[PATCH for-next 0/8] IB/hfi1: patches for next 08/13/2017

[PATCH for-next 0/8] IB/hfi1: patches for next 08/13/2017

[Dennis Dalessandro]

Hi Doug,

Here is the next set of patches for our drivers for 4.14. One of these patches
gets rid of a config option that is no longer needed. There are some cleanups
as usual and minor bug fixes. Nothing that I would say needs to go into an RC.
These can all just land in your for-next tree and apply on top of that branch
with the addition of the pending patches still on the list. Everything applied
cleanly.

There is one patch to an OPA header file that addresses a question that Leon
had on the list about OPA physical port states.

Patches can can also be found in my GitHub repo at:
https://github.com/ddalessa/kernel/tree/for-4.14

---

Dennis Dalessandro (1):
      IB/hfi1: Document phys port state bits not used in IB

Grzegorz Morys (1):
      IB/hfi1: Remove HFI1_VERBS_31BIT_PSN option

Jakub Byczkowski (3):
      IB/hfi1: Add flag for platform config scratch register read
      IB/hfi1: Load fallback platform configuration per HFI device
      IB/hfi1: Remove pstate from hfi1_pportdata

Kaike Wan (1):
      IB/hfi1: Add kernel receive context info to debugfs

Kamenee Arumugame (1):
      IB/hfi1: Stricter bounds checking of MAD trap index

Sebastian Sanchez (1):
      IB/hfi1: Check xchg returned value for queuing link down entry


 drivers/infiniband/hw/hfi1/Kconfig      |    7 ---
 drivers/infiniband/hw/hfi1/chip.c       |   72 +++++++++++++++++++------------
 drivers/infiniband/hw/hfi1/chip.h       |    4 --
 drivers/infiniband/hw/hfi1/debugfs.c    |   47 ++++++++++++++++++++
 drivers/infiniband/hw/hfi1/driver.c     |   59 +++++++++++++++++++++++++
 drivers/infiniband/hw/hfi1/firmware.c   |   57 +++----------------------
 drivers/infiniband/hw/hfi1/hfi.h        |   28 ++----------
 drivers/infiniband/hw/hfi1/mad.c        |   15 ++++++
 drivers/infiniband/hw/hfi1/opa_compat.h |   21 ++++++++-
 drivers/infiniband/hw/hfi1/platform.c   |   43 +++++++++++++------
 drivers/infiniband/hw/hfi1/verbs.h      |    5 --
 include/rdma/rdma_vt.h                  |    2 -
 12 files changed, 224 insertions(+), 136 deletions(-)

--
-Denny
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH for-next 1/8] IB/hfi1: Check xchg returned value for queuing link down entry

[Dennis Dalessandro]

From: Sebastian Sanchez <sebastian.sanchez-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Check xchg returned value for queuing link down entry
to guarantee proper atomic value reads.

Fixes: 626c077c025f ("IB/hfi1: Prevent link down request double queuing")
Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/chip.c |   18 ++++++++++++------
 1 files changed, 12 insertions(+), 6 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c
index 24fda52..a6c6534 100644
--- a/drivers/infiniband/hw/hfi1/chip.c
+++ b/drivers/infiniband/hw/hfi1/chip.c
@@ -7810,16 +7810,22 @@ static void handle_8051_interrupt(struct hfi1_devdata *dd, u32 unused, u64 reg)
 	if (queue_link_down) {
 		/*
 		 * if the link is already going down or disabled, do not
-		 * queue another
+		 * queue another. If there's a link down entry already
+		 * queued, don't queue another one.
 		 */
 		if ((ppd->host_link_state &
 		    (HLS_GOING_OFFLINE | HLS_LINK_COOLDOWN)) ||
-		    ppd->link_enabled == 0 || ppd->is_link_down_queued) {
-			dd_dev_info(dd, "%s: not queuing link down\n",
-				    __func__);
+		    ppd->link_enabled == 0) {
+			dd_dev_info(dd, "%s: not queuing link down. host_link_state %x, link_enabled %x\n",
+				    __func__, ppd->host_link_state,
+				    ppd->link_enabled);
 		} else {
-			xchg(&ppd->is_link_down_queued, 1);
-			queue_work(ppd->link_wq, &ppd->link_down_work);
+			if (xchg(&ppd->is_link_down_queued, 1) == 1)
+				dd_dev_info(dd,
+					    "%s: link down request already queued\n",
+					    __func__);
+			else
+				queue_work(ppd->link_wq, &ppd->link_down_work);
 		}
 	}
 }

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH for-next 2/8] IB/hfi1: Document phys port state bits not used in IB

[Dennis Dalessandro]

A couple bits are used by OPA for link physical state that are not present
as part of InfiniBand. Add a short blurb what those states mean and removed
an unused state.

Cc: Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Reviewed-by: Todd Rimmer <todd.rimmer-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Reviewed-by: Brent Rothermel <brent.r.rothermel-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/opa_compat.h |   21 +++++++++++++++++++--
 1 files changed, 19 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/opa_compat.h b/drivers/infiniband/hw/hfi1/opa_compat.h
index 6ef3c1c..774215b 100644
--- a/drivers/infiniband/hw/hfi1/opa_compat.h
+++ b/drivers/infiniband/hw/hfi1/opa_compat.h
@@ -84,7 +84,8 @@ static inline u8 port_states_to_phys_state(struct opa_port_states *ps)
 /*
  * OPA port physical states
  * IB Volume 1, Table 146 PortInfo/IB Volume 2 Section 5.4.2(1) PortPhysState
- * values.
+ * values are the same in OmniPath Architecture. OPA leverages some of the same
+ * concepts as InfiniBand, but has a few other states as well.
  *
  * When writing, only values 0-3 are valid, other values are ignored.
  * When reading, 0 is reserved.
@@ -92,6 +93,8 @@ static inline u8 port_states_to_phys_state(struct opa_port_states *ps)
  * Returned by the ibphys_portstate() routine.
  */
 enum opa_port_phys_state {
+	/* Values 0-7 have the same meaning in OPA as in InfiniBand. */
+
 	IB_PORTPHYSSTATE_NOP = 0,
 	/* 1 is reserved */
 	IB_PORTPHYSSTATE_POLLING = 2,
@@ -101,9 +104,23 @@ enum opa_port_phys_state {
 	IB_PORTPHYSSTATE_LINK_ERROR_RECOVERY = 6,
 	IB_PORTPHYSSTATE_PHY_TEST = 7,
 	/* 8 is reserved */
+
+	/*
+	 * Offline: Port is quiet (transmitters disabled) due to lack of
+	 * physical media, unsupported media, or transition between link up
+	 * and next link up attempt
+	 */
 	OPA_PORTPHYSSTATE_OFFLINE = 9,
-	OPA_PORTPHYSSTATE_GANGED = 10,
+
+	/* 10 is reserved */
+
+	/*
+	 * Phy_Test: Specific test patterns are transmitted, and receiver BER
+	 * can be monitored. This facilitates signal integrity testing for the
+	 * physical layer of the port.
+	 */
 	OPA_PORTPHYSSTATE_TEST = 11,
+
 	OPA_PORTPHYSSTATE_MAX = 11,
 	/* values 12-15 are reserved/ignored */
 };

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH for-next 2/8] IB/hfi1: Document phys port state bits not used in IB

[Leon Romanovsky]

[-- Attachment #1: Type: text/plain, Size: 802 bytes --]

On Sun, Aug 13, 2017 at 08:08:28AM -0700, Dennis Dalessandro wrote:
> A couple bits are used by OPA for link physical state that are not present
> as part of InfiniBand. Add a short blurb what those states mean and removed
> an unused state.
>
> Cc: Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
> Reviewed-by: Todd Rimmer <todd.rimmer-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Reviewed-by: Brent Rothermel <brent.r.rothermel-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> ---
>  drivers/infiniband/hw/hfi1/opa_compat.h |   21 +++++++++++++++++++--
>  1 files changed, 19 insertions(+), 2 deletions(-)
>

Thanks,
Reviewed-by: Leon Romanovsky <leonro-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[PATCH for-next 3/8] IB/hfi1: Add flag for platform config scratch register read

[Dennis Dalessandro]

From: Jakub Byczkowski <jakub.byczkowski-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Add flag in pport data structure to determine when platform config was
read from scratch registers. Change conditions in parse_platform_config
and get_platform_config_field to use the new flag.

Reviewed-by: Easwar Hariharan <easwar.hariharan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Reviewed-by: Ira Weiny <ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Jakub Byczkowski <jakub.byczkowski-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/firmware.c |    6 ++++--
 drivers/infiniband/hw/hfi1/hfi.h      |    3 +++
 drivers/infiniband/hw/hfi1/platform.c |    2 ++
 3 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/firmware.c b/drivers/infiniband/hw/hfi1/firmware.c
index 885714b..fb095e6 100644
--- a/drivers/infiniband/hw/hfi1/firmware.c
+++ b/drivers/infiniband/hw/hfi1/firmware.c
@@ -1789,6 +1789,7 @@ static int check_meta_version(struct hfi1_devdata *dd, u32 *system_table)
 int parse_platform_config(struct hfi1_devdata *dd)
 {
 	struct platform_config_cache *pcfgcache = &dd->pcfg_cache;
+	struct hfi1_pportdata *ppd = dd->pport;
 	u32 *ptr = NULL;
 	u32 header1 = 0, header2 = 0, magic_num = 0, crc = 0, file_length = 0;
 	u32 record_idx = 0, table_type = 0, table_length_dwords = 0;
@@ -1800,7 +1801,7 @@ int parse_platform_config(struct hfi1_devdata *dd)
 	 * scratch register bitmap, thus there is no platform config to parse.
 	 * Skip parsing in these situations.
 	 */
-	if (is_integrated(dd) && !platform_config_load)
+	if (ppd->config_from_scratch)
 		return 0;
 
 	if (!dd->platform_config.data) {
@@ -2089,13 +2090,14 @@ int get_platform_config_field(struct hfi1_devdata *dd,
 	int ret = 0, wlen = 0, seek = 0;
 	u32 field_len_bits = 0, field_start_bits = 0, *src_ptr = NULL;
 	struct platform_config_cache *pcfgcache = &dd->pcfg_cache;
+	struct hfi1_pportdata *ppd = dd->pport;
 
 	if (data)
 		memset(data, 0, len);
 	else
 		return -EINVAL;
 
-	if (is_integrated(dd) && !platform_config_load) {
+	if (ppd->config_from_scratch) {
 		/*
 		 * Use saved configuration from ppd for integrated platforms
 		 */
diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index dff3d3f..181feca 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -677,6 +677,9 @@ struct hfi1_pportdata {
 	u8  default_atten;
 	u8  max_power_class;
 
+	/* did we read platform config from scratch registers? */
+	bool config_from_scratch;
+
 	/* GUIDs for this interface, in host order, guids[0] is a port guid */
 	u64 guids[HFI1_GUIDS_PER_PORT];
 
diff --git a/drivers/infiniband/hw/hfi1/platform.c b/drivers/infiniband/hw/hfi1/platform.c
index 5c38a24..ab221b0 100644
--- a/drivers/infiniband/hw/hfi1/platform.c
+++ b/drivers/infiniband/hw/hfi1/platform.c
@@ -138,6 +138,8 @@ static void save_platform_config_fields(struct hfi1_devdata *dd)
 
 	ppd->max_power_class = (temp_scratch & QSFP_MAX_POWER_SMASK) >>
 				QSFP_MAX_POWER_SHIFT;
+
+	ppd->config_from_scratch = true;
 }
 
 void get_platform_config(struct hfi1_devdata *dd)

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH for-next 4/8] IB/hfi1: Load fallback platform configuration per HFI device

[Dennis Dalessandro]

From: Jakub Byczkowski <jakub.byczkowski-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Currently fallback configuration is loaded once per driver instance.
With multiple HFI devices in the same system the current code may not
load the platform config data for the device. Change fallback platform
config data loading to be per device.

Reviewed-by: Easwar Hariharan <easwar.hariharan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Reviewed-by: Ira Weiny <ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Jakub Byczkowski <jakub.byczkowski-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/chip.h     |    1 -
 drivers/infiniband/hw/hfi1/firmware.c |   51 ++-------------------------------
 drivers/infiniband/hw/hfi1/platform.c |   41 ++++++++++++++++++---------
 3 files changed, 31 insertions(+), 62 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/chip.h b/drivers/infiniband/hw/hfi1/chip.h
index f708302..fc21097 100644
--- a/drivers/infiniband/hw/hfi1/chip.h
+++ b/drivers/infiniband/hw/hfi1/chip.h
@@ -648,7 +648,6 @@ u64 create_pbc(struct hfi1_pportdata *ppd, u64 flags, int srate_mbs, u32 vl,
 #define NUM_PCIE_SERDES 16	/* number of PCIe serdes on the SBus */
 extern const u8 pcie_serdes_broadcast[];
 extern const u8 pcie_pcs_addrs[2][NUM_PCIE_SERDES];
-extern uint platform_config_load;
 
 /* SBus commands */
 #define RESET_SBUS_RECEIVER 0x20
diff --git a/drivers/infiniband/hw/hfi1/firmware.c b/drivers/infiniband/hw/hfi1/firmware.c
index fb095e6..5aea8f4 100644
--- a/drivers/infiniband/hw/hfi1/firmware.c
+++ b/drivers/infiniband/hw/hfi1/firmware.c
@@ -64,7 +64,6 @@
 #define DEFAULT_FW_FABRIC_NAME "hfi1_fabric.fw"
 #define DEFAULT_FW_SBUS_NAME "hfi1_sbus.fw"
 #define DEFAULT_FW_PCIE_NAME "hfi1_pcie.fw"
-#define DEFAULT_PLATFORM_CONFIG_NAME "hfi1_platform.dat"
 #define ALT_FW_8051_NAME_ASIC "hfi1_dc8051_d.fw"
 #define ALT_FW_FABRIC_NAME "hfi1_fabric_d.fw"
 #define ALT_FW_SBUS_NAME "hfi1_sbus_d.fw"
@@ -76,19 +75,11 @@
 static uint fw_pcie_serdes_load = 1;
 static uint fw_sbus_load = 1;
 
-/*
- * Access required in platform.c
- * Maintains state of whether the platform config was fetched via the
- * fallback option
- */
-uint platform_config_load;
-
 /* Firmware file names get set in hfi1_firmware_init() based on the above */
 static char *fw_8051_name;
 static char *fw_fabric_serdes_name;
 static char *fw_sbus_name;
 static char *fw_pcie_serdes_name;
-static char *platform_config_name;
 
 #define SBUS_MAX_POLL_COUNT 100
 #define SBUS_COUNTER(reg, name) \
@@ -178,7 +169,6 @@ enum fw_state {
 static struct firmware_details fw_fabric;
 static struct firmware_details fw_pcie;
 static struct firmware_details fw_sbus;
-static const struct firmware *platform_config;
 
 /* flags for turn_off_spicos() */
 #define SPICO_SBUS   0x1
@@ -684,7 +674,6 @@ static void __obtain_firmware(struct hfi1_devdata *dd)
 static int obtain_firmware(struct hfi1_devdata *dd)
 {
 	unsigned long timeout;
-	int err = 0;
 
 	mutex_lock(&fw_mutex);
 
@@ -708,38 +697,11 @@ static int obtain_firmware(struct hfi1_devdata *dd)
 	}
 	/* not in FW_TRY state */
 
-	if (fw_state == FW_FINAL) {
-		if (platform_config) {
-			dd->platform_config.data = platform_config->data;
-			dd->platform_config.size = platform_config->size;
-		}
-		goto done;	/* already acquired */
-	} else if (fw_state == FW_ERR) {
-		goto done;	/* already tried and failed */
-	}
-	/* fw_state is FW_EMPTY */
-
 	/* set fw_state to FW_TRY, FW_FINAL, or FW_ERR, and fw_err */
-	__obtain_firmware(dd);
-
-	if (platform_config_load) {
-		platform_config = NULL;
-		err = request_firmware(&platform_config, platform_config_name,
-				       &dd->pcidev->dev);
-		if (err) {
-			platform_config = NULL;
-			dd_dev_err(dd,
-				   "%s: No default platform config file found\n",
-				   __func__);
-			goto done;
-		}
-		dd->platform_config.data = platform_config->data;
-		dd->platform_config.size = platform_config->size;
-	}
+	if (fw_state == FW_EMPTY)
+		__obtain_firmware(dd);
 
-done:
 	mutex_unlock(&fw_mutex);
-
 	return fw_err;
 }
 
@@ -761,9 +723,6 @@ void dispose_firmware(void)
 	dispose_one_firmware(&fw_pcie);
 	dispose_one_firmware(&fw_sbus);
 
-	release_firmware(platform_config);
-	platform_config = NULL;
-
 	/* retain the error state, otherwise revert to empty */
 	if (fw_state != FW_ERR)
 		fw_state = FW_EMPTY;
@@ -1725,10 +1684,8 @@ int hfi1_firmware_init(struct hfi1_devdata *dd)
 	}
 
 	/* no 8051 or QSFP on simulator */
-	if (dd->icode == ICODE_FUNCTIONAL_SIMULATOR) {
+	if (dd->icode == ICODE_FUNCTIONAL_SIMULATOR)
 		fw_8051_load = 0;
-		platform_config_load = 0;
-	}
 
 	if (!fw_8051_name) {
 		if (dd->icode == ICODE_RTL_SILICON)
@@ -1742,8 +1699,6 @@ int hfi1_firmware_init(struct hfi1_devdata *dd)
 		fw_sbus_name = DEFAULT_FW_SBUS_NAME;
 	if (!fw_pcie_serdes_name)
 		fw_pcie_serdes_name = DEFAULT_FW_PCIE_NAME;
-	if (!platform_config_name)
-		platform_config_name = DEFAULT_PLATFORM_CONFIG_NAME;
 
 	return obtain_firmware(dd);
 }
diff --git a/drivers/infiniband/hw/hfi1/platform.c b/drivers/infiniband/hw/hfi1/platform.c
index ab221b0..8004f1d 100644
--- a/drivers/infiniband/hw/hfi1/platform.c
+++ b/drivers/infiniband/hw/hfi1/platform.c
@@ -45,10 +45,14 @@
  *
  */
 
+#include <linux/firmware.h>
+
 #include "hfi.h"
 #include "efivar.h"
 #include "eprom.h"
 
+#define DEFAULT_PLATFORM_CONFIG_NAME "hfi1_platform.dat"
+
 static int validate_scratch_checksum(struct hfi1_devdata *dd)
 {
 	u64 checksum = 0, temp_scratch = 0;
@@ -147,6 +151,7 @@ void get_platform_config(struct hfi1_devdata *dd)
 	int ret = 0;
 	u8 *temp_platform_config = NULL;
 	u32 esize;
+	const struct firmware *platform_config_file = NULL;
 
 	if (is_integrated(dd)) {
 		if (validate_scratch_checksum(dd)) {
@@ -167,23 +172,33 @@ void get_platform_config(struct hfi1_devdata *dd)
 	dd_dev_err(dd,
 		   "%s: Failed to get platform config, falling back to sub-optimal default file\n",
 		   __func__);
-	/* fall back to request firmware */
-	platform_config_load = 1;
-}
 
-void free_platform_config(struct hfi1_devdata *dd)
-{
-	if (!platform_config_load) {
-		/*
-		 * was loaded from EFI or the EPROM, release memory
-		 * allocated by read_efi_var/eprom_read_platform_config
-		 */
-		kfree(dd->platform_config.data);
+	ret = request_firmware(&platform_config_file,
+			       DEFAULT_PLATFORM_CONFIG_NAME,
+			       &dd->pcidev->dev);
+	if (ret) {
+		dd_dev_err(dd,
+			   "%s: No default platform config file found\n",
+			   __func__);
+		return;
 	}
+
 	/*
-	 * else do nothing, dispose_firmware will release
-	 * struct firmware platform_config on driver exit
+	 * Allocate separate memory block to store data and free firmware
+	 * structure. This allows free_platform_config to treat EPROM and
+	 * fallback configs in the same manner.
 	 */
+	dd->platform_config.data = kmemdup(platform_config_file->data,
+					   platform_config_file->size,
+					   GFP_KERNEL);
+	dd->platform_config.size = platform_config_file->size;
+	release_firmware(platform_config_file);
+}
+
+void free_platform_config(struct hfi1_devdata *dd)
+{
+	/* Release memory allocated for eprom or fallback file read. */
+	kfree(dd->platform_config.data);
 }
 
 void get_port_type(struct hfi1_pportdata *ppd)

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH for-next 5/8] IB/hfi1: Stricter bounds checking of MAD trap index

[Dennis Dalessandro]

From: Kamenee Arumugame <kamenee.arumugam-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

The macro size is valid. This change makes it less ambiguous.
Bounds check trap type for better security.

Reviewed-by: Michael J. Ruhl <michael.j.ruhl-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Kamenee Arumugam <kamenee.arumugam-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/mad.c |   13 ++++++++++++-
 include/rdma/rdma_vt.h           |    2 +-
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/mad.c b/drivers/infiniband/hw/hfi1/mad.c
index 1be1d74..ba6f58e 100644
--- a/drivers/infiniband/hw/hfi1/mad.c
+++ b/drivers/infiniband/hw/hfi1/mad.c
@@ -151,13 +151,24 @@ static void cleanup_traps(struct hfi1_ibport *ibp, struct trap_node *trap)
 	unsigned long flags;
 	unsigned long timeout;
 	int found = 0;
+	unsigned int queue_id;
+	static int trap_count;
+
+	queue_id = trap->data.generic_type & 0x0F;
+	if (queue_id >= RVT_MAX_TRAP_LISTS) {
+		trap_count++;
+		pr_err_ratelimited("hfi1: Invalid trap 0x%0x dropped. Total dropped: %d\n",
+				  trap->data.generic_type, trap_count);
+		kfree(trap);
+		return NULL;
+	}
 
 	/*
 	 * Since the retry (handle timeout) does not remove a trap request
 	 * from the list, all we have to do is compare the node.
 	 */
 	spin_lock_irqsave(&ibp->rvp.lock, flags);
-	trap_list = &ibp->rvp.trap_lists[trap->data.generic_type & 0x0F];
+	trap_list = &ibp->rvp.trap_lists[queue_id];
 
 	list_for_each_entry(node, &trap_list->list, list) {
 		if (node == trap) {
diff --git a/include/rdma/rdma_vt.h b/include/rdma/rdma_vt.h
index 1d94f3c..1ba84a7 100644
--- a/include/rdma/rdma_vt.h
+++ b/include/rdma/rdma_vt.h
@@ -64,7 +64,7 @@
 #define RVT_MAX_PKEY_VALUES 16
 
 #define RVT_MAX_TRAP_LEN 100 /* Limit pending trap list */
-#define RVT_MAX_TRAP_LISTS ((IB_NOTICE_TYPE_INFO & 0x0F) + 1)
+#define RVT_MAX_TRAP_LISTS 5 /*((IB_NOTICE_TYPE_INFO & 0x0F) + 1)*/
 #define RVT_TRAP_TIMEOUT 4096 /* 4.096 usec */
 
 struct trap_list {

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH for-next 6/8] IB/hfi1: Remove pstate from hfi1_pportdata

[Dennis Dalessandro]

From: Jakub Byczkowski <jakub.byczkowski-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Do not track physical state separately from host_link_state.
Deduce physical state from host_link_state when required.
Change cache_physical_state to log_physical_state to make
sure host_link_state reflects hardwares physical state properly.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Reviewed-by: Ira Weiny <ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Jakub Byczkowski <jakub.byczkowski-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/chip.c |   54 +++++++++++++++++++++----------------
 drivers/infiniband/hw/hfi1/chip.h |    3 +-
 drivers/infiniband/hw/hfi1/hfi.h  |   24 ----------------
 drivers/infiniband/hw/hfi1/mad.c  |    2 +
 4 files changed, 33 insertions(+), 50 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c
index a6c6534..45f6874 100644
--- a/drivers/infiniband/hw/hfi1/chip.c
+++ b/drivers/infiniband/hw/hfi1/chip.c
@@ -1068,6 +1068,8 @@ static int do_8051_command(struct hfi1_devdata *dd, u32 type, u64 in_data,
 static void update_statusp(struct hfi1_pportdata *ppd, u32 state);
 static int wait_logical_linkstate(struct hfi1_pportdata *ppd, u32 state,
 				  int msecs);
+static void log_state_transition(struct hfi1_pportdata *ppd, u32 state);
+static void log_physical_state(struct hfi1_pportdata *ppd, u32 state);
 static int wait_physical_linkstate(struct hfi1_pportdata *ppd, u32 state,
 				   int msecs);
 static void read_planned_down_reason_code(struct hfi1_devdata *dd, u8 *pdrrc);
@@ -10450,11 +10452,11 @@ static int goto_offline(struct hfi1_pportdata *ppd, u8 rem_reason)
 }
 
 /*
- * driver_physical_state - convert the driver's notion of a port's
+ * driver_pstate - convert the driver's notion of a port's
  * state (an HLS_*) into a physical state (a {IB,OPA}_PORTPHYSSTATE_*).
  * Return -1 (converted to a u32) to indicate error.
  */
-u32 driver_physical_state(struct hfi1_pportdata *ppd)
+u32 driver_pstate(struct hfi1_pportdata *ppd)
 {
 	switch (ppd->host_link_state) {
 	case HLS_UP_INIT:
@@ -10720,7 +10722,7 @@ int set_link_state(struct hfi1_pportdata *ppd, u32 state)
 		if (ret)
 			goto_offline(ppd, 0);
 		else
-			cache_physical_state(ppd);
+			log_physical_state(ppd, PLS_POLLING);
 		break;
 	case HLS_DN_DISABLE:
 		/* link is disabled */
@@ -10769,7 +10771,7 @@ int set_link_state(struct hfi1_pportdata *ppd, u32 state)
 		if (ppd->host_link_state != HLS_DN_POLL)
 			goto unexpected;
 		ppd->host_link_state = HLS_VERIFY_CAP;
-		cache_physical_state(ppd);
+		log_physical_state(ppd, PLS_CONFIGPHY_VERIFYCAP);
 		break;
 	case HLS_GOING_UP:
 		if (ppd->host_link_state != HLS_VERIFY_CAP)
@@ -12743,25 +12745,30 @@ static int wait_logical_linkstate(struct hfi1_pportdata *ppd, u32 state,
 	return 0;
 }
 
+static void log_state_transition(struct hfi1_pportdata *ppd, u32 state)
+{
+	u32 ib_pstate = chip_to_opa_pstate(ppd->dd, state);
+
+	dd_dev_info(ppd->dd,
+		    "physical state changed to %s (0x%x), phy 0x%x\n",
+		    opa_pstate_name(ib_pstate), ib_pstate, state);
+}
+
 /*
- * Read the physical hardware link state and set the driver's cached value
- * of it.
+ * Read the physical hardware link state and check if it matches host
+ * drivers anticipated state.
  */
-void cache_physical_state(struct hfi1_pportdata *ppd)
+static void log_physical_state(struct hfi1_pportdata *ppd, u32 state)
 {
-	u32 read_pstate;
-	u32 ib_pstate;
+	u32 read_state = read_physical_state(ppd->dd);
 
-	read_pstate = read_physical_state(ppd->dd);
-	ib_pstate = chip_to_opa_pstate(ppd->dd, read_pstate);
-	/* check if OPA pstate changed */
-	if (chip_to_opa_pstate(ppd->dd, ppd->pstate) != ib_pstate) {
-		dd_dev_info(ppd->dd,
-			    "%s: physical state changed to %s (0x%x), phy 0x%x\n",
-			    __func__, opa_pstate_name(ib_pstate), ib_pstate,
-			    read_pstate);
+	if (read_state == state) {
+		log_state_transition(ppd, state);
+	} else {
+		dd_dev_err(ppd->dd,
+			   "anticipated phy link state 0x%x, read 0x%x\n",
+			   state, read_state);
 	}
-	ppd->pstate = read_pstate;
 }
 
 /*
@@ -12776,22 +12783,24 @@ void cache_physical_state(struct hfi1_pportdata *ppd)
 static int wait_physical_linkstate(struct hfi1_pportdata *ppd, u32 state,
 				   int msecs)
 {
+	u32 read_state;
 	unsigned long timeout;
 
 	timeout = jiffies + msecs_to_jiffies(msecs);
 	while (1) {
-		cache_physical_state(ppd);
-		if (ppd->pstate == state)
+		read_state = read_physical_state(ppd->dd);
+		if (read_state == state)
 			break;
 		if (time_after(jiffies, timeout)) {
 			dd_dev_err(ppd->dd,
-				   "timeout waiting for phy link state 0x%x, current state is 0x%x\n",
-				   state, ppd->pstate);
+				   "timeout waiting for phy link state 0x%x\n",
+				   state);
 			return -ETIMEDOUT;
 		}
 		usleep_range(1950, 2050); /* sleep 2ms-ish */
 	}
 
+	log_state_transition(ppd, state);
 	return 0;
 }
 
@@ -14896,7 +14905,6 @@ struct hfi1_devdata *hfi1_init_dd(struct pci_dev *pdev,
 		/* start in offline */
 		ppd->host_link_state = HLS_DN_OFFLINE;
 		init_vl_arb_caches(ppd);
-		ppd->pstate = PLS_OFFLINE;
 	}
 
 	dd->link_default = HLS_DN_POLL;
diff --git a/drivers/infiniband/hw/hfi1/chip.h b/drivers/infiniband/hw/hfi1/chip.h
index fc21097..b8345a6 100644
--- a/drivers/infiniband/hw/hfi1/chip.h
+++ b/drivers/infiniband/hw/hfi1/chip.h
@@ -747,10 +747,9 @@ void update_usrhead(struct hfi1_ctxtdata *rcd, u32 hd, u32 updegr, u32 egrhd,
 int is_bx(struct hfi1_devdata *dd);
 u32 read_physical_state(struct hfi1_devdata *dd);
 u32 chip_to_opa_pstate(struct hfi1_devdata *dd, u32 chip_pstate);
-void cache_physical_state(struct hfi1_pportdata *ppd);
 const char *opa_lstate_name(u32 lstate);
 const char *opa_pstate_name(u32 pstate);
-u32 driver_physical_state(struct hfi1_pportdata *ppd);
+u32 driver_pstate(struct hfi1_pportdata *ppd);
 u32 driver_lstate(struct hfi1_pportdata *ppd);
 
 int acquire_lcb_access(struct hfi1_devdata *dd, int sleep_ok);
diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index 181feca..b15749e 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -758,7 +758,6 @@ struct hfi1_pportdata {
 	u8 link_enabled;	/* link enabled? */
 	u8 linkinit_reason;
 	u8 local_tx_rate;	/* rate given to 8051 firmware */
-	u8 pstate;		/* info only */
 	u8 qsfp_retry_count;
 
 	/* placeholders for IB MAD packet settings */
@@ -1428,29 +1427,6 @@ void hfi1_make_ud_req_16B(struct rvt_qp *qp,
 
 int hfi1_reset_device(int);
 
-/* return the driver's idea of the physical OPA port state */
-static inline u32 driver_pstate(struct hfi1_pportdata *ppd)
-{
-	/*
-	 * When DC is shut down and state is changed, its CSRs are not
-	 * impacted, therefore host_link_state should be used to get
-	 * current physical state.
-	 */
-	if (ppd->dd->dc_shutdown)
-		return driver_physical_state(ppd);
-	/*
-	 * The driver does some processing from the time the physical
-	 * link state is at LINKUP to the time the SM can be notified
-	 * as such. Return IB_PORTPHYSSTATE_TRAINING until the software
-	 * state is ready.
-	 */
-	if (ppd->pstate == PLS_LINKUP &&
-	    !(ppd->host_link_state & HLS_UP))
-		return IB_PORTPHYSSTATE_TRAINING;
-	else
-		return chip_to_opa_pstate(ppd->dd, ppd->pstate);
-}
-
 void receive_interrupt_work(struct work_struct *work);
 
 /* extract service channel from header and rhf */
diff --git a/drivers/infiniband/hw/hfi1/mad.c b/drivers/infiniband/hw/hfi1/mad.c
index ba6f58e..9336e78 100644
--- a/drivers/infiniband/hw/hfi1/mad.c
+++ b/drivers/infiniband/hw/hfi1/mad.c
@@ -1181,7 +1181,7 @@ static int physical_transition_allowed(int old, int new)
 static int port_states_transition_allowed(struct hfi1_pportdata *ppd,
 					  u32 logical_new, u32 physical_new)
 {
-	u32 physical_old = driver_physical_state(ppd);
+	u32 physical_old = driver_pstate(ppd);
 	u32 logical_old = driver_lstate(ppd);
 	int ret, logical_allowed, physical_allowed;
 

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH for-next 7/8] IB/hfi1: Remove HFI1_VERBS_31BIT_PSN option

[Dennis Dalessandro]

From: Grzegorz Morys <grzegorz.morys-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Remove HFI1_VERBS_31BIT_PSN Kconfig option leaving only 31-bit PSNs
available. The option was implemented in the early days of the driver
and is no longer needed.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Grzegorz Morys <grzegorz.morys-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/Kconfig |    7 -------
 drivers/infiniband/hw/hfi1/verbs.h |    5 -----
 2 files changed, 0 insertions(+), 12 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/Kconfig b/drivers/infiniband/hw/hfi1/Kconfig
index f6ea088..7b146b6 100644
--- a/drivers/infiniband/hw/hfi1/Kconfig
+++ b/drivers/infiniband/hw/hfi1/Kconfig
@@ -13,13 +13,6 @@ config HFI1_DEBUG_SDMA_ORDER
 	---help---
 	This is a debug flag to test for out of order
 	sdma completions for unit testing
-config HFI1_VERBS_31BIT_PSN
-	bool "HFI1 enable 31 bit PSN"
-	depends on INFINIBAND_HFI1
-	default y
-	---help---
-	Setting this enables 31 BIT PSN
-	For verbs RC/UC
 config SDMA_VERBOSITY
 	bool "Config SDMA Verbosity"
 	depends on INFINIBAND_HFI1
diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h
index 56ead87..87d1285 100644
--- a/drivers/infiniband/hw/hfi1/verbs.h
+++ b/drivers/infiniband/hw/hfi1/verbs.h
@@ -280,13 +280,8 @@ int hfi1_process_mad(struct ib_device *ibdev, int mad_flags, u8 port,
  * necessarily be at least one bit less than
  * the container holding the PSN.
  */
-#ifndef CONFIG_HFI1_VERBS_31BIT_PSN
-#define PSN_MASK 0xFFFFFF
-#define PSN_SHIFT 8
-#else
 #define PSN_MASK 0x7FFFFFFF
 #define PSN_SHIFT 1
-#endif
 #define PSN_MODIFY_MASK 0xFFFFFF
 
 /*

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH for-next 8/8] IB/hfi1: Add kernel receive context info to debugfs

[Dennis Dalessandro]

From: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/debugfs.c |   47 +++++++++++++++++++++++++++
 drivers/infiniband/hw/hfi1/driver.c  |   59 ++++++++++++++++++++++++++++++++++
 drivers/infiniband/hw/hfi1/hfi.h     |    1 +
 3 files changed, 107 insertions(+), 0 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/debugfs.c b/drivers/infiniband/hw/hfi1/debugfs.c
index 550119c..c6a472f 100644
--- a/drivers/infiniband/hw/hfi1/debugfs.c
+++ b/drivers/infiniband/hw/hfi1/debugfs.c
@@ -368,6 +368,52 @@ static int _sdes_seq_show(struct seq_file *s, void *v)
 DEBUGFS_SEQ_FILE_OPEN(sdes)
 DEBUGFS_FILE_OPS(sdes);
 
+static void *_rcds_seq_start(struct seq_file *s, loff_t *pos)
+{
+	struct hfi1_ibdev *ibd;
+	struct hfi1_devdata *dd;
+
+	ibd = (struct hfi1_ibdev *)s->private;
+	dd = dd_from_dev(ibd);
+	if (!dd->rcd || *pos >= dd->n_krcv_queues)
+		return NULL;
+	return pos;
+}
+
+static void *_rcds_seq_next(struct seq_file *s, void *v, loff_t *pos)
+{
+	struct hfi1_ibdev *ibd = (struct hfi1_ibdev *)s->private;
+	struct hfi1_devdata *dd = dd_from_dev(ibd);
+
+	++*pos;
+	if (!dd->rcd || *pos >= dd->n_krcv_queues)
+		return NULL;
+	return pos;
+}
+
+static void _rcds_seq_stop(struct seq_file *s, void *v)
+{
+}
+
+static int _rcds_seq_show(struct seq_file *s, void *v)
+{
+	struct hfi1_ibdev *ibd = (struct hfi1_ibdev *)s->private;
+	struct hfi1_devdata *dd = dd_from_dev(ibd);
+	struct hfi1_ctxtdata *rcd;
+	loff_t *spos = v;
+	loff_t i = *spos;
+
+	rcd = hfi1_rcd_get_by_index(dd, i);
+	if (rcd)
+		seqfile_dump_rcd(s, rcd);
+	hfi1_rcd_put(rcd);
+	return 0;
+}
+
+DEBUGFS_SEQ_FILE_OPS(rcds);
+DEBUGFS_SEQ_FILE_OPEN(rcds)
+DEBUGFS_FILE_OPS(rcds);
+
 /* read the per-device counters */
 static ssize_t dev_counters_read(struct file *file, char __user *buf,
 				 size_t count, loff_t *ppos)
@@ -1321,6 +1367,7 @@ void hfi1_dbg_ibdev_init(struct hfi1_ibdev *ibd)
 	DEBUGFS_SEQ_FILE_CREATE(ctx_stats, ibd->hfi1_ibdev_dbg, ibd);
 	DEBUGFS_SEQ_FILE_CREATE(qp_stats, ibd->hfi1_ibdev_dbg, ibd);
 	DEBUGFS_SEQ_FILE_CREATE(sdes, ibd->hfi1_ibdev_dbg, ibd);
+	DEBUGFS_SEQ_FILE_CREATE(rcds, ibd->hfi1_ibdev_dbg, ibd);
 	DEBUGFS_SEQ_FILE_CREATE(sdma_cpu_list, ibd->hfi1_ibdev_dbg, ibd);
 	/* dev counter files */
 	for (i = 0; i < ARRAY_SIZE(cntr_ops); i++)
diff --git a/drivers/infiniband/hw/hfi1/driver.c b/drivers/infiniband/hw/hfi1/driver.c
index fc7085d..7372cc0 100644
--- a/drivers/infiniband/hw/hfi1/driver.c
+++ b/drivers/infiniband/hw/hfi1/driver.c
@@ -1689,3 +1689,62 @@ int process_receive_invalid(struct hfi1_packet *packet)
 		   rhf_rcv_type(packet->rhf));
 	return RHF_RCV_CONTINUE;
 }
+
+void seqfile_dump_rcd(struct seq_file *s, struct hfi1_ctxtdata *rcd)
+{
+	struct hfi1_packet packet;
+	struct ps_mdata mdata;
+
+	seq_printf(s, "Rcd %u: RcvHdr cnt %u entsize %u %s head %llu tail %llu\n",
+		   rcd->ctxt, rcd->rcvhdrq_cnt, rcd->rcvhdrqentsize,
+		   HFI1_CAP_KGET_MASK(rcd->flags, DMA_RTAIL) ?
+		   "dma_rtail" : "nodma_rtail",
+		   read_uctxt_csr(rcd->dd, rcd->ctxt, RCV_HDR_HEAD) &
+		   RCV_HDR_HEAD_HEAD_MASK,
+		   read_uctxt_csr(rcd->dd, rcd->ctxt, RCV_HDR_TAIL));
+
+	init_packet(rcd, &packet);
+	init_ps_mdata(&mdata, &packet);
+
+	while (1) {
+		struct hfi1_devdata *dd = rcd->dd;
+		__le32 *rhf_addr = (__le32 *)rcd->rcvhdrq + mdata.ps_head +
+					 dd->rhf_offset;
+		struct ib_header *hdr;
+		u64 rhf = rhf_to_cpu(rhf_addr);
+		u32 etype = rhf_rcv_type(rhf), qpn;
+		u8 opcode;
+		u32 psn;
+		u8 lnh;
+
+		if (ps_done(&mdata, rhf, rcd))
+			break;
+
+		if (ps_skip(&mdata, rhf, rcd))
+			goto next;
+
+		if (etype > RHF_RCV_TYPE_IB)
+			goto next;
+
+		packet.hdr = hfi1_get_msgheader(dd, rhf_addr);
+		hdr = packet.hdr;
+
+		lnh = be16_to_cpu(hdr->lrh[0]) & 3;
+
+		if (lnh == HFI1_LRH_BTH)
+			packet.ohdr = &hdr->u.oth;
+		else if (lnh == HFI1_LRH_GRH)
+			packet.ohdr = &hdr->u.l.oth;
+		else
+			goto next; /* just in case */
+
+		opcode = (be32_to_cpu(packet.ohdr->bth[0]) >> 24);
+		qpn = be32_to_cpu(packet.ohdr->bth[1]) & RVT_QPN_MASK;
+		psn = mask_psn(be32_to_cpu(packet.ohdr->bth[2]));
+
+		seq_printf(s, "\tEnt %u: opcode 0x%x, qpn 0x%x, psn 0x%x\n",
+			   mdata.ps_head, opcode, qpn, psn);
+next:
+		update_ps_mdata(&mdata, rcd);
+	}
+}
diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index b15749e..eadb735 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -1987,6 +1987,7 @@ static inline void flush_wc(void)
 int kdeth_process_expected(struct hfi1_packet *packet);
 int kdeth_process_eager(struct hfi1_packet *packet);
 int process_receive_invalid(struct hfi1_packet *packet);
+void seqfile_dump_rcd(struct seq_file *s, struct hfi1_ctxtdata *rcd);
 
 /* global module parameter variables */
 extern unsigned int hfi1_max_mtu;

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH for-next 0/8] IB/hfi1: patches for next 08/13/2017

[Leon Romanovsky]

[-- Attachment #1: Type: text/plain, Size: 248 bytes --]

On Sun, Aug 13, 2017 at 08:08:15AM -0700, Dennis Dalessandro wrote:
>
> There is one patch to an OPA header file that addresses a question that Leon
> had on the list about OPA physical port states.

Thanks, I'll base my UAPI patches on that code.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH for-next 0/8] IB/hfi1: patches for next 08/13/2017

[Doug Ledford]

On Sun, 2017-08-13 at 08:08 -0700, Dennis Dalessandro wrote:
> Hi Doug,
> 
> Here is the next set of patches for our drivers for 4.14. One of
> these patches
> gets rid of a config option that is no longer needed. There are some
> cleanups
> as usual and minor bug fixes. Nothing that I would say needs to go
> into an RC.
> These can all just land in your for-next tree and apply on top of
> that branch
> with the addition of the pending patches still on the list.
> Everything applied
> cleanly.
> 
> There is one patch to an OPA header file that addresses a question
> that Leon
> had on the list about OPA physical port states.
> 
> Patches can can also be found in my GitHub repo at:
> https://github.com/ddalessa/kernel/tree/for-4.14

Thanks Denny, series applied.

-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
    GPG KeyID: B826A3330E572FDD
    Key fingerprint = AE6B 1BDA 122B 23B4 265B  1274 B826 A333 0E57 2FDD

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html