].vaddr) {
- pci_err(pdev, "failed to allocate BAR0 region\n");
- return -ENOMEM;
- }
+ if (!ok) {
+ idpf_decfg_lan_memory_regions(adapter);
+ return -ENOMEM;
}
return 0;
@@ -1413,7 +1417,7 @@ static int __idpf_queue_reg_init(struct idpf_vport *vport,
struct idpf_q_vec_rsrc *rsrc, u32 *reg_vals,
int num_regs, u32 q_type)
{
- struct idpf_adapter *adapter = vport->adapter;
+ struct libie_mmio_info *mmio = &vport->adapter->ctlq_ctx.mmio_info;
int i, j, k = 0;
switch (q_type) {
@@ -1423,7 +1427,8 @@ static int __idpf_queue_reg_init(struct idpf_vport *vport,
for (j = 0; j < tx_qgrp->num_txq && k < num_regs; j++, k++)
tx_qgrp->txqs[j]->tail =
- idpf_get_reg_addr(adapter, reg_vals[k]);
+ libie_pci_get_mmio_addr(mmio,
+ reg_vals[k]);
}
break;
case VIRTCHNL2_QUEUE_TYPE_RX:
@@ -1435,8 +1440,8 @@ static int __idpf_queue_reg_init(struct idpf_vport *vport,
struct idpf_rx_queue *q;
q = rx_qgrp->singleq.rxqs[j];
- q->tail = idpf_get_reg_addr(adapter,
- reg_vals[k]);
+ q->tail = libie_pci_get_mmio_addr(mmio,
+ reg_vals[k]);
}
}
break;
@@ -1449,8 +1454,8 @@ static int __idpf_queue_reg_init(struct idpf_vport *vport,
struct idpf_buf_queue *q;
q = &rx_qgrp->splitq.bufq_sets[j].bufq;
- q->tail = idpf_get_reg_addr(adapter,
- reg_vals[k]);
+ q->tail = libie_pci_get_mmio_addr(mmio,
+ reg_vals[k]);
}
}
break;
@@ -3445,35 +3450,30 @@ int idpf_vc_core_init(struct idpf_adapter *adapter)
}
if (idpf_is_cap_ena(adapter, IDPF_OTHER_CAPS, VIRTCHNL2_CAP_LAN_MEMORY_REGIONS)) {
- err = idpf_send_get_lan_memory_regions(adapter);
+ err = idpf_cfg_lan_memory_regions(adapter);
if (err) {
- dev_err(&adapter->pdev->dev, "Failed to get LAN memory regions: %d\n",
+ dev_err(&adapter->pdev->dev, "Failed to configure LAN memory regions: %d\n",
err);
return -EINVAL;
}
} else {
/* Fallback to mapping the remaining regions of the entire BAR */
- err = idpf_calc_remaining_mmio_regs(adapter);
+ err = idpf_map_remaining_mmio_regs(adapter);
if (err) {
- dev_err(&adapter->pdev->dev, "Failed to allocate BAR0 region(s): %d\n",
+ dev_err(&adapter->pdev->dev, "Failed to configure BAR0 region(s): %d\n",
err);
- return -ENOMEM;
+ return err;
}
}
- err = idpf_map_lan_mmio_regs(adapter);
- if (err) {
- dev_err(&adapter->pdev->dev, "Failed to map BAR0 region(s): %d\n",
- err);
- return -ENOMEM;
- }
-
pci_sriov_set_totalvfs(adapter->pdev, idpf_get_max_vfs(adapter));
num_max_vports = idpf_get_max_vports(adapter);
adapter->max_vports = num_max_vports;
adapter->vports = kzalloc_objs(*adapter->vports, num_max_vports);
- if (!adapter->vports)
- return -ENOMEM;
+ if (!adapter->vports) {
+ err = -ENOMEM;
+ goto decfg_regions;
+ }
if (!adapter->netdevs) {
adapter->netdevs = kzalloc_objs(struct net_device *,
@@ -3539,6 +3539,8 @@ int idpf_vc_core_init(struct idpf_adapter *adapter)
err_netdev_alloc:
kfree(adapter->vports);
adapter->vports = NULL;
+decfg_regions:
+ idpf_decfg_lan_memory_regions(adapter);
return err;
init_failed:
@@ -3572,7 +3574,6 @@ int idpf_vc_core_init(struct idpf_adapter *adapter)
*/
void idpf_vc_core_deinit(struct idpf_adapter *adapter)
{
- struct idpf_hw *hw = &adapter->hw;
bool remove_in_prog;
if (!test_bit(IDPF_VC_CORE_INIT, adapter->flags))
@@ -3597,12 +3598,10 @@ void idpf_vc_core_deinit(struct idpf_adapter *adapter)
idpf_vport_params_buf_rel(adapter);
- kfree(hw->lan_regs);
- hw->lan_regs = NULL;
-
kfree(adapter->vports);
adapter->vports = NULL;
+ idpf_decfg_lan_memory_regions(adapter);
clear_bit(IDPF_VC_CORE_INIT, adapter->flags);
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
index d9bcc3f61c65..8d8fb498e092 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
@@ -31,6 +31,7 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
struct virtchnl2_ptp_cross_time_reg_offsets cross_tstamp_offsets;
+ struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
struct virtchnl2_ptp_clk_adj_reg_offsets clk_adj_offsets;
struct virtchnl2_ptp_clk_reg_offsets clock_offsets;
struct idpf_ptp_secondary_mbx *scnd_mbx;
@@ -76,19 +77,20 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
clock_offsets = recv_ptp_caps_msg->clk_offsets;
temp_offset = le32_to_cpu(clock_offsets.dev_clk_ns_l);
- ptp->dev_clk_regs.dev_clk_ns_l = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.dev_clk_ns_l =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clock_offsets.dev_clk_ns_h);
- ptp->dev_clk_regs.dev_clk_ns_h = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.dev_clk_ns_h =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clock_offsets.phy_clk_ns_l);
- ptp->dev_clk_regs.phy_clk_ns_l = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.phy_clk_ns_l =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clock_offsets.phy_clk_ns_h);
- ptp->dev_clk_regs.phy_clk_ns_h = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.phy_clk_ns_h =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clock_offsets.cmd_sync_trigger);
- ptp->dev_clk_regs.cmd_sync = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.cmd_sync =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
cross_tstamp:
access_type = ptp->get_cross_tstamp_access;
@@ -98,13 +100,14 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
cross_tstamp_offsets = recv_ptp_caps_msg->cross_time_offsets;
temp_offset = le32_to_cpu(cross_tstamp_offsets.sys_time_ns_l);
- ptp->dev_clk_regs.sys_time_ns_l = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.sys_time_ns_l =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(cross_tstamp_offsets.sys_time_ns_h);
- ptp->dev_clk_regs.sys_time_ns_h = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.sys_time_ns_h =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(cross_tstamp_offsets.cmd_sync_trigger);
- ptp->dev_clk_regs.cmd_sync = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.cmd_sync =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
discipline_clock:
access_type = ptp->adj_dev_clk_time_access;
@@ -115,29 +118,32 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
/* Device clock offsets */
temp_offset = le32_to_cpu(clk_adj_offsets.dev_clk_cmd_type);
- ptp->dev_clk_regs.cmd = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.cmd = libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.dev_clk_incval_l);
- ptp->dev_clk_regs.incval_l = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.incval_l = libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.dev_clk_incval_h);
- ptp->dev_clk_regs.incval_h = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.incval_h = libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.dev_clk_shadj_l);
- ptp->dev_clk_regs.shadj_l = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.shadj_l = libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.dev_clk_shadj_h);
- ptp->dev_clk_regs.shadj_h = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.shadj_h = libie_pci_get_mmio_addr(mmio, temp_offset);
/* PHY clock offsets */
temp_offset = le32_to_cpu(clk_adj_offsets.phy_clk_cmd_type);
- ptp->dev_clk_regs.phy_cmd = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.phy_cmd =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.phy_clk_incval_l);
- ptp->dev_clk_regs.phy_incval_l = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.phy_incval_l =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.phy_clk_incval_h);
- ptp->dev_clk_regs.phy_incval_h = idpf_get_reg_addr(adapter,
- temp_offset);
+ ptp->dev_clk_regs.phy_incval_h =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.phy_clk_shadj_l);
- ptp->dev_clk_regs.phy_shadj_l = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.phy_shadj_l =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
temp_offset = le32_to_cpu(clk_adj_offsets.phy_clk_shadj_h);
- ptp->dev_clk_regs.phy_shadj_h = idpf_get_reg_addr(adapter, temp_offset);
+ ptp->dev_clk_regs.phy_shadj_h =
+ libie_pci_get_mmio_addr(mmio, temp_offset);
return 0;
}
--
2.47.1
^ permalink raw reply related
* [PATCH net-next v5 10/15] idpf: make mbx_task queueing and cancelling more consistent
From: Tony Nguyen @ 2026-07-15 18:00 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, netdev
Cc: Larysa Zaremba, anthony.l.nguyen, przemyslaw.kitszel,
aleksander.lobakin, sridhar.samudrala, michal.swiatkowski,
maciej.fijalkowski, emil.s.tantilov, madhu.chittim, joshua.a.hay,
jacob.e.keller, jayaprakash.shanmugam, jiri, horms, corbet,
richardcochran, skhan, linux-doc, Samuel Salin
In-Reply-To: <20260715180042.1972010-1-anthony.l.nguyen@intel.com>
From: Larysa Zaremba <larysa.zaremba@intel.com>
One of the assumptions of libie_cp and pre-refactor idpf control queue
handling is such that all Rx processing is handled by a single task, which
is to be cancelled before the mailbox destruction. Aside from cancelling,
it is also important to make sure that idpf_intr_rel() never reschedules
it afterwards.
In order to comply, in the init path, do the first queueing of mbx_task in
idpf_init_dflt_mbx(), and in deinit and reset, always cancel the task in
idpf_deinit_dflt_mbx(), in every single flow call idpf_mb_intr_rel_irq()
beforehand.
Reviewed-by: Emil Tantilov <emil.s.tantilov@intel.com>
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Tested-by: Samuel Salin <Samuel.salin@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/idpf/idpf.h | 1 +
drivers/net/ethernet/intel/idpf/idpf_lib.c | 9 ++++-----
drivers/net/ethernet/intel/idpf/idpf_virtchnl.c | 5 +++++
3 files changed, 10 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/intel/idpf/idpf.h b/drivers/net/ethernet/intel/idpf/idpf.h
index d7d751e2a781..470bc23c844c 100644
--- a/drivers/net/ethernet/intel/idpf/idpf.h
+++ b/drivers/net/ethernet/intel/idpf/idpf.h
@@ -984,6 +984,7 @@ void idpf_vc_event_task(struct work_struct *work);
void idpf_dev_ops_init(struct idpf_adapter *adapter);
void idpf_vf_dev_ops_init(struct idpf_adapter *adapter);
int idpf_intr_req(struct idpf_adapter *adapter);
+void idpf_mb_intr_rel_irq(struct idpf_adapter *adapter);
void idpf_intr_rel(struct idpf_adapter *adapter);
u16 idpf_get_max_tx_hdr_size(struct idpf_adapter *adapter);
int idpf_initiate_soft_reset(struct idpf_vport *vport,
diff --git a/drivers/net/ethernet/intel/idpf/idpf_lib.c b/drivers/net/ethernet/intel/idpf/idpf_lib.c
index 6751534e022b..1c19cd1f9dd1 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_lib.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_lib.c
@@ -68,9 +68,11 @@ static void idpf_deinit_vector_stack(struct idpf_adapter *adapter)
* This will also disable interrupt mode and queue up mailbox task. Mailbox
* task will reschedule itself if not in interrupt mode.
*/
-static void idpf_mb_intr_rel_irq(struct idpf_adapter *adapter)
+void idpf_mb_intr_rel_irq(struct idpf_adapter *adapter)
{
- clear_bit(IDPF_MB_INTR_MODE, adapter->flags);
+ if (!test_and_clear_bit(IDPF_MB_INTR_MODE, adapter->flags))
+ return;
+
kfree(free_irq(adapter->msix_entries[0].vector, adapter));
queue_delayed_work(adapter->mbx_wq, &adapter->mbx_task, 0);
}
@@ -1939,14 +1941,11 @@ static void idpf_init_hard_reset(struct idpf_adapter *adapter)
goto unlock_mutex;
}
- queue_delayed_work(adapter->mbx_wq, &adapter->mbx_task, 0);
-
/* Initialize the state machine, also allocate memory and request
* resources
*/
err = idpf_vc_core_init(adapter);
if (err) {
- cancel_delayed_work_sync(&adapter->mbx_task);
idpf_deinit_dflt_mbx(adapter);
goto unlock_mutex;
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
index 84155d683773..7c166651f90d 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
@@ -2931,6 +2931,8 @@ int idpf_init_dflt_mbx(struct idpf_adapter *adapter)
adapter->xnm = params.xnm;
adapter->state = __IDPF_VER_CHECK;
+ queue_delayed_work(adapter->mbx_wq, &adapter->mbx_task, 0);
+
return 0;
}
@@ -2940,6 +2942,9 @@ int idpf_init_dflt_mbx(struct idpf_adapter *adapter)
*/
void idpf_deinit_dflt_mbx(struct idpf_adapter *adapter)
{
+ idpf_mb_intr_rel_irq(adapter);
+ cancel_delayed_work_sync(&adapter->mbx_task);
+
if (adapter->xnm) {
idpf_mb_clean(adapter, adapter->asq, true);
libie_ctlq_xn_deinit(adapter->xnm, &adapter->ctlq_ctx);
--
2.47.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox