* Re: [PATCH] net: add regs attribute to phy device for user diagnose
From: David Miller @ 2017-01-16 21:54 UTC (permalink / raw)
To: f.fainelli; +Cc: cugyly, andrew, netdev, Linyu.Yuan
In-Reply-To: <cf3e3125-582f-8b91-9c23-05dbdd2c6d24@gmail.com>
From: Florian Fainelli <f.fainelli@gmail.com>
Date: Mon, 16 Jan 2017 12:22:16 -0800
> On 01/16/2017 04:59 AM, yuan linyu wrote:
>> On 日, 2017-01-15 at 18:21 +0100, Andrew Lunn wrote:
>>> On Sun, Jan 15, 2017 at 09:51:03AM +0800, yuan linyu wrote:
>>>>
>>>> I hope user/developer can read this attribute file "regs" to do
>>>> a full check of all registers value, and they can write any register
>>>> inside PHY through this file.
>>> Since this is intended for debug, it should not be sysfs, but debugfs.
>> agree,
>>> However, in general, Linux does not allow user space to peek and poke
>>> device registers. Can you point me at examples where i can do the same
>>> to my GPU? SATA controller? Ethernet controller, I2C temperature
>>> sensor? Any device?
>> we can read registers of ethernet controller(memory register accessed) through devmem or ethtool
>
> So why not add support in ethtool for reading PHY registers if you need
> it? There are handful of PHY "things" in ethtool, such as reading
> counters, configuring downshift etc., adding support for dumping
> registers does not sound out of space.
Agreed.
^ permalink raw reply
* Re: [PATCH net-next 6/9] net/mlx4_en: Adding support of turning off link autonegotiation via ethtool
From: Or Gerlitz @ 2017-01-16 21:48 UTC (permalink / raw)
To: Tariq Toukan, Ariel Levkovich
Cc: David S. Miller, Linux Netdev List, Eran Ben Elisha
In-Reply-To: <1484587805-12666-7-git-send-email-tariqt@mellanox.com>
On Mon, Jan 16, 2017 at 7:30 PM, Tariq Toukan <tariqt@mellanox.com> wrote:
0644
> --- a/include/linux/mlx4/device.h
> +++ b/include/linux/mlx4/device.h
> @@ -1539,8 +1539,13 @@ enum mlx4_ptys_proto {
> MLX4_PTYS_EN = 1<<2,
> };
>
> +enum mlx4_ptys_flags {
> + MLX4_PTYS_AN_DISABLE_CAP = 1 << 5,
> + MLX4_PTYS_AN_DISABLE_ADMIN = 1 << 6,
> +};
> +
> struct mlx4_ptys_reg {
> - u8 resrvd1;
> + u8 flags;
> u8 local_port;
> u8 resrvd2;
> u8 proto_mask;
Hi Ariel,
I didn't see any new dev cap bit, what happens with FW version which
don't support this, is the result well defined? what is it?
Or.
^ permalink raw reply
* Re: [PATCH 1/2] qed: Add support for hardware offloaded FCoE.
From: Chad Dupuis @ 2017-01-16 21:47 UTC (permalink / raw)
To: martin.petersen
Cc: linux-scsi, fcoe-devel, netdev, yuval.mintz,
QLogic-Storage-Upstream
In-Reply-To: <1484596437-27637-2-git-send-email-chad.dupuis@cavium.com>
I forgot to add netdev-next to the subject line. Is a repost needed here?
On Mon, 16 Jan 2017, 7:53pm -0000, Dupuis, Chad wrote:
> From: Arun Easi <arun.easi@qlogic.com>
>
> This adds the backbone required for the various HW initalizations
> which are necessary for the FCoE driver (qedf) for QLogic FastLinQ
> 4xxxx line of adapters - FW notification, resource initializations, etc.
>
> Signed-off-by: Arun Easi <arun.easi@cavium.com>
> Signed-off-by: Yuval Mintz <yuval.mintz@cavium.com>
> ---
> drivers/net/ethernet/qlogic/Kconfig | 3 +
> drivers/net/ethernet/qlogic/qed/Makefile | 1 +
> drivers/net/ethernet/qlogic/qed/qed.h | 11 +
> drivers/net/ethernet/qlogic/qed/qed_cxt.c | 98 ++-
> drivers/net/ethernet/qlogic/qed/qed_cxt.h | 3 +
> drivers/net/ethernet/qlogic/qed/qed_dcbx.c | 11 +
> drivers/net/ethernet/qlogic/qed/qed_dcbx.h | 1 +
> drivers/net/ethernet/qlogic/qed/qed_dev.c | 205 ++++-
> drivers/net/ethernet/qlogic/qed/qed_dev_api.h | 42 +
> drivers/net/ethernet/qlogic/qed/qed_fcoe.c | 990 ++++++++++++++++++++++
> drivers/net/ethernet/qlogic/qed/qed_fcoe.h | 52 ++
> drivers/net/ethernet/qlogic/qed/qed_hsi.h | 781 ++++++++++++++++-
> drivers/net/ethernet/qlogic/qed/qed_hw.c | 3 +
> drivers/net/ethernet/qlogic/qed/qed_ll2.c | 25 +
> drivers/net/ethernet/qlogic/qed/qed_ll2.h | 2 +-
> drivers/net/ethernet/qlogic/qed/qed_main.c | 7 +
> drivers/net/ethernet/qlogic/qed/qed_mcp.c | 3 +
> drivers/net/ethernet/qlogic/qed/qed_mcp.h | 1 +
> drivers/net/ethernet/qlogic/qed/qed_reg_addr.h | 8 +
> drivers/net/ethernet/qlogic/qed/qed_sp.h | 4 +
> drivers/net/ethernet/qlogic/qed/qed_sp_commands.c | 3 +
> include/linux/qed/common_hsi.h | 10 +-
> include/linux/qed/fcoe_common.h | 715 ++++++++++++++++
> include/linux/qed/qed_fcoe_if.h | 145 ++++
> include/linux/qed/qed_if.h | 39 +
> 25 files changed, 3152 insertions(+), 11 deletions(-)
> create mode 100644 drivers/net/ethernet/qlogic/qed/qed_fcoe.c
> create mode 100644 drivers/net/ethernet/qlogic/qed/qed_fcoe.h
> create mode 100644 include/linux/qed/fcoe_common.h
> create mode 100644 include/linux/qed/qed_fcoe_if.h
>
^ permalink raw reply
* [PATCH v2] net: marvell: skge: use new api ethtool_{get|set}_link_ksettings
From: Philippe Reynes @ 2017-01-16 21:46 UTC (permalink / raw)
To: mlindner, stephen, davem; +Cc: netdev, linux-kernel, Philippe Reynes
The ethtool api {get|set}_settings is deprecated.
We move this driver to new api {get|set}_link_ksettings.
The callback set_link_ksettings no longer update the value
of advertising, as the struct ethtool_link_ksettings is
defined as const.
As I don't have the hardware, I'd be very pleased if
someone may test this patch.
Signed-off-by: Philippe Reynes <tremyfr@gmail.com>
---
Changelog:
v2:
- update the commit log with a note to explain that this code
was not tested on hardware (feedback from Stephen Hemminger)
drivers/net/ethernet/marvell/skge.c | 63 ++++++++++++++++++++--------------
1 files changed, 37 insertions(+), 26 deletions(-)
diff --git a/drivers/net/ethernet/marvell/skge.c b/drivers/net/ethernet/marvell/skge.c
index 9146a51..81106b7 100644
--- a/drivers/net/ethernet/marvell/skge.c
+++ b/drivers/net/ethernet/marvell/skge.c
@@ -300,65 +300,76 @@ static u32 skge_supported_modes(const struct skge_hw *hw)
return supported;
}
-static int skge_get_settings(struct net_device *dev,
- struct ethtool_cmd *ecmd)
+static int skge_get_link_ksettings(struct net_device *dev,
+ struct ethtool_link_ksettings *cmd)
{
struct skge_port *skge = netdev_priv(dev);
struct skge_hw *hw = skge->hw;
+ u32 supported, advertising;
- ecmd->transceiver = XCVR_INTERNAL;
- ecmd->supported = skge_supported_modes(hw);
+ supported = skge_supported_modes(hw);
if (hw->copper) {
- ecmd->port = PORT_TP;
- ecmd->phy_address = hw->phy_addr;
+ cmd->base.port = PORT_TP;
+ cmd->base.phy_address = hw->phy_addr;
} else
- ecmd->port = PORT_FIBRE;
+ cmd->base.port = PORT_FIBRE;
+
+ advertising = skge->advertising;
+ cmd->base.autoneg = skge->autoneg;
+ cmd->base.speed = skge->speed;
+ cmd->base.duplex = skge->duplex;
+
+ ethtool_convert_legacy_u32_to_link_mode(cmd->link_modes.supported,
+ supported);
+ ethtool_convert_legacy_u32_to_link_mode(cmd->link_modes.advertising,
+ advertising);
- ecmd->advertising = skge->advertising;
- ecmd->autoneg = skge->autoneg;
- ethtool_cmd_speed_set(ecmd, skge->speed);
- ecmd->duplex = skge->duplex;
return 0;
}
-static int skge_set_settings(struct net_device *dev, struct ethtool_cmd *ecmd)
+static int skge_set_link_ksettings(struct net_device *dev,
+ const struct ethtool_link_ksettings *cmd)
{
struct skge_port *skge = netdev_priv(dev);
const struct skge_hw *hw = skge->hw;
u32 supported = skge_supported_modes(hw);
int err = 0;
+ u32 advertising;
+
+ ethtool_convert_link_mode_to_legacy_u32(&advertising,
+ cmd->link_modes.advertising);
- if (ecmd->autoneg == AUTONEG_ENABLE) {
- ecmd->advertising = supported;
+ if (cmd->base.autoneg == AUTONEG_ENABLE) {
+ advertising = supported;
skge->duplex = -1;
skge->speed = -1;
} else {
u32 setting;
- u32 speed = ethtool_cmd_speed(ecmd);
+ u32 speed = cmd->base.speed;
switch (speed) {
case SPEED_1000:
- if (ecmd->duplex == DUPLEX_FULL)
+ if (cmd->base.duplex == DUPLEX_FULL)
setting = SUPPORTED_1000baseT_Full;
- else if (ecmd->duplex == DUPLEX_HALF)
+ else if (cmd->base.duplex == DUPLEX_HALF)
setting = SUPPORTED_1000baseT_Half;
else
return -EINVAL;
break;
case SPEED_100:
- if (ecmd->duplex == DUPLEX_FULL)
+ if (cmd->base.duplex == DUPLEX_FULL)
setting = SUPPORTED_100baseT_Full;
- else if (ecmd->duplex == DUPLEX_HALF)
+ else if (cmd->base.duplex == DUPLEX_HALF)
setting = SUPPORTED_100baseT_Half;
else
return -EINVAL;
break;
case SPEED_10:
- if (ecmd->duplex == DUPLEX_FULL)
+ if (cmd->base.duplex == DUPLEX_FULL)
setting = SUPPORTED_10baseT_Full;
- else if (ecmd->duplex == DUPLEX_HALF)
+ else if (cmd->base.duplex == DUPLEX_HALF)
setting = SUPPORTED_10baseT_Half;
else
return -EINVAL;
@@ -371,11 +382,11 @@ static int skge_set_settings(struct net_device *dev, struct ethtool_cmd *ecmd)
return -EINVAL;
skge->speed = speed;
- skge->duplex = ecmd->duplex;
+ skge->duplex = cmd->base.duplex;
}
- skge->autoneg = ecmd->autoneg;
- skge->advertising = ecmd->advertising;
+ skge->autoneg = cmd->base.autoneg;
+ skge->advertising = advertising;
if (netif_running(dev)) {
skge_down(dev);
@@ -875,8 +886,6 @@ static int skge_set_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom
}
static const struct ethtool_ops skge_ethtool_ops = {
- .get_settings = skge_get_settings,
- .set_settings = skge_set_settings,
.get_drvinfo = skge_get_drvinfo,
.get_regs_len = skge_get_regs_len,
.get_regs = skge_get_regs,
@@ -899,6 +908,8 @@ static int skge_set_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom
.set_phys_id = skge_set_phys_id,
.get_sset_count = skge_get_sset_count,
.get_ethtool_stats = skge_get_ethtool_stats,
+ .get_link_ksettings = skge_get_link_ksettings,
+ .set_link_ksettings = skge_set_link_ksettings,
};
/*
--
1.7.4.4
^ permalink raw reply related
* Re: [patch net-next] stmmac: indent an if statement
From: Julia Lawall @ 2017-01-16 21:46 UTC (permalink / raw)
To: Dan Carpenter
Cc: David Miller, peppe.cavallaro, Joao.Pinto, alexandre.torgue,
netdev, linux-kernel, kernel-janitors
In-Reply-To: <20170116093954.GB4104@mwanda>
On Mon, 16 Jan 2017, Dan Carpenter wrote:
> On Mon, Jan 16, 2017 at 12:19:24PM +0300, Dan Carpenter wrote:
> > On Sun, Jan 15, 2017 at 10:14:38PM -0500, David Miller wrote:
> > > From: Dan Carpenter <dan.carpenter@oracle.com>
> > > Date: Thu, 12 Jan 2017 21:46:32 +0300
> > >
> > > > The break statement should be indented one more tab.
> > > >
> > > > Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> > >
> > > Applied, but like Julia I think we might have a missing of_node_put()
> > > here.
> >
> > Of course, sorry for dropping the ball on this. I'll send a patch for
> > that.
> >
>
> Actually, I've looked at it some more and I think this function is OK.
> We're supposed to do an of_node_put() later... I can't find where that
> happens, but presumably that's because I don't know stmmac well. This
> code here, though, is fine.
Why do you think it is fine? Does anyone in the calling context know
which child would have caused the break? An extra put is only needed on
that one. Is there a guarantee that the break is always taken?
julia
^ permalink raw reply
* Re: [PATCH net-next 3/9] net/mlx4_core: Set EQ affinity hint to local NUMA CPUs
From: Or Gerlitz @ 2017-01-16 21:44 UTC (permalink / raw)
To: Tariq Toukan, Daniel Jurgens
Cc: David S. Miller, Linux Netdev List, Eran Ben Elisha
In-Reply-To: <1484587805-12666-4-git-send-email-tariqt@mellanox.com>
On Mon, Jan 16, 2017 at 7:29 PM, Tariq Toukan <tariqt@mellanox.com> wrote:
> From: Daniel Jurgens <danielj@mellanox.com>
>
> Use CPUs on the close NUMA when setting the EQ affinity hints.
Dan, are we sure there are no down-sides for always doing this? this
code is probably there for many years and we're introducing here new
behaviour to potentially to many Ms installs when they get distro
update that includes this patch.
^ permalink raw reply
* Re: [PATCH] net: marvell: skge: use new api ethtool_{get|set}_link_ksettings
From: Philippe Reynes @ 2017-01-16 21:27 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: David Miller, mlindner, netdev, linux-kernel
In-Reply-To: <20170116095003.6998b773@xeon-e3>
Hi Stephen,
On 1/16/17, Stephen Hemminger <stephen@networkplumber.org> wrote:
> On Mon, 16 Jan 2017 12:36:17 -0500 (EST)
> David Miller <davem@davemloft.net> wrote:
>
>> From: Stephen Hemminger <stephen@networkplumber.org>
>> Date: Mon, 16 Jan 2017 09:29:51 -0800
>>
>> > On Sat, 14 Jan 2017 13:08:28 +0100
>> > Philippe Reynes <tremyfr@gmail.com> wrote:
>> >
>> >> The ethtool api {get|set}_settings is deprecated.
>> >> We move this driver to new api {get|set}_link_ksettings.
>> >>
>> >> The callback set_link_ksettings no longer update the value
>> >> of advertising, as the struct ethtool_link_ksettings is
>> >> defined as const.
>> >>
>> >> Signed-off-by: Philippe Reynes <tremyfr@gmail.com>
>> >
>> > Did you test this on real hardware?
>>
>> Philippe probably doesn't have physical access to most of the
>> drivers he is converting.
>>
>> But, he is the only person working on converting all of the drivers,
>> and therefore when the change looks straightforward I am going to
>> reward his work and effort by applying his changes and hope there
>> isn't any fallout.
>>
>> Those who really care can test his patches and give a Tested-by:
>>
>> Thanks.
>
> Yes, it looks mechanical and should be applied. There are lots of pieces of
> old hardware that no developer is running, and if we required full test
> suite runs
> on all drivers, then no refactoring would ever be possible.
>
> My preference is to always add commit note that the patch was compile
> tested only so that if someone has a problem with real hardware then they
> know
> what to suspect.
David is right, I don't have the hardware to test all drivers.
I haven't added a little note on all commit, because I was
thinking it would be a lot of noise. But if you prefer (and
David agrees), I'll add a note on all the following patches.
Btw, thanks David for applying my patches.
Philippe
^ permalink raw reply
* Re: [PATCH v2] net/irda: fix lockdep annotation
From: Dmitry Vyukov @ 2017-01-16 21:11 UTC (permalink / raw)
To: kbuild test robot
Cc: kbuild-all, Dave Jones, Samuel Ortiz, David Miller,
Alexander Potapenko, andreyknvl, netdev
In-Reply-To: <201701170149.G5mmkfWa%fengguang.wu@intel.com>
On Mon, Jan 16, 2017 at 7:09 PM, kbuild test robot <lkp@intel.com> wrote:
> Hi Dmitry,
>
> [auto build test ERROR on tip/locking/core]
> [also build test ERROR on v4.10-rc4 next-20170116]
> [if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
>
> url: https://github.com/0day-ci/linux/commits/Dmitry-Vyukov/net-irda-fix-lockdep-annotation/20170117-001737
> config: i386-defconfig (attached as .config)
> compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
> reproduce:
> # save the attached .config to linux build tree
> make ARCH=i386
>
> All error/warnings (new ones prefixed by >>):
>
> In file included from include/linux/mmzone.h:7:0,
> from include/linux/gfp.h:5,
> from include/linux/slab.h:14,
> from drivers/gpu/drm/i915/i915_sw_fence.c:10:
> drivers/gpu/drm/i915/i915_sw_fence.c: In function '__i915_sw_fence_wake_up_all':
>>> include/linux/spinlock.h:217:3: error: void value not ignored as it ought to be
> (void)subclass; \
>
>>> include/linux/spinlock.h:335:2: note: in expansion of macro 'raw_spin_lock_irqsave_nested'
> raw_spin_lock_irqsave_nested(spinlock_check(lock), flags, subclass); \
> ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>> drivers/gpu/drm/i915/i915_sw_fence.c:68:2: note: in expansion of macro 'spin_lock_irqsave_nested'
> spin_lock_irqsave_nested(&x->lock, flags, 1 + !!continuation);
> ^~~~~~~~~~~~~~~~~~~~~~~~
Mailed v3.
> vim +217 include/linux/spinlock.h
>
> 211 typecheck(unsigned long, flags); \
> 212 flags = _raw_spin_lock_irqsave_nested(lock, subclass); \
> 213 } while (0)
> 214 #else
> 215 #define raw_spin_lock_irqsave_nested(lock, flags, subclass) \
> 216 do { \
> > 217 (void)subclass; \
> 218 typecheck(unsigned long, flags); \
> 219 flags = _raw_spin_lock_irqsave(lock); \
> 220 } while (0)
> 221 #endif
> 222
> 223 #else
> 224
> 225 #define raw_spin_lock_irqsave(lock, flags) \
> 226 do { \
> 227 typecheck(unsigned long, flags); \
> 228 _raw_spin_lock_irqsave(lock, flags); \
> 229 } while (0)
> 230
> 231 #define raw_spin_lock_irqsave_nested(lock, flags, subclass) \
> 232 raw_spin_lock_irqsave(lock, flags)
> 233
> 234 #endif
> 235
> 236 #define raw_spin_lock_irq(lock) _raw_spin_lock_irq(lock)
> 237 #define raw_spin_lock_bh(lock) _raw_spin_lock_bh(lock)
> 238 #define raw_spin_unlock(lock) _raw_spin_unlock(lock)
> 239 #define raw_spin_unlock_irq(lock) _raw_spin_unlock_irq(lock)
> 240
> 241 #define raw_spin_unlock_irqrestore(lock, flags) \
> 242 do { \
> 243 typecheck(unsigned long, flags); \
> 244 _raw_spin_unlock_irqrestore(lock, flags); \
> 245 } while (0)
> 246 #define raw_spin_unlock_bh(lock) _raw_spin_unlock_bh(lock)
> 247
> 248 #define raw_spin_trylock_bh(lock) \
> 249 __cond_lock(lock, _raw_spin_trylock_bh(lock))
> 250
> 251 #define raw_spin_trylock_irq(lock) \
> 252 ({ \
> 253 local_irq_disable(); \
> 254 raw_spin_trylock(lock) ? \
> 255 1 : ({ local_irq_enable(); 0; }); \
> 256 })
> 257
> 258 #define raw_spin_trylock_irqsave(lock, flags) \
> 259 ({ \
> 260 local_irq_save(flags); \
> 261 raw_spin_trylock(lock) ? \
> 262 1 : ({ local_irq_restore(flags); 0; }); \
> 263 })
> 264
> 265 /**
> 266 * raw_spin_can_lock - would raw_spin_trylock() succeed?
> 267 * @lock: the spinlock in question.
> 268 */
> 269 #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock))
> 270
> 271 /* Include rwlock functions */
> 272 #include <linux/rwlock.h>
> 273
> 274 /*
> 275 * Pull the _spin_*()/_read_*()/_write_*() functions/declarations:
> 276 */
> 277 #if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
> 278 # include <linux/spinlock_api_smp.h>
> 279 #else
> 280 # include <linux/spinlock_api_up.h>
> 281 #endif
> 282
> 283 /*
> 284 * Map the spin_lock functions to the raw variants for PREEMPT_RT=n
> 285 */
> 286
> 287 static __always_inline raw_spinlock_t *spinlock_check(spinlock_t *lock)
> 288 {
> 289 return &lock->rlock;
> 290 }
> 291
> 292 #define spin_lock_init(_lock) \
> 293 do { \
> 294 spinlock_check(_lock); \
> 295 raw_spin_lock_init(&(_lock)->rlock); \
> 296 } while (0)
> 297
> 298 static __always_inline void spin_lock(spinlock_t *lock)
> 299 {
> 300 raw_spin_lock(&lock->rlock);
> 301 }
> 302
> 303 static __always_inline void spin_lock_bh(spinlock_t *lock)
> 304 {
> 305 raw_spin_lock_bh(&lock->rlock);
> 306 }
> 307
> 308 static __always_inline int spin_trylock(spinlock_t *lock)
> 309 {
> 310 return raw_spin_trylock(&lock->rlock);
> 311 }
> 312
> 313 #define spin_lock_nested(lock, subclass) \
> 314 do { \
> 315 raw_spin_lock_nested(spinlock_check(lock), subclass); \
> 316 } while (0)
> 317
> 318 #define spin_lock_nest_lock(lock, nest_lock) \
> 319 do { \
> 320 raw_spin_lock_nest_lock(spinlock_check(lock), nest_lock); \
> 321 } while (0)
> 322
> 323 static __always_inline void spin_lock_irq(spinlock_t *lock)
> 324 {
> 325 raw_spin_lock_irq(&lock->rlock);
> 326 }
> 327
> 328 #define spin_lock_irqsave(lock, flags) \
> 329 do { \
> 330 raw_spin_lock_irqsave(spinlock_check(lock), flags); \
> 331 } while (0)
> 332
> 333 #define spin_lock_irqsave_nested(lock, flags, subclass) \
> 334 do { \
> > 335 raw_spin_lock_irqsave_nested(spinlock_check(lock), flags, subclass); \
> 336 } while (0)
> 337
> 338 static __always_inline void spin_unlock(spinlock_t *lock)
>
> ---
> 0-DAY kernel test infrastructure Open Source Technology Center
> https://lists.01.org/pipermail/kbuild-all Intel Corporation
^ permalink raw reply
* [PATCH v3] net/irda: fix lockdep annotation
From: Dmitry Vyukov @ 2017-01-16 21:10 UTC (permalink / raw)
To: davej, samuel, davem; +Cc: glider, andreyknvl, Dmitry Vyukov, netdev
The current annotation uses a global variable as recursion counter.
The variable is not atomic nor protected with a mutex, but mutated
by multiple threads. This causes lockdep bug reports episodically:
BUG: looking up invalid subclass: 4294967295
...
_raw_spin_lock_irqsave_nested+0x120/0x180
hashbin_delete+0x4fe/0x750
__irias_delete_object+0xab/0x170
irias_delete_object+0x5f/0xc0
ircomm_tty_detach_cable+0x1d5/0x3f0
...
Make the hashbin_lock_depth variable atomic to prevent bug reports.
As is this causes "unused variable 'depth'" warning without LOCKDEP.
So also change raw_spin_lock_irqsave_nested() macro to not cause
the warning without LOCKDEP. Similar to what raw_spin_lock_nested()
already does.
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Dave Jones <davej@redhat.com>
Cc: Samuel Ortiz <samuel@sortiz.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Fixes: c7630a4b932af ("[IrDA]: irda lockdep annotation")
---
Changes since v1:
- Added raw_spin_lock_irqsave_nested() change
as 0-DAY bot reported compiler warning without LOCKDEP.
Changes since v2:
- Use ((void)(subclass), (lock)) instead of (void)subclass
to prevent unused variable warning. Now 0-DAY complaints
"error: void value not ignored as it ought to be".
My compiler does not produce warning with W=1 either way,
but the comma trick is what already used in raw_spin_lock_nested().
---
include/linux/spinlock.h | 2 +-
net/irda/irqueue.c | 12 +++++++-----
2 files changed, 8 insertions(+), 6 deletions(-)
diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h
index 47dd0cebd204..beeb5a39bd8b 100644
--- a/include/linux/spinlock.h
+++ b/include/linux/spinlock.h
@@ -218,7 +218,7 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock)
#define raw_spin_lock_irqsave_nested(lock, flags, subclass) \
do { \
typecheck(unsigned long, flags); \
- flags = _raw_spin_lock_irqsave(lock); \
+ flags = _raw_spin_lock_irqsave(((void)(subclass), (lock))); \
} while (0)
#endif
diff --git a/net/irda/irqueue.c b/net/irda/irqueue.c
index acbe61c7e683..b9fd74e6ca99 100644
--- a/net/irda/irqueue.c
+++ b/net/irda/irqueue.c
@@ -384,21 +384,23 @@ EXPORT_SYMBOL(hashbin_new);
* just supply kfree, which should take care of the job.
*/
#ifdef CONFIG_LOCKDEP
-static int hashbin_lock_depth = 0;
+static atomic_t hashbin_lock_depth = ATOMIC_INIT(0);
#endif
int hashbin_delete( hashbin_t* hashbin, FREE_FUNC free_func)
{
irda_queue_t* queue;
unsigned long flags = 0;
- int i;
+ int i, depth = 0;
IRDA_ASSERT(hashbin != NULL, return -1;);
IRDA_ASSERT(hashbin->magic == HB_MAGIC, return -1;);
/* Synchronize */
if ( hashbin->hb_type & HB_LOCK ) {
- spin_lock_irqsave_nested(&hashbin->hb_spinlock, flags,
- hashbin_lock_depth++);
+#ifdef CONFIG_LOCKDEP
+ depth = atomic_inc_return(&hashbin_lock_depth) - 1;
+#endif
+ spin_lock_irqsave_nested(&hashbin->hb_spinlock, flags, depth);
}
/*
@@ -423,7 +425,7 @@ int hashbin_delete( hashbin_t* hashbin, FREE_FUNC free_func)
if ( hashbin->hb_type & HB_LOCK) {
spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
#ifdef CONFIG_LOCKDEP
- hashbin_lock_depth--;
+ atomic_dec(&hashbin_lock_depth);
#endif
}
--
2.11.0.483.g087da7b7c-goog
^ permalink raw reply related
* Re: [pull request][for-next] Mellanox mlx5 Reorganize core driver directory layout
From: Tom Herbert @ 2017-01-16 21:06 UTC (permalink / raw)
To: Saeed Mahameed
Cc: David Miller, Saeed Mahameed, Doug Ledford, Linux Netdev List,
linux-rdma, Leon Romanovsky
In-Reply-To: <CALzJLG--ZRVk0Zy+6UaGCWbOhki+PZnkmrh9rjojU1ZDm_5Y6w@mail.gmail.com>
On Mon, Jan 16, 2017 at 12:30 PM, Saeed Mahameed
<saeedm@dev.mellanox.co.il> wrote:
> On Sat, Jan 14, 2017 at 12:07 AM, Saeed Mahameed
> <saeedm@dev.mellanox.co.il> wrote:
>> On Fri, Jan 13, 2017 at 7:14 PM, David Miller <davem@davemloft.net> wrote:
>>> From: Saeed Mahameed <saeedm@mellanox.com>
>>> Date: Thu, 12 Jan 2017 19:22:34 +0200
>>>
>>>> This pull request includes one patch from Leon, this patch as described
>>>> below will change the driver directory structure and layout for better,
>>>> logical and modular driver files separation.
>>>>
>>>> This change is important to both rdma and net maintainers in order to
>>>> have smoother management of driver patches for different mlx5 sub modules
>>>> and smoother rdma-next vs. net-next features submissions.
>>>>
>>>> Please find more info below -in the tag commit message-,
>>>> review and let us know if there's any problem.
>>>>
>>>> This change doesn't introduce any conflicts with the current mlx5
>>>> fixes and cleanups posted on 2017-01-10 to net branch, and merge tests
>>>> worked flawlessly with no issues.
>>>>
>>>> This is the last pull request meant for both rdma-next and net-next.
>>>> Once pulled, this will be the base shared code for both trees.
>>>
>>> This is pretty crazy, it will make all bug fix backporting to -stable
>>> a complete nightmare for myself, Doug, various distribution maintainers
>>> and many other people who quietly have to maintain their own trees and
>>> do backporting.
>>>
>>
>> I hear you,
>> But please bear with me here, what if we queue this patch up to -stable ? and we
>> (Mellanox) and specifically our dedicated inbox team, will make sure
>> that this patch
>> will land on the various distributions and for those maintaining their
>> own trees.
>> This patch is really straight forward (rename files) and I already
>> tried to cherry-pick it
>> on older kernels, I only got a couple of conflicts on some of the
>> "#inlcude" lines we've
>> changed, and they are pretty straightforward to fix, we can even avoid
>> this if we decide to
>> not move mlx5 header files in this phase.
>>
>> If this is possible then all trees will be aligned and it will be a
>> win win situation.
>>
>
> Hi Dave,
>
> Any chance you saw my -stable suggestion above ?
> I think it would really close the backporting gap.
>
> Sorry i am bothering you with this topic, but we really desire the new
> structure and
> I never got your feedback on this suggestion, so i would like to hear
> your thoughts.
>
Saeed,
I've already you specific suggestions on your new structure, please
consider your reviewers feedback more carefully. Again, the starting
point for your restructuring should be to separate out the minimum set
of files required to build reasonable driver and then cleanly
compartmentalize the rest of the features to make it easy for your
users to include or not include those in their build. Unless you've
done this I'm not seeing much benefit for this restructuring. Also, I
would rather see this done in one shot then expecting some sort of
evolution over time to the right solution-- as Dave said the
complexity of this driver is to far down the road to do that.
Tom
> Thanks,
> Saeed.
>
>>> I really don't think you can justify this rearrangement based upon the
>>> consequences and how much activity happens in this driver.
>>>
>>
>> Right, but this is not the only justification, I can sum it up to that
>> we would like
>> to lay out the foundation for many years to come for a well designed
>> driver with a modular
>> sub modules break down and scalable infrastructure. We already plan to
>> submit more mlx5
>> independent sub modules - just like mlx5e (en_*) files and mlx5_ib
>> driver- so this was also
>> a reason for us to consider this re-engagement at this stage.
>>
>>> You should have thought long and hard about the layout a long time ago
>>> rather than after the driver has been in the tree for many years.
>>>
>>
>> I had this Idea for the separation before the mlx5 Ethernet
>> submission, but I wasn't the maintainer
>> back then, and i have been itching to submit such patch for long as
>> well, still i don't think
>> it is too late, We (Me and Leon) will keep maintaining this driver for
>> only god knows how many years to come,
>> and the mlx5 drivers are meant to serve at least 3-4 more future HW generations.
>>
>> Long story short, We would like to re-arrange the driver in a way that
>> would serve us (the maintainers) and serve those
>> who are going do develop the future Stack features and the future HW
>> generations over the well designed (Hopefully)
>> mlx5 infrastructure.
>> Keeping it as it is today, will only make the situation worst in the
>> future and it will be really hard to avoid having a spaghetti code
>> in the mlx5 driver. All i want to point out here is that maintaining
>> such a flat subtree is also nightmare.
>>
>> So i am only asking you to reconsider this change and give my -stable
>> suggestion a thought.
>>
>> Thank you.
>> Saeed.
^ permalink raw reply
* Re: [PATCH] net: constify mdiobb_ops structures
From: Sergei Shtylyov @ 2017-01-16 20:48 UTC (permalink / raw)
To: Bhumika Goyal, julia.lawall, pantelis.antoniou, vbordug,
f.fainelli, netdev, linux-kernel, linuxppc-dev, linux-renesas-soc
In-Reply-To: <1484330546-16569-1-git-send-email-bhumirks@gmail.com>
Hello!
On 01/13/2017 09:02 PM, Bhumika Goyal wrote:
> Declare mdiobb_ops structures as const as they are only stored in the
> ops field of mdiobb_ctrl structures. This field is of type const, so
> mdiobb_ops structures having this property can be declared const too.
> Done using Coccinelle:
[...]
> Signed-off-by: Bhumika Goyal <bhumirks@gmail.com>
[...]
> diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
> index 92d7692..1b0acd1 100644
> --- a/drivers/net/ethernet/renesas/ravb_main.c
> +++ b/drivers/net/ethernet/renesas/ravb_main.c
> @@ -171,7 +171,7 @@ static int ravb_get_mdio_data(struct mdiobb_ctrl *ctrl)
> }
>
> /* MDIO bus control struct */
> -static struct mdiobb_ops bb_ops = {
> +static const struct mdiobb_ops bb_ops = {
> .owner = THIS_MODULE,
> .set_mdc = ravb_set_mdc,
> .set_mdio_dir = ravb_set_mdio_dir,
> diff --git a/drivers/net/ethernet/renesas/sh_eth.c b/drivers/net/ethernet/renesas/sh_eth.c
> index 00fafab..6ef5dd8 100644
> --- a/drivers/net/ethernet/renesas/sh_eth.c
> +++ b/drivers/net/ethernet/renesas/sh_eth.c
> @@ -1052,7 +1052,7 @@ static void sh_mdc_ctrl(struct mdiobb_ctrl *ctrl, int bit)
> }
>
> /* mdio bus control struct */
> -static struct mdiobb_ops bb_ops = {
> +static const struct mdiobb_ops bb_ops = {
> .owner = THIS_MODULE,
> .set_mdc = sh_mdc_ctrl,
> .set_mdio_dir = sh_mmd_ctrl,
[...]
For the above 2 drivers:
Acked-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
MBR, Sergei
^ permalink raw reply
* Re: [PATCH/RFC v2 net-next] ravb: unmap descriptors when freeing rings
From: Sergei Shtylyov @ 2017-01-16 20:41 UTC (permalink / raw)
To: Simon Horman; +Cc: David Miller, Magnus Damm, netdev, linux-renesas-soc
In-Reply-To: <20170112131821.GA16060@verge.net.au>
Hello!
On 01/12/2017 04:18 PM, Simon Horman wrote:
> ...
>>>> Here, it stop once an untransmitted buffer is encountered...
>>>
>>> Yes, I see that now.
>>>
>>> I wonder if we should:
>>>
>>> a) paramatise ravb_tx_free() so it may either clear all transmitted buffers
>>> (current behaviour) or all buffers (new behaviour).
>>> b) provide a different version of this loop in ravb_ring_free()
>>>
>>> What are your thoughts?
>>
>> I'm voting for (b).
After looking at this issue for another time, I'll vote for (a). Sorry for
back-and-forth on this matter -- I somehow thought we could do a better job by
scanning all the TX ring...
> Ok, something like this?
>
> @@ -215,6 +225,30 @@ static void ravb_ring_free(struct net_device *ndev, int q)
> }
>
> if (priv->tx_ring[q]) {
> + for (; priv->cur_tx[q] - priv->dirty_tx[q] > 0; priv->dirty_tx[q]++) {
> + struct ravb_tx_desc *desc;
> + int entry;
> +
> + entry = priv->dirty_tx[q] % (priv->num_tx_ring[q] *
> + NUM_TX_DESC);
> + desc = &priv->tx_ring[q][entry];
> +
> + /* Free the original skb. */
> + if (priv->tx_skb[q][entry / NUM_TX_DESC]) {
> + u32 size = le16_to_cpu(desc->ds_tagl) & TX_DS;
> +
> + dma_unmap_single(ndev->dev.parent,
> + le32_to_cpu(desc->dptr),
> + size, DMA_TO_DEVICE);
> + /* Last packet descriptor? */
> + if (entry % NUM_TX_DESC == NUM_TX_DESC - 1) {
> + entry /= NUM_TX_DESC;
> + dev_kfree_skb_any(priv->tx_skb[q][entry]);
> + priv->tx_skb[q][entry] = NULL;
> + }
> + }
> + }
> +
This is only different from ravb_tx_free() by not stopping on unfinished
descriptor, so we must be good with adding an extra param to it...
[...]
MBR, Sergei
^ permalink raw reply
* Re: [PATCH V2] audit: log 32-bit socketcalls
From: Paul Moore @ 2017-01-16 20:38 UTC (permalink / raw)
To: David Miller; +Cc: rgb, netdev, linux-kernel, linux-audit
In-Reply-To: <20170116.132745.1831536278596554491.davem@davemloft.net>
On Mon, Jan 16, 2017 at 1:27 PM, David Miller <davem@davemloft.net> wrote:
> From: Richard Guy Briggs <rgb@redhat.com>
> Date: Fri, 13 Jan 2017 04:51:48 -0500
>
>> diff --git a/include/linux/audit.h b/include/linux/audit.h
>> index 9d4443f..43d8003 100644
>> --- a/include/linux/audit.h
>> +++ b/include/linux/audit.h
>> @@ -387,6 +387,18 @@ static inline int audit_socketcall(int nargs, unsigned long *args)
>> return __audit_socketcall(nargs, args);
>> return 0;
>> }
>> +static inline int audit_socketcall_compat(int nargs, u32 *args)
>> +{
>
> Please put an empty line between function definitions.
David, assuming Richard makes your requested changes, any objection if
I merge this via the audit tree instead of the netdev tree? It's a
bit easier for us from a testing perspective this way ...
>> + if (unlikely(!audit_dummy_context())) {
>> + int i;
>> + unsigned long a[AUDITSC_ARGS];
>
> Please order local variable declarations from longest to shortest line.
>
>> +
>> + for (i=0; i<nargs; i++)
>
> Please put a space around operators such as "=" and "<".
>
>> + a[i] = (unsigned long)args[i];
>> + return __audit_socketcall(nargs, a);
>> + }
>> + return 0;
>> +}
>> static inline int audit_sockaddr(int len, void *addr)
>
> Again, empty line between function definitions please.
>
>> @@ -781,14 +782,24 @@ COMPAT_SYSCALL_DEFINE5(recvmmsg, int, fd, struct compat_mmsghdr __user *, mmsg,
>>
>> COMPAT_SYSCALL_DEFINE2(socketcall, int, call, u32 __user *, args)
>> {
>> + unsigned int len;
>> int ret;
>> - u32 a[6];
>> + u32 a[AUDITSC_ARGS];
>> u32 a0, a1;
>
> Longest to shortest line for local variable declarations please.
>
> --
> Linux-audit mailing list
> Linux-audit@redhat.com
> https://www.redhat.com/mailman/listinfo/linux-audit
--
paul moore
www.paul-moore.com
^ permalink raw reply
* RE: [net-next 0/3] tipc: improve interaction socket-link
From: Jon Maloy @ 2017-01-16 20:35 UTC (permalink / raw)
To: David Miller
Cc: netdev@vger.kernel.org, viro@zeniv.linux.org.uk,
Parthasarathy Bhuvaragan, Ying Xue, maloy@donjonn.com,
tipc-discussion@lists.sourceforge.net
In-Reply-To: <20170116.144514.1071967219363714348.davem@davemloft.net>
> -----Original Message-----
> From: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org]
> On Behalf Of David Miller
> Sent: Monday, 16 January, 2017 14:45
> To: Jon Maloy <jon.maloy@ericsson.com>
> Cc: netdev@vger.kernel.org; viro@zeniv.linux.org.uk; Parthasarathy Bhuvaragan
> <parthasarathy.bhuvaragan@ericsson.com>; Ying Xue
> <ying.xue@windriver.com>; maloy@donjonn.com; tipc-
> discussion@lists.sourceforge.net
> Subject: Re: [net-next 0/3] tipc: improve interaction socket-link
>
> From: Jon Maloy <jon.maloy@ericsson.com>
> Date: Tue, 3 Jan 2017 10:26:45 -0500
>
> > We fix a very real starvation problem that may occur when a link
> > encounters send buffer congestion. At the same time we make the
> > interaction between the socket and link layer simpler and more
> > consistent.
>
> This doesn't apply to net-next, also the Date in your emails is 10 days
> in the past. What's going on here?
I don't know. This series was sent in on Jan 3rd, and applied by you the same day. Maybe the mail server decided to send you a duplicate?
///jon
^ permalink raw reply
* Re: [pull request][for-next] Mellanox mlx5 Reorganize core driver directory layout
From: Saeed Mahameed @ 2017-01-16 20:30 UTC (permalink / raw)
To: David Miller
Cc: Saeed Mahameed, Doug Ledford, Linux Netdev List,
linux-rdma-u79uwXL29TY76Z2rM5mHXA, Leon Romanovsky
In-Reply-To: <CALzJLG_fjiGyuB_buZqOSMU7SaCnuqFiZ0UyW+hPDq+3yx7TWg-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
On Sat, Jan 14, 2017 at 12:07 AM, Saeed Mahameed
<saeedm-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org> wrote:
> On Fri, Jan 13, 2017 at 7:14 PM, David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org> wrote:
>> From: Saeed Mahameed <saeedm-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
>> Date: Thu, 12 Jan 2017 19:22:34 +0200
>>
>>> This pull request includes one patch from Leon, this patch as described
>>> below will change the driver directory structure and layout for better,
>>> logical and modular driver files separation.
>>>
>>> This change is important to both rdma and net maintainers in order to
>>> have smoother management of driver patches for different mlx5 sub modules
>>> and smoother rdma-next vs. net-next features submissions.
>>>
>>> Please find more info below -in the tag commit message-,
>>> review and let us know if there's any problem.
>>>
>>> This change doesn't introduce any conflicts with the current mlx5
>>> fixes and cleanups posted on 2017-01-10 to net branch, and merge tests
>>> worked flawlessly with no issues.
>>>
>>> This is the last pull request meant for both rdma-next and net-next.
>>> Once pulled, this will be the base shared code for both trees.
>>
>> This is pretty crazy, it will make all bug fix backporting to -stable
>> a complete nightmare for myself, Doug, various distribution maintainers
>> and many other people who quietly have to maintain their own trees and
>> do backporting.
>>
>
> I hear you,
> But please bear with me here, what if we queue this patch up to -stable ? and we
> (Mellanox) and specifically our dedicated inbox team, will make sure
> that this patch
> will land on the various distributions and for those maintaining their
> own trees.
> This patch is really straight forward (rename files) and I already
> tried to cherry-pick it
> on older kernels, I only got a couple of conflicts on some of the
> "#inlcude" lines we've
> changed, and they are pretty straightforward to fix, we can even avoid
> this if we decide to
> not move mlx5 header files in this phase.
>
> If this is possible then all trees will be aligned and it will be a
> win win situation.
>
Hi Dave,
Any chance you saw my -stable suggestion above ?
I think it would really close the backporting gap.
Sorry i am bothering you with this topic, but we really desire the new
structure and
I never got your feedback on this suggestion, so i would like to hear
your thoughts.
Thanks,
Saeed.
>> I really don't think you can justify this rearrangement based upon the
>> consequences and how much activity happens in this driver.
>>
>
> Right, but this is not the only justification, I can sum it up to that
> we would like
> to lay out the foundation for many years to come for a well designed
> driver with a modular
> sub modules break down and scalable infrastructure. We already plan to
> submit more mlx5
> independent sub modules - just like mlx5e (en_*) files and mlx5_ib
> driver- so this was also
> a reason for us to consider this re-engagement at this stage.
>
>> You should have thought long and hard about the layout a long time ago
>> rather than after the driver has been in the tree for many years.
>>
>
> I had this Idea for the separation before the mlx5 Ethernet
> submission, but I wasn't the maintainer
> back then, and i have been itching to submit such patch for long as
> well, still i don't think
> it is too late, We (Me and Leon) will keep maintaining this driver for
> only god knows how many years to come,
> and the mlx5 drivers are meant to serve at least 3-4 more future HW generations.
>
> Long story short, We would like to re-arrange the driver in a way that
> would serve us (the maintainers) and serve those
> who are going do develop the future Stack features and the future HW
> generations over the well designed (Hopefully)
> mlx5 infrastructure.
> Keeping it as it is today, will only make the situation worst in the
> future and it will be really hard to avoid having a spaghetti code
> in the mlx5 driver. All i want to point out here is that maintaining
> such a flat subtree is also nightmare.
>
> So i am only asking you to reconsider this change and give my -stable
> suggestion a thought.
>
> Thank you.
> Saeed.
--
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
^ permalink raw reply
* Re: [PATCH] net: add regs attribute to phy device for user diagnose
From: Florian Fainelli @ 2017-01-16 20:22 UTC (permalink / raw)
To: cugyly, Andrew Lunn; +Cc: David S . Miller, netdev, yuan linyu
In-Reply-To: <1484571579.3374.4.camel@163.com>
On 01/16/2017 04:59 AM, yuan linyu wrote:
> On 日, 2017-01-15 at 18:21 +0100, Andrew Lunn wrote:
>> On Sun, Jan 15, 2017 at 09:51:03AM +0800, yuan linyu wrote:
>>>
>>> I hope user/developer can read this attribute file "regs" to do
>>> a full check of all registers value, and they can write any register
>>> inside PHY through this file.
>> Since this is intended for debug, it should not be sysfs, but debugfs.
> agree,
>> However, in general, Linux does not allow user space to peek and poke
>> device registers. Can you point me at examples where i can do the same
>> to my GPU? SATA controller? Ethernet controller, I2C temperature
>> sensor? Any device?
> we can read registers of ethernet controller(memory register accessed) through devmem or ethtool
So why not add support in ethtool for reading PHY registers if you need
it? There are handful of PHY "things" in ethtool, such as reading
counters, configuring downshift etc., adding support for dumping
registers does not sound out of space.
--
Florian
^ permalink raw reply
* Re: [PATCH net 0/3] mlx4 core fixes
From: David Miller @ 2017-01-16 20:08 UTC (permalink / raw)
To: tariqt; +Cc: netdev, eranbe, jackm
In-Reply-To: <1484584299-3950-1-git-send-email-tariqt@mellanox.com>
From: Tariq Toukan <tariqt@mellanox.com>
Date: Mon, 16 Jan 2017 18:31:36 +0200
> This patchset contains bug fixes from Jack to the mlx4 Core driver.
>
> Patch 1 solves a race in the flow of CQ free.
> Patch 2 moves some qp context flags update to the correct qp transition.
> Patch 3 eliminates warnings from the path of SRQ_LIMIT that flood the message log,
> and keeps them only in the path of SRQ_CATAS_ERROR.
>
> Series generated against net commit:
> 1a717fcf8bbe Merge tag 'mac80211-for-davem-2017-01-13' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
Series applied, thanks.
^ permalink raw reply
* Re: [pull request][for-next] Mellanox mlx5 Reorganize core driver directory layout
From: Saeed Mahameed @ 2017-01-16 20:15 UTC (permalink / raw)
To: Tom Herbert
Cc: Leon Romanovsky, David Miller, Saeed Mahameed, Doug Ledford,
Linux Kernel Network Developers,
linux-rdma-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <CALx6S35cSnMoBtxi+Eb_mH=LCLy9yY6U5sxofndce9r0eCs8+w-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
On Sat, Jan 14, 2017 at 12:56 AM, Tom Herbert <tom-BjP2VixgY4xUbtYUoyoikg@public.gmane.org> wrote:
> On Fri, Jan 13, 2017 at 2:45 PM, Saeed Mahameed
> <saeedm-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org> wrote:
>> On Sat, Jan 14, 2017 at 12:06 AM, Tom Herbert <tom-BjP2VixgY4xUbtYUoyoikg@public.gmane.org> wrote:
>>> Btw, we did hit one issue in the backport. We started to get rx csum
>>> faults (checksum complete value indicates TCP checksum is bad, but
>>> host computation says checksum is good). I ran against 4.9 upstream
>>> kernel and do see these, however don't see them in 4.10. I haven't
>>> bisected yet. Is this a known issue?
>>>
>>
>> Not to me, I don't recall any csum related fixes or feature submitted
>> lately to mlx5,
>> Maybe something changed in the stack ?
>>
>> what configuration are you running ? what traffic ?
>>
> Nothing fancy. 8 queues and 20 concurrent netperf TCP_STREAMs trips
> it. Not a lot of them, but I don't think we really should ever see
> these errors.
>
Hi Tom
I've been trying to repro with no luck with kernel v4.9.
I am looking for checksum errors in dmesg, is that the place where you
saw the errors ?
can you dump the error in here ?
--
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
^ permalink raw reply
* [PATCH 0/2] Add QLogic FastLinQ FCoE (qedf) driver
From: Dupuis, Chad @ 2017-01-16 19:53 UTC (permalink / raw)
To: martin.petersen
Cc: linux-scsi, fcoe-devel, netdev, yuval.mintz,
QLogic-Storage-Upstream
From: "Dupuis, Chad" <chad.dupuis@cavium.com>
This series introduces the hardware offload FCoE initiator driver for the
41000 Series Converged Network Adapters (579xx chip) by Cavium. The overall
driver design includes a common module ('qed') and protocol specific
dependent modules ('qedf' for FCoE).
This driver uses the kernel components of libfc and libfcoe as is and does not
make use of the open-fcoe user space components. Therefore, no changes will need to be
made to any open-fcoe components.
The 'qed' common module, under drivers/net/ethernet/qlogic/qed/, is
enhanced with functionality required for FCoE support.
Changes from RFC -> V1
- Squash qedf patches to one patch now that the initial review has taken place
- Convert qedf to use hotplug state machine
- Return via va_end to match corresponding va_start in logging functions
- Convert qedf_ctx offloaded port list to a RCU list so searches do not need
to make use of spinlocks. Also eliminates the need to fcport conn_id's.
- Use IS_ERR(fp) in qedf_flogi_resp() instead of checking individual FC_EX_* errors.
- Remove scsi_block_target when executing TMF request.
- Checkpatch fixes in the qed and qedf patches
At the current time I couldn't make use of Christoph's IRQ allocation changes
as changes are needed to the core module (qed) which would affect other drivers
that are already in tree (qede, qedr, qedi).
Arun Easi (1):
qed: Add support for hardware offloaded FCoE.
Dupuis, Chad (1):
qedf: Add QLogic FastLinQ offload FCoE driver framework.
MAINTAINERS | 6 +
drivers/net/ethernet/qlogic/Kconfig | 3 +
drivers/net/ethernet/qlogic/qed/Makefile | 1 +
drivers/net/ethernet/qlogic/qed/qed.h | 11 +
drivers/net/ethernet/qlogic/qed/qed_cxt.c | 98 +-
drivers/net/ethernet/qlogic/qed/qed_cxt.h | 3 +
drivers/net/ethernet/qlogic/qed/qed_dcbx.c | 11 +
drivers/net/ethernet/qlogic/qed/qed_dcbx.h | 1 +
drivers/net/ethernet/qlogic/qed/qed_dev.c | 205 +-
drivers/net/ethernet/qlogic/qed/qed_dev_api.h | 42 +
drivers/net/ethernet/qlogic/qed/qed_fcoe.c | 990 ++++++
drivers/net/ethernet/qlogic/qed/qed_fcoe.h | 52 +
drivers/net/ethernet/qlogic/qed/qed_hsi.h | 781 ++++-
drivers/net/ethernet/qlogic/qed/qed_hw.c | 3 +
drivers/net/ethernet/qlogic/qed/qed_ll2.c | 25 +
drivers/net/ethernet/qlogic/qed/qed_ll2.h | 2 +-
drivers/net/ethernet/qlogic/qed/qed_main.c | 7 +
drivers/net/ethernet/qlogic/qed/qed_mcp.c | 3 +
drivers/net/ethernet/qlogic/qed/qed_mcp.h | 1 +
drivers/net/ethernet/qlogic/qed/qed_reg_addr.h | 8 +
drivers/net/ethernet/qlogic/qed/qed_sp.h | 4 +
drivers/net/ethernet/qlogic/qed/qed_sp_commands.c | 3 +
drivers/scsi/Kconfig | 1 +
drivers/scsi/qedf/Kconfig | 11 +
drivers/scsi/qedf/Makefile | 5 +
drivers/scsi/qedf/qedf.h | 555 ++++
drivers/scsi/qedf/qedf_attr.c | 165 +
drivers/scsi/qedf/qedf_dbg.c | 195 ++
drivers/scsi/qedf/qedf_dbg.h | 154 +
drivers/scsi/qedf/qedf_debugfs.c | 471 +++
drivers/scsi/qedf/qedf_els.c | 984 ++++++
drivers/scsi/qedf/qedf_fip.c | 267 ++
drivers/scsi/qedf/qedf_hsi.h | 427 +++
drivers/scsi/qedf/qedf_io.c | 2301 ++++++++++++++
drivers/scsi/qedf/qedf_main.c | 3483 +++++++++++++++++++++
drivers/scsi/qedf/qedf_version.h | 15 +
include/linux/qed/common_hsi.h | 10 +-
include/linux/qed/fcoe_common.h | 715 +++++
include/linux/qed/qed_fcoe_if.h | 145 +
include/linux/qed/qed_if.h | 39 +
40 files changed, 12192 insertions(+), 11 deletions(-)
create mode 100644 drivers/net/ethernet/qlogic/qed/qed_fcoe.c
create mode 100644 drivers/net/ethernet/qlogic/qed/qed_fcoe.h
create mode 100644 drivers/scsi/qedf/Kconfig
create mode 100644 drivers/scsi/qedf/Makefile
create mode 100644 drivers/scsi/qedf/qedf.h
create mode 100644 drivers/scsi/qedf/qedf_attr.c
create mode 100644 drivers/scsi/qedf/qedf_dbg.c
create mode 100644 drivers/scsi/qedf/qedf_dbg.h
create mode 100644 drivers/scsi/qedf/qedf_debugfs.c
create mode 100644 drivers/scsi/qedf/qedf_els.c
create mode 100644 drivers/scsi/qedf/qedf_fip.c
create mode 100644 drivers/scsi/qedf/qedf_hsi.h
create mode 100644 drivers/scsi/qedf/qedf_io.c
create mode 100644 drivers/scsi/qedf/qedf_main.c
create mode 100644 drivers/scsi/qedf/qedf_version.h
create mode 100644 include/linux/qed/fcoe_common.h
create mode 100644 include/linux/qed/qed_fcoe_if.h
--
1.8.5.6
^ permalink raw reply
* Re: [PATCH net-next] net/sched: cls_flower: Disallow duplicate internal elements
From: David Miller @ 2017-01-16 20:06 UTC (permalink / raw)
To: paulb; +Cc: netdev, jiri, roid, ogerlitz, hadarh
In-Reply-To: <1484556313-33614-1-git-send-email-paulb@mellanox.com>
From: Paul Blakey <paulb@mellanox.com>
Date: Mon, 16 Jan 2017 10:45:13 +0200
> Flower currently allows having the same filter twice with the same
> priority. Actions (and statistics update) will always execute on the
> first inserted rule leaving the second rule unused.
> This patch disallows that.
>
> Signed-off-by: Paul Blakey <paulb@mellanox.com>
> Acked-by: Jiri Pirko <jiri@mellanox.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH V2] audit: log 32-bit socketcalls
From: Paul Moore @ 2017-01-16 20:04 UTC (permalink / raw)
To: Eric Paris, Richard Guy Briggs; +Cc: netdev, linux-kernel, linux-audit, Steve
In-Reply-To: <1484318543.5300.1.camel@redhat.com>
On Fri, Jan 13, 2017 at 9:42 AM, Eric Paris <eparis@redhat.com> wrote:
> On Fri, 2017-01-13 at 04:51 -0500, Richard Guy Briggs wrote:
>> diff --git a/include/linux/audit.h b/include/linux/audit.h
>> index 9d4443f..43d8003 100644
>> --- a/include/linux/audit.h
>> +++ b/include/linux/audit.h
>> @@ -387,6 +387,18 @@ static inline int audit_socketcall(int nargs,
>> unsigned long *args)
>> return __audit_socketcall(nargs, args);
>> return 0;
>> }
>> +static inline int audit_socketcall_compat(int nargs, u32 *args)
>> +{
>> + if (unlikely(!audit_dummy_context())) {
>
> I've always hated these likely/unlikely. Mostly because I think they
> are so often wrong. I believe this says that you compiled audit in but
> you expect it to be explicitly disabled. While that is (recently) true
> in Fedora I highly doubt that's true on the vast majority of systems
> that have audit compiled in.
Richard and I have talked about the likely/unlikely optimization
before and I know Richard likes to use them, but I don't for the
reasons Eric has already mentioned. Richard, since you're respinning
the patch, go ahead and yank out the unlikely() call.
--
paul moore
www.paul-moore.com
^ permalink raw reply
* Re: [PATCH] net: stmmac: don't use netdev_[dbg, info, ..] before net_device is registered
From: David Miller @ 2017-01-16 19:52 UTC (permalink / raw)
To: hkallweit1; +Cc: netdev
In-Reply-To: <01b736a6-b9e9-05cf-0937-f6ac18ac8e3c@gmail.com>
From: Heiner Kallweit <hkallweit1@gmail.com>
Date: Sun, 15 Jan 2017 19:19:00 +0100
> Don't use netdev_info and friends before the net_device is registered.
> This avoids ugly messages like
> "meson8b-dwmac c9410000.ethernet (unnamed net_device) (uninitialized):
> Enable RX Mitigation via HW Watchdog Timer"
>
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Indeed, this is a side-effect of the MDIO vs. register_netdev race fix
that went in the other week.
Applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next v3 06/10] net: dsa: Migrate to device_find_class()
From: Florian Fainelli @ 2017-01-16 20:01 UTC (permalink / raw)
To: Andrew Lunn
Cc: Greg KH, netdev, Jason Cooper, Sebastian Hesselbarth,
Gregory Clement, Russell King, Vivien Didelot, David S. Miller,
moderated list:ARM SUB-ARCHITECTURES, open list
In-Reply-To: <20170115191632.GD5643@lunn.ch>
On 01/15/2017 11:16 AM, Andrew Lunn wrote:
>>> What exactly is the relationship between these devices (a ascii-art tree
>>> or sysfs tree output might be nice) so I can try to understand what is
>>> going on here.
>
> Hi Greg, Florian
>
> A few diagrams and trees which might help understand what is going on.
>
> The first diagram comes from the 2008 patch which added all this code:
>
> +-----------+ +-----------+
> | | RGMII | |
> | +-------+ +------ 1000baseT MDI ("WAN")
> | | | 6-port +------ 1000baseT MDI ("LAN1")
> | CPU | | ethernet +------ 1000baseT MDI ("LAN2")
> | |MIImgmt| switch +------ 1000baseT MDI ("LAN3")
> | +-------+ w/5 PHYs +------ 1000baseT MDI ("LAN4")
> | | | |
> +-----------+ +-----------+
>
> We have an ethernet switch and a host CPU. The switch is connected to
> the CPU in two different ways. RGMII allows us to get Ethernet frames
> from the CPU into the switch. MIImgmt, is the management bus normally
> used for Ethernet PHYs, but Marvell switches also use it for Managing
> switches.
>
> The diagram above is the simplest setup. You can have multiple
> Ethernet switches, connected together via switch ports. Each switch
> has its own MIImgmt connect to the CPU, but there is only one RGMII
> link.
>
> When this code was designed back in 2008, it was decided to represent
> this is a platform device, and it has a platform_data, which i have
> slightly edited to keep it simple:
>
> struct dsa_platform_data {
> /*
> * Reference to a Linux network interface that connects
> * to the root switch chip of the tree.
> */
> struct device *netdev;
>
> /*
> * Info structs describing each of the switch chips
> * connected via this network interface.
> */
> int nr_chips;
> struct dsa_chip_data *chip;
> };
>
> This netdev is the CPU side of the RGMII interface.
>
> Each switch has a dsa_chip_data, again edited:
>
> struct dsa_chip_data {
> /*
> * How to access the switch configuration registers.
> */
> struct device *host_dev;
> int sw_addr;
> ...
> }
>
> The host_dev is the CPU side of the MIImgmt, and we have the address
> the switch is using on the bus.
>
> During probe of this platform device, we need to get from the
> struct device *netdev to a struct net_device *dev.
>
> So the code looks in the device net class to find the device
>
> | | | |-- f1074000.ethernet
> | | | | |-- deferred_probe
> | | | | |-- driver -> ../../../../../bus/platform/drivers/mvneta
> | | | | |-- driver_override
> | | | | |-- modalias
> | | | | |-- net
> | | | | | `-- eth1
> | | | | | |-- addr_assign_type
> | | | | | |-- address
> | | | | | |-- addr_len
> | | | | | |-- broadcast
> | | | | | |-- carrier
> | | | | | |-- carrier_changes
> | | | | | |-- deferred_probe
> | | | | | |-- device -> ../../../f1074000.ethernet
>
> and then use container_of() to get the net_device.
>
> Similarly, the code needs to get from struct device *host_dev to a struct mii_bus *.
>
> | | | |-- f1072004.mdio
> | | | | |-- deferred_probe
> | | | | |-- driver -> ../../../../../bus/platform/drivers/orion-mdio
> | | | | |-- driver_override
> | | | | |-- mdio_bus
> | | | | | `-- f1072004.mdio-mi
> | | | | | |-- deferred_probe
> | | | | | |-- device -> ../../../f1072004.mdio
>
Thanks Andrew! Greg, does that make it clearer how these devices
references are used, do you still think the way this is done is wrong,
too cautious, or valid?
--
Florian
^ permalink raw reply
* Re: [PATCH] qed: Replace memset with eth_zero_addr
From: David Miller @ 2017-01-16 19:58 UTC (permalink / raw)
To: mayhs11saini; +Cc: Yuval.Mintz, Ariel.Elior, everest-linux-l2, netdev
In-Reply-To: <1484592698-19747-1-git-send-email-mayhs11saini@gmail.com>
From: Shyam Saini <mayhs11saini@gmail.com>
Date: Tue, 17 Jan 2017 00:21:38 +0530
> Use eth_zero_addr to assign zero address to the given address array
> instead of memset when the second argument in memset is address
> of zero. Also, it makes the code clearer
>
> Signed-off-by: Shyam Saini <mayhs11saini@gmail.com>
This patch does not apply cleanly to the net-next tree.
Please state if you don't understand what that means instead of
submitting this patch over and over again against the wrong source
tree.
Thanks.
^ permalink raw reply
* [PATCH 2/2] qedf: Add QLogic FastLinQ offload FCoE driver framework.
From: Dupuis, Chad @ 2017-01-16 19:53 UTC (permalink / raw)
To: martin.petersen-QHcLZuEGTsvQT0dZR+AlfA
Cc: fcoe-devel-s9riP+hp16TNLxjTenLetw, netdev-u79uwXL29TY76Z2rM5mHXA,
QLogic-Storage-Upstream-YGCgFSpz5w/QT0dZR+AlfA,
linux-scsi-u79uwXL29TY76Z2rM5mHXA,
yuval.mintz-YGCgFSpz5w/QT0dZR+AlfA
In-Reply-To: <1484596437-27637-1-git-send-email-chad.dupuis-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org>
From: "Dupuis, Chad" <chad.dupuis-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org>
The QLogic FastLinQ Driver for FCoE (qedf) is the FCoE specific module for 41000
Series Converged Network Adapters by QLogic. This patch consists of following
changes:
- MAINTAINERS Makefile and Kconfig changes for qedf
- PCI driver registration
- libfc/fcoe host level initialization
- SCSI host template initialization and callbacks
- Debugfs and log level infrastructure
- Link handling
- Firmware interface structures
- QED core module initialization
- Light L2 interface callbacks
- I/O request initialization
- Firmware I/O completion handling
- Firmware ELS request/response handling
- FIP request/response handled by the driver itself
Signed-off-by: Nilesh Javali <nilesh.javali-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org>
Signed-off-by: Manish Rangankar <manish.rangankar-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org>
Signed-off-by: Saurav Kashyap <saurav.kashyap-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org>
Signed-off-by: Arun Easi <arun.easi-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org>
Signed-off-by: Chad Dupuis <chad.dupuis-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org>
---
MAINTAINERS | 6 +
drivers/scsi/Kconfig | 1 +
drivers/scsi/qedf/Kconfig | 11 +
drivers/scsi/qedf/Makefile | 5 +
drivers/scsi/qedf/qedf.h | 555 ++++++
drivers/scsi/qedf/qedf_attr.c | 165 ++
drivers/scsi/qedf/qedf_dbg.c | 195 +++
drivers/scsi/qedf/qedf_dbg.h | 154 ++
drivers/scsi/qedf/qedf_debugfs.c | 471 ++++++
drivers/scsi/qedf/qedf_els.c | 984 +++++++++++
drivers/scsi/qedf/qedf_fip.c | 267 +++
drivers/scsi/qedf/qedf_hsi.h | 427 +++++
drivers/scsi/qedf/qedf_io.c | 2301 +++++++++++++++++++++++++
drivers/scsi/qedf/qedf_main.c | 3483 ++++++++++++++++++++++++++++++++++++++
drivers/scsi/qedf/qedf_version.h | 15 +
15 files changed, 9040 insertions(+)
create mode 100644 drivers/scsi/qedf/Kconfig
create mode 100644 drivers/scsi/qedf/Makefile
create mode 100644 drivers/scsi/qedf/qedf.h
create mode 100644 drivers/scsi/qedf/qedf_attr.c
create mode 100644 drivers/scsi/qedf/qedf_dbg.c
create mode 100644 drivers/scsi/qedf/qedf_dbg.h
create mode 100644 drivers/scsi/qedf/qedf_debugfs.c
create mode 100644 drivers/scsi/qedf/qedf_els.c
create mode 100644 drivers/scsi/qedf/qedf_fip.c
create mode 100644 drivers/scsi/qedf/qedf_hsi.h
create mode 100644 drivers/scsi/qedf/qedf_io.c
create mode 100644 drivers/scsi/qedf/qedf_main.c
create mode 100644 drivers/scsi/qedf/qedf_version.h
diff --git a/MAINTAINERS b/MAINTAINERS
index d14e42b..9ab20fa 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -10157,6 +10157,12 @@ L: linux-scsi-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
S: Supported
F: drivers/scsi/qedi/
+QLOGIC QL41xxx FCOE DRIVER
+M: QLogic-Storage-Upstream-YGCgFSpz5w/QT0dZR+AlfA@public.gmane.org
+L: linux-scsi-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
+S: Supported
+F: drivers/scsi/qedf/
+
QNX4 FILESYSTEM
M: Anders Larsen <al-V9/YLgxv/GvR7s880joybQ@public.gmane.org>
W: http://www.alarsen.net/linux/qnx4fs/
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index a4f6b0d..e9fce78b 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -1234,6 +1234,7 @@ config SCSI_QLOGICPTI
source "drivers/scsi/qla2xxx/Kconfig"
source "drivers/scsi/qla4xxx/Kconfig"
source "drivers/scsi/qedi/Kconfig"
+source "drivers/scsi/qedf/Kconfig"
config SCSI_LPFC
tristate "Emulex LightPulse Fibre Channel Support"
diff --git a/drivers/scsi/qedf/Kconfig b/drivers/scsi/qedf/Kconfig
new file mode 100644
index 0000000..943f5ee
--- /dev/null
+++ b/drivers/scsi/qedf/Kconfig
@@ -0,0 +1,11 @@
+config QEDF
+ tristate "QLogic QEDF 25/40/100Gb FCoE Initiator Driver Support"
+ depends on PCI && SCSI
+ depends on QED
+ depends on LIBFC
+ depends on LIBFCOE
+ select QED_LL2
+ select QED_FCOE
+ ---help---
+ This driver supports FCoE offload for the QLogic FastLinQ
+ 41000 Series Converged Network Adapters.
diff --git a/drivers/scsi/qedf/Makefile b/drivers/scsi/qedf/Makefile
new file mode 100644
index 0000000..b2077d5
--- /dev/null
+++ b/drivers/scsi/qedf/Makefile
@@ -0,0 +1,5 @@
+obj-$(CONFIG_QEDF) := qedf.o
+qedf-objs = qedf_dbg.o qedf_main.o qedf_io.o qedf_fip.o \
+ qedf_attr.o qedf_els.o
+
+qedf-$(CONFIG_DEBUG_FS) += qedf_debugfs.o
diff --git a/drivers/scsi/qedf/qedf.h b/drivers/scsi/qedf/qedf.h
new file mode 100644
index 0000000..fdd5873
--- /dev/null
+++ b/drivers/scsi/qedf/qedf.h
@@ -0,0 +1,555 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#ifndef _QEDFC_H_
+#define _QEDFC_H_
+
+#include <scsi/libfcoe.h>
+#include <scsi/libfc.h>
+#include <scsi/fc/fc_fip.h>
+#include <scsi/fc/fc_fc2.h>
+#include <scsi/scsi_tcq.h>
+#include <scsi/fc_encode.h>
+#include <linux/version.h>
+
+
+/* qedf_hsi.h needs to before included any qed includes */
+#include "qedf_hsi.h"
+
+#include <linux/qed/qed_if.h>
+#include <linux/qed/qed_fcoe_if.h>
+#include <linux/qed/qed_ll2_if.h>
+#include "qedf_version.h"
+#include "qedf_dbg.h"
+
+/* Helpers to extract upper and lower 32-bits of pointer */
+#define U64_HI(val) ((u32)(((u64)(val)) >> 32))
+#define U64_LO(val) ((u32)(((u64)(val)) & 0xffffffff))
+
+#define QEDF_DESCR "QLogic FCoE Offload Driver"
+#define QEDF_MODULE_NAME "qedf"
+
+#define QEDF_MIN_XID 0
+#define QEDF_MAX_SCSI_XID (NUM_TASKS_PER_CONNECTION - 1)
+#define QEDF_MAX_ELS_XID 4095
+#define QEDF_FLOGI_RETRY_CNT 3
+#define QEDF_RPORT_RETRY_CNT 255
+#define QEDF_MAX_SESSIONS 1024
+#define QEDF_MAX_PAYLOAD 2048
+#define QEDF_MAX_BDS_PER_CMD 256
+#define QEDF_MAX_BD_LEN 0xffff
+#define QEDF_BD_SPLIT_SZ 0x1000
+#define QEDF_PAGE_SIZE 4096
+#define QED_HW_DMA_BOUNDARY 0xfff
+#define QEDF_MAX_SGLEN_FOR_CACHESGL ((1U << 16) - 1)
+#define QEDF_MFS (QEDF_MAX_PAYLOAD + \
+ sizeof(struct fc_frame_header))
+#define QEDF_MAX_NPIV 64
+#define QEDF_TM_TIMEOUT 10
+#define QEDF_ABORT_TIMEOUT 10
+#define QEDF_CLEANUP_TIMEOUT 10
+#define QEDF_MAX_CDB_LEN 16
+
+#define UPSTREAM_REMOVE 1
+#define UPSTREAM_KEEP 1
+
+struct qedf_mp_req {
+ uint8_t tm_flags;
+
+ uint32_t req_len;
+ void *req_buf;
+ dma_addr_t req_buf_dma;
+ struct fcoe_sge *mp_req_bd;
+ dma_addr_t mp_req_bd_dma;
+ struct fc_frame_header req_fc_hdr;
+
+ uint32_t resp_len;
+ void *resp_buf;
+ dma_addr_t resp_buf_dma;
+ struct fcoe_sge *mp_resp_bd;
+ dma_addr_t mp_resp_bd_dma;
+ struct fc_frame_header resp_fc_hdr;
+};
+
+struct qedf_els_cb_arg {
+ struct qedf_ioreq *aborted_io_req;
+ struct qedf_ioreq *io_req;
+ u8 op; /* Used to keep track of ELS op */
+ uint16_t l2_oxid;
+ u32 offset; /* Used for sequence cleanup */
+ u8 r_ctl; /* Used for sequence cleanup */
+};
+
+enum qedf_ioreq_event {
+ QEDF_IOREQ_EV_ABORT_SUCCESS,
+ QEDF_IOREQ_EV_ABORT_FAILED,
+ QEDF_IOREQ_EV_SEND_RRQ,
+ QEDF_IOREQ_EV_ELS_TMO,
+ QEDF_IOREQ_EV_ELS_ERR_DETECT,
+ QEDF_IOREQ_EV_ELS_FLUSH,
+ QEDF_IOREQ_EV_CLEANUP_SUCCESS,
+ QEDF_IOREQ_EV_CLEANUP_FAILED,
+};
+
+#define FC_GOOD 0
+#define FCOE_FCP_RSP_FLAGS_FCP_RESID_OVER (0x1<<2)
+#define FCOE_FCP_RSP_FLAGS_FCP_RESID_UNDER (0x1<<3)
+#define CMD_SCSI_STATUS(Cmnd) ((Cmnd)->SCp.Status)
+#define FCOE_FCP_RSP_FLAGS_FCP_RSP_LEN_VALID (0x1<<0)
+#define FCOE_FCP_RSP_FLAGS_FCP_SNS_LEN_VALID (0x1<<1)
+struct qedf_ioreq {
+ struct list_head link;
+ uint16_t xid;
+ struct scsi_cmnd *sc_cmd;
+ bool use_slowpath; /* Use slow SGL for this I/O */
+#define QEDF_SCSI_CMD 1
+#define QEDF_TASK_MGMT_CMD 2
+#define QEDF_ABTS 3
+#define QEDF_ELS 4
+#define QEDF_CLEANUP 5
+#define QEDF_SEQ_CLEANUP 6
+ u8 cmd_type;
+#define QEDF_CMD_OUTSTANDING 0x0
+#define QEDF_CMD_IN_ABORT 0x1
+#define QEDF_CMD_IN_CLEANUP 0x2
+#define QEDF_CMD_SRR_SENT 0x3
+ u8 io_req_flags;
+ struct qedf_rport *fcport;
+ unsigned long flags;
+ enum qedf_ioreq_event event;
+ size_t data_xfer_len;
+ struct kref refcount;
+ struct qedf_cmd_mgr *cmd_mgr;
+ struct io_bdt *bd_tbl;
+ struct delayed_work timeout_work;
+ struct completion tm_done;
+ struct completion abts_done;
+ struct fcoe_task_context *task;
+ int idx;
+/*
+ * Need to allocate enough room for both sense data and FCP response data
+ * which has a max length of 8 bytes according to spec.
+ */
+#define QEDF_SCSI_SENSE_BUFFERSIZE (SCSI_SENSE_BUFFERSIZE + 8)
+ uint8_t *sense_buffer;
+ dma_addr_t sense_buffer_dma;
+ u32 fcp_resid;
+ u32 fcp_rsp_len;
+ u32 fcp_sns_len;
+ u8 cdb_status;
+ u8 fcp_status;
+ u8 fcp_rsp_code;
+ u8 scsi_comp_flags;
+#define QEDF_MAX_REUSE 0xfff
+ u16 reuse_count;
+ struct qedf_mp_req mp_req;
+ void (*cb_func)(struct qedf_els_cb_arg *cb_arg);
+ struct qedf_els_cb_arg *cb_arg;
+ int fp_idx;
+ unsigned int cpu;
+ unsigned int int_cpu;
+#define QEDF_IOREQ_SLOW_SGE 0
+#define QEDF_IOREQ_SINGLE_SGE 1
+#define QEDF_IOREQ_FAST_SGE 2
+ u8 sge_type;
+ struct delayed_work rrq_work;
+
+ /* Used for sequence level recovery; i.e. REC/SRR */
+ uint32_t rx_buf_off;
+ uint32_t tx_buf_off;
+ uint32_t rx_id;
+ uint32_t task_retry_identifier;
+
+ /*
+ * Used to tell if we need to return a SCSI command
+ * during some form of error processing.
+ */
+ bool return_scsi_cmd_on_abts;
+};
+
+struct qedf_percpu_iothread_s {
+ struct task_struct *iothread;
+ struct list_head work_list;
+ spinlock_t work_lock;
+ uint64_t requests;
+ uint64_t responses;
+};
+
+struct qedf_rport {
+ spinlock_t rport_lock;
+#define QEDF_RPORT_SESSION_READY 1
+#define QEDF_RPORT_UPLOADING_CONNECTION 2
+ unsigned long flags;
+ unsigned long retry_delay_timestamp;
+ struct fc_rport *rport;
+ struct fc_rport_priv *rdata;
+ struct qedf_ctx *qedf;
+ u32 handle; /* Handle from qed */
+ u32 fw_cid; /* fw_cid from qed */
+ void __iomem *p_doorbell;
+ /* Send queue management */
+ atomic_t free_sqes;
+ atomic_t num_active_ios;
+ struct fcoe_wqe *sq;
+ dma_addr_t sq_dma;
+ u16 sq_prod_idx;
+ u16 fw_sq_prod_idx;
+ u16 sq_con_idx;
+ u32 sq_mem_size;
+ void *sq_pbl;
+ dma_addr_t sq_pbl_dma;
+ u32 sq_pbl_size;
+ u32 sid;
+#define QEDF_RPORT_TYPE_DISK 1
+#define QEDF_RPORT_TYPE_TAPE 2
+ uint dev_type; /* Disk or tape */
+ struct list_head peers;
+};
+
+/* Used to contain LL2 skb's in ll2_skb_list */
+struct qedf_skb_work {
+ struct list_head list;
+ struct sk_buff *skb;
+};
+
+struct qedf_fastpath {
+#define QEDF_SB_ID_NULL 0xffff
+ u16 sb_id;
+ struct qed_sb_info *sb_info;
+ struct qedf_ctx *qedf;
+ /* Keep track of number of completions on this fastpath */
+ unsigned long completions;
+ uint32_t cq_num_entries;
+};
+
+/* Used to pass fastpath information needed to process CQEs */
+struct qedf_io_work {
+ struct list_head list;
+ struct fcoe_cqe cqe;
+ struct qedf_ctx *qedf;
+ struct fc_frame *fp;
+};
+
+struct qedf_glbl_q_params {
+ u64 hw_p_cq; /* Completion queue PBL */
+ u64 hw_p_rq; /* Request queue PBL */
+ u64 hw_p_cmdq; /* Command queue PBL */
+};
+
+struct global_queue {
+ struct fcoe_cqe *cq;
+ dma_addr_t cq_dma;
+ u32 cq_mem_size;
+ u32 cq_cons_idx; /* Completion queue consumer index */
+ u32 cq_prod_idx;
+
+ void *cq_pbl;
+ dma_addr_t cq_pbl_dma;
+ u32 cq_pbl_size;
+};
+
+/* I/O tracing entry */
+#define QEDF_IO_TRACE_SIZE 2048
+struct qedf_io_log {
+#define QEDF_IO_TRACE_REQ 0
+#define QEDF_IO_TRACE_RSP 1
+ uint8_t direction;
+ uint16_t task_id;
+ uint32_t port_id; /* Remote port fabric ID */
+ int lun;
+ char op; /* SCSI CDB */
+ uint8_t lba[4];
+ unsigned int bufflen; /* SCSI buffer length */
+ unsigned int sg_count; /* Number of SG elements */
+ int result; /* Result passed back to mid-layer */
+ unsigned long jiffies; /* Time stamp when I/O logged */
+ int refcount; /* Reference count for task id */
+ unsigned int req_cpu; /* CPU that the task is queued on */
+ unsigned int int_cpu; /* Interrupt CPU that the task is received on */
+ unsigned int rsp_cpu; /* CPU that task is returned on */
+ u8 sge_type; /* Did we take the slow, single or fast SGE path */
+};
+
+/* Number of entries in BDQ */
+#define QEDF_BDQ_SIZE 256
+#define QEDF_BDQ_BUF_SIZE 2072
+
+/* DMA coherent buffers for BDQ */
+struct qedf_bdq_buf {
+ void *buf_addr;
+ dma_addr_t buf_dma;
+};
+
+/* Main adapter struct */
+struct qedf_ctx {
+ struct qedf_dbg_ctx dbg_ctx;
+ struct fcoe_ctlr ctlr;
+ struct fc_lport *lport;
+ u8 data_src_addr[ETH_ALEN];
+#define QEDF_LINK_DOWN 0
+#define QEDF_LINK_UP 1
+ atomic_t link_state;
+#define QEDF_DCBX_PENDING 0
+#define QEDF_DCBX_DONE 1
+ atomic_t dcbx;
+ uint16_t max_scsi_xid;
+ uint16_t max_els_xid;
+#define QEDF_NULL_VLAN_ID -1
+#define QEDF_FALLBACK_VLAN 1002
+#define QEDF_DEFAULT_PRIO 3
+ int vlan_id;
+ uint vlan_hw_insert:1;
+ struct qed_dev *cdev;
+ struct qed_dev_fcoe_info dev_info;
+ struct qed_int_info int_info;
+ uint16_t last_command;
+ spinlock_t hba_lock;
+ struct pci_dev *pdev;
+ u64 wwnn;
+ u64 wwpn;
+ u8 __aligned(16) mac[ETH_ALEN];
+ struct list_head fcports;
+ atomic_t num_offloads;
+ unsigned int curr_conn_id;
+ spinlock_t ll2_lock;
+ struct list_head ll2_skb_list;
+ struct task_struct *ll2_recv_thread;
+ struct workqueue_struct *link_update_wq;
+ struct delayed_work link_update;
+ struct delayed_work link_recovery;
+ struct completion flogi_compl;
+ struct completion fipvlan_compl;
+
+ /*
+ * Used to tell if we're in the window where we are waiting for
+ * the link to come back up before informting fcoe that the link is
+ * done.
+ */
+ atomic_t link_down_tmo_valid;
+#define QEDF_TIMER_INTERVAL (1 * HZ)
+ struct timer_list timer; /* One second book keeping timer */
+#define QEDF_DRAIN_ACTIVE 1
+#define QEDF_LL2_STARTED 2
+#define QEDF_UNLOADING 3
+#define QEDF_GRCDUMP_CAPTURE 4
+#define QEDF_IN_RECOVERY 5
+ unsigned long flags; /* Miscellaneous state flags */
+ int fipvlan_retries;
+ u8 num_queues;
+ struct global_queue **global_queues;
+ /* Pointer to array of queue structures */
+ struct qedf_glbl_q_params *p_cpuq;
+ /* Physical address of array of queue structures */
+ dma_addr_t hw_p_cpuq;
+
+ struct qedf_bdq_buf bdq[QEDF_BDQ_SIZE];
+ void *bdq_pbl;
+ dma_addr_t bdq_pbl_dma;
+ size_t bdq_pbl_mem_size;
+ void *bdq_pbl_list;
+ dma_addr_t bdq_pbl_list_dma;
+ u8 bdq_pbl_list_num_entries;
+ void __iomem *bdq_primary_prod;
+ void __iomem *bdq_secondary_prod;
+ uint16_t bdq_prod_idx;
+
+ /* Structure for holding all the fastpath for this qedf_ctx */
+ struct qedf_fastpath *fp_array;
+ struct qed_fcoe_tid tasks;
+ struct qedf_cmd_mgr *cmd_mgr;
+ /* Holds the PF parameters we pass to qed to start he FCoE function */
+ struct qed_pf_params pf_params;
+ /* Used to time middle path ELS and TM commands */
+ struct workqueue_struct *timer_work_queue;
+
+#define QEDF_IO_WORK_MIN 64
+ mempool_t *io_mempool;
+ struct workqueue_struct *dpc_wq;
+
+ u32 slow_sge_ios;
+ u32 fast_sge_ios;
+ u32 single_sge_ios;
+
+ uint8_t *grcdump;
+ uint32_t grcdump_size;
+
+ struct qedf_io_log io_trace_buf[QEDF_IO_TRACE_SIZE];
+ spinlock_t io_trace_lock;
+ uint16_t io_trace_idx;
+
+ bool stop_io_on_error;
+
+ u32 flogi_cnt;
+ u32 flogi_failed;
+
+ /* Used for fc statistics */
+ u64 input_requests;
+ u64 output_requests;
+ u64 control_requests;
+ u64 packet_aborts;
+ u64 alloc_failures;
+};
+
+/*
+ * 4 regs size $$KEEP_ENDIANNESS$$
+ */
+
+struct io_bdt {
+ struct qedf_ioreq *io_req;
+ struct fcoe_sge *bd_tbl;
+ dma_addr_t bd_tbl_dma;
+ u16 bd_valid;
+};
+
+struct qedf_cmd_mgr {
+ struct qedf_ctx *qedf;
+ u16 idx;
+ struct io_bdt **io_bdt_pool;
+#define FCOE_PARAMS_NUM_TASKS 4096
+ struct qedf_ioreq cmds[FCOE_PARAMS_NUM_TASKS];
+ spinlock_t lock;
+ atomic_t free_list_cnt;
+};
+
+/* Stolen from qed_cxt_api.h and adapted for qed_fcoe_info
+ * Usage:
+ *
+ * void *ptr;
+ * ptr = qedf_get_task_mem(&qedf->tasks, 128);
+ */
+static inline void *qedf_get_task_mem(struct qed_fcoe_tid *info, u32 tid)
+{
+ return (void *)(info->blocks[tid / info->num_tids_per_block] +
+ (tid % info->num_tids_per_block) * info->size);
+}
+
+static inline void qedf_stop_all_io(struct qedf_ctx *qedf)
+{
+ set_bit(QEDF_UNLOADING, &qedf->flags);
+}
+
+/*
+ * Externs
+ */
+#define QEDF_DEFAULT_LOG_MASK 0x3CFB6
+extern const struct qed_fcoe_ops *qed_ops;
+extern uint qedf_dump_frames;
+extern uint qedf_io_tracing;
+extern uint qedf_stop_io_on_error;
+extern uint qedf_link_down_tmo;
+#define QEDF_RETRY_DELAY_MAX 20 /* 2 seconds */
+extern bool qedf_retry_delay;
+extern uint qedf_debug;
+
+extern struct qedf_cmd_mgr *qedf_cmd_mgr_alloc(struct qedf_ctx *qedf);
+extern void qedf_cmd_mgr_free(struct qedf_cmd_mgr *cmgr);
+extern int qedf_queuecommand(struct Scsi_Host *host,
+ struct scsi_cmnd *sc_cmd);
+extern void qedf_fip_send(struct fcoe_ctlr *fip, struct sk_buff *skb);
+extern void qedf_update_src_mac(struct fc_lport *lport, u8 *addr);
+extern u8 *qedf_get_src_mac(struct fc_lport *lport);
+extern void qedf_fip_recv(struct qedf_ctx *qedf, struct sk_buff *skb);
+extern void qedf_fcoe_send_vlan_req(struct qedf_ctx *qedf);
+extern void qedf_scsi_completion(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req);
+extern void qedf_process_warning_compl(struct qedf_ctx *qedf,
+ struct fcoe_cqe *cqe,struct qedf_ioreq *io_req);
+extern void qedf_process_error_detect(struct qedf_ctx *qedf,
+ struct fcoe_cqe *cqe, struct qedf_ioreq *io_req);
+extern void qedf_flush_active_ios(struct qedf_rport *fcport, int lun);
+extern void qedf_release_cmd(struct kref *ref);
+extern int qedf_initiate_abts(struct qedf_ioreq *io_req,
+ bool return_scsi_cmd_on_abts);
+extern void qedf_process_abts_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req);
+extern struct qedf_ioreq *qedf_alloc_cmd(struct qedf_rport *fcport,
+ u8 cmd_type);
+
+extern struct device_attribute *qedf_host_attrs[];
+extern void qedf_cmd_timer_set(struct qedf_ctx *qedf, struct qedf_ioreq *io_req,
+ unsigned int timer_msec);
+extern int qedf_init_mp_req(struct qedf_ioreq *io_req);
+extern void qedf_init_mp_task(struct qedf_ioreq *io_req,
+ struct fcoe_task_context *task_ctx);
+extern void qedf_add_to_sq(struct qedf_rport *fcport, u16 xid,
+ u32 ptu_invalidate, enum fcoe_task_type req_type, u32 offset);
+extern void qedf_ring_doorbell(struct qedf_rport *fcport);
+extern void qedf_process_els_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *els_req);
+extern int qedf_send_rrq(struct qedf_ioreq *aborted_io_req);
+extern int qedf_send_adisc(struct qedf_rport *fcport, struct fc_frame *fp);
+extern int qedf_initiate_cleanup(struct qedf_ioreq *io_req,
+ bool return_scsi_cmd_on_abts);
+extern void qedf_process_cleanup_compl(struct qedf_ctx *qedf,
+ struct fcoe_cqe *cqe, struct qedf_ioreq *io_req);
+extern int qedf_initiate_tmf(struct scsi_cmnd *sc_cmd, u8 tm_flags);
+extern void qedf_process_tmf_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req);
+extern void qedf_process_cqe(struct qedf_ctx *qedf, struct fcoe_cqe *cqe);
+extern void qedf_scsi_done(struct qedf_ctx *qedf, struct qedf_ioreq *io_req,
+ int result);
+extern void qedf_set_vlan_id(struct qedf_ctx *qedf, int vlan_id);
+extern void qedf_create_sysfs_ctx_attr(struct qedf_ctx *qedf);
+extern void qedf_remove_sysfs_ctx_attr(struct qedf_ctx *qedf);
+extern void qedf_capture_grc_dump(struct qedf_ctx *qedf);
+extern void qedf_wait_for_upload(struct qedf_ctx *qedf);
+extern void qedf_process_unsol_compl(struct qedf_ctx *qedf, uint16_t que_idx,
+ struct fcoe_cqe *cqe);
+extern void qedf_inc_percpu_requests(unsigned long cpu);
+extern void qedf_restart_rport(struct qedf_rport *fcport);
+extern int qedf_send_rec(struct qedf_ioreq *orig_io_req);
+extern int qedf_post_io_req(struct qedf_rport *fcport,
+ struct qedf_ioreq *io_req);
+extern void qedf_process_seq_cleanup_compl(struct qedf_ctx *qedf,
+ struct fcoe_cqe *cqe, struct qedf_ioreq *io_req);
+extern int qedf_send_flogi(struct qedf_ctx *qedf);
+
+#define FCOE_WORD_TO_BYTE 4
+#define QEDF_MAX_TASK_NUM 0xFFFF
+
+struct fip_vlan {
+ struct ethhdr eth;
+ struct fip_header fip;
+ struct {
+ struct fip_mac_desc mac;
+ struct fip_wwn_desc wwnn;
+ } desc;
+};
+
+/* SQ/CQ Sizes */
+#define GBL_RSVD_TASKS 16
+#define NUM_TASKS_PER_CONNECTION 1024
+#define NUM_RW_TASKS_PER_CONNECTION 512
+#define FCOE_PARAMS_CQ_NUM_ENTRIES FCOE_PARAMS_NUM_TASKS
+
+#define FCOE_PARAMS_CMDQ_NUM_ENTRIES FCOE_PARAMS_NUM_TASKS
+#define SQ_NUM_ENTRIES NUM_TASKS_PER_CONNECTION
+
+#define QEDF_FCOE_PARAMS_GL_RQ_PI 0
+#define QEDF_FCOE_PARAMS_GL_CMD_PI 1
+
+#define QEDF_READ (1 << 1)
+#define QEDF_WRITE (1 << 0)
+#define MAX_FIBRE_LUNS 0xffffffff
+
+#define QEDF_MAX_NUM_CQS 8
+
+/*
+ * PCI function probe defines
+ */
+/* Probe/remove called during normal PCI probe */
+#define QEDF_MODE_NORMAL 0
+/* Probe/remove called from qed error recovery */
+#define QEDF_MODE_RECOVERY 1
+
+#define SUPPORTED_25000baseKR_Full (1<<27)
+#define SUPPORTED_50000baseKR2_Full (1<<28)
+#define SUPPORTED_100000baseKR4_Full (1<<29)
+#define SUPPORTED_100000baseCR4_Full (1<<30)
+
+#endif
diff --git a/drivers/scsi/qedf/qedf_attr.c b/drivers/scsi/qedf/qedf_attr.c
new file mode 100644
index 0000000..4772061
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_attr.c
@@ -0,0 +1,165 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#include "qedf.h"
+
+static ssize_t
+qedf_fcoe_mac_show(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct fc_lport *lport = shost_priv(class_to_shost(dev));
+ u32 port_id;
+ u8 lport_src_id[3];
+ u8 fcoe_mac[6];
+
+ port_id = fc_host_port_id(lport->host);
+ lport_src_id[2] = (port_id & 0x000000FF);
+ lport_src_id[1] = (port_id & 0x0000FF00) >> 8;
+ lport_src_id[0] = (port_id & 0x00FF0000) >> 16;
+ fc_fcoe_set_mac(fcoe_mac, lport_src_id);
+
+ return scnprintf(buf, PAGE_SIZE, "%pM\n", fcoe_mac);
+}
+
+static DEVICE_ATTR(fcoe_mac, S_IRUGO, qedf_fcoe_mac_show, NULL);
+
+struct device_attribute *qedf_host_attrs[] = {
+ &dev_attr_fcoe_mac,
+ NULL,
+};
+
+extern const struct qed_fcoe_ops *qed_ops;
+
+inline bool qedf_is_vport(struct qedf_ctx *qedf)
+{
+ return (!(qedf->lport->vport == NULL));
+}
+
+/* Get base qedf for physical port from vport */
+static struct qedf_ctx *qedf_get_base_qedf(struct qedf_ctx *qedf)
+{
+ struct fc_lport *lport;
+ struct fc_lport *base_lport;
+
+ if (!(qedf_is_vport(qedf)))
+ return NULL;
+
+ lport = qedf->lport;
+ base_lport = shost_priv(vport_to_shost(lport->vport));
+ return (struct qedf_ctx *)(lport_priv(base_lport));
+}
+
+void qedf_capture_grc_dump(struct qedf_ctx *qedf)
+{
+ struct qedf_ctx *base_qedf;
+
+ /* Make sure we use the base qedf to take the GRC dump */
+ if (qedf_is_vport(qedf))
+ base_qedf = qedf_get_base_qedf(qedf);
+ else
+ base_qedf = qedf;
+
+ if (test_bit(QEDF_GRCDUMP_CAPTURE, &base_qedf->flags)) {
+ QEDF_INFO(&(base_qedf->dbg_ctx), QEDF_LOG_INFO,
+ "GRC Dump already captured.\n");
+ return;
+ }
+
+
+ qedf_get_grc_dump(base_qedf->cdev, qed_ops->common,
+ &base_qedf->grcdump, &base_qedf->grcdump_size);
+ QEDF_ERR(&(base_qedf->dbg_ctx), "GRC Dump captured.\n");
+ set_bit(QEDF_GRCDUMP_CAPTURE, &base_qedf->flags);
+ qedf_uevent_emit(base_qedf->lport->host, QEDF_UEVENT_CODE_GRCDUMP,
+ NULL);
+}
+
+static ssize_t
+qedf_sysfs_read_grcdump(struct file *filep, struct kobject *kobj,
+ struct bin_attribute *ba, char *buf, loff_t off,
+ size_t count)
+{
+ ssize_t ret = 0;
+ struct fc_lport *lport = shost_priv(dev_to_shost(container_of(kobj,
+ struct device, kobj)));
+ struct qedf_ctx *qedf = lport_priv(lport);
+
+ if (test_bit(QEDF_GRCDUMP_CAPTURE, &qedf->flags)) {
+ ret = memory_read_from_buffer(buf, count, &off,
+ qedf->grcdump, qedf->grcdump_size);
+ } else {
+ QEDF_ERR(&(qedf->dbg_ctx), "GRC Dump not captured!\n");
+ }
+
+ return ret;
+}
+
+static ssize_t
+qedf_sysfs_write_grcdump(struct file *filep, struct kobject *kobj,
+ struct bin_attribute *ba, char *buf, loff_t off,
+ size_t count)
+{
+ struct fc_lport *lport = NULL;
+ struct qedf_ctx *qedf = NULL;
+ long reading;
+ int ret = 0;
+ char msg[40];
+
+ if (off != 0)
+ return ret;
+
+
+ lport = shost_priv(dev_to_shost(container_of(kobj,
+ struct device, kobj)));
+ qedf = lport_priv(lport);
+
+ buf[1] = 0;
+ ret = kstrtol(buf, 10, &reading);
+ if (ret) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Invalid input, err(%d)\n", ret);
+ return ret;
+ }
+
+ memset(msg, 0, sizeof(msg));
+ switch (reading) {
+ case 0:
+ memset(qedf->grcdump, 0, qedf->grcdump_size);
+ clear_bit(QEDF_GRCDUMP_CAPTURE, &qedf->flags);
+ break;
+ case 1:
+ qedf_capture_grc_dump(qedf);
+ break;
+ }
+
+ return count;
+}
+
+static struct bin_attribute sysfs_grcdump_attr = {
+ .attr = {
+ .name = "grcdump",
+ .mode = S_IRUSR | S_IWUSR,
+ },
+ .size = 0,
+ .read = qedf_sysfs_read_grcdump,
+ .write = qedf_sysfs_write_grcdump,
+};
+
+static struct sysfs_bin_attrs bin_file_entries[] = {
+ {"grcdump", &sysfs_grcdump_attr},
+ {NULL},
+};
+
+void qedf_create_sysfs_ctx_attr(struct qedf_ctx *qedf)
+{
+ qedf_create_sysfs_attr(qedf->lport->host, bin_file_entries);
+}
+
+void qedf_remove_sysfs_ctx_attr(struct qedf_ctx *qedf)
+{
+ qedf_remove_sysfs_attr(qedf->lport->host, bin_file_entries);
+}
diff --git a/drivers/scsi/qedf/qedf_dbg.c b/drivers/scsi/qedf/qedf_dbg.c
new file mode 100644
index 0000000..e023f5d
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_dbg.c
@@ -0,0 +1,195 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#include "qedf_dbg.h"
+#include <linux/vmalloc.h>
+
+void
+qedf_dbg_err(struct qedf_dbg_ctx *qedf, const char *func, u32 line,
+ const char *fmt, ...)
+{
+ va_list va;
+ struct va_format vaf;
+ char nfunc[32];
+
+ memset(nfunc, 0, sizeof(nfunc));
+ memcpy(nfunc, func, sizeof(nfunc) - 1);
+
+ va_start(va, fmt);
+
+ vaf.fmt = fmt;
+ vaf.va = &va;
+
+ if (likely(qedf) && likely(qedf->pdev))
+ pr_err("[%s]:[%s:%d]:%d: %pV", dev_name(&(qedf->pdev->dev)),
+ nfunc, line, qedf->host_no, &vaf);
+ else
+ pr_err("[0000:00:00.0]:[%s:%d]: %pV", nfunc, line, &vaf);
+
+ va_end(va);
+}
+
+void
+qedf_dbg_warn(struct qedf_dbg_ctx *qedf, const char *func, u32 line,
+ const char *fmt, ...)
+{
+ va_list va;
+ struct va_format vaf;
+ char nfunc[32];
+
+ memset(nfunc, 0, sizeof(nfunc));
+ memcpy(nfunc, func, sizeof(nfunc) - 1);
+
+ va_start(va, fmt);
+
+ vaf.fmt = fmt;
+ vaf.va = &va;
+
+ if (!(qedf_debug & QEDF_LOG_WARN))
+ goto ret;
+
+ if (likely(qedf) && likely(qedf->pdev))
+ pr_warn("[%s]:[%s:%d]:%d: %pV", dev_name(&(qedf->pdev->dev)),
+ nfunc, line, qedf->host_no, &vaf);
+ else
+ pr_warn("[0000:00:00.0]:[%s:%d]: %pV", nfunc, line, &vaf);
+
+ret:
+ va_end(va);
+}
+
+void
+qedf_dbg_notice(struct qedf_dbg_ctx *qedf, const char *func, u32 line,
+ const char *fmt, ...)
+{
+ va_list va;
+ struct va_format vaf;
+ char nfunc[32];
+
+ memset(nfunc, 0, sizeof(nfunc));
+ memcpy(nfunc, func, sizeof(nfunc) - 1);
+
+ va_start(va, fmt);
+
+ vaf.fmt = fmt;
+ vaf.va = &va;
+
+ if (!(qedf_debug & QEDF_LOG_NOTICE))
+ goto ret;
+
+ if (likely(qedf) && likely(qedf->pdev))
+ pr_notice("[%s]:[%s:%d]:%d: %pV",
+ dev_name(&(qedf->pdev->dev)), nfunc, line,
+ qedf->host_no, &vaf);
+ else
+ pr_notice("[0000:00:00.0]:[%s:%d]: %pV", nfunc, line, &vaf);
+
+ret:
+ va_end(va);
+}
+
+void
+qedf_dbg_info(struct qedf_dbg_ctx *qedf, const char *func, u32 line,
+ u32 level, const char *fmt, ...)
+{
+ va_list va;
+ struct va_format vaf;
+ char nfunc[32];
+
+ memset(nfunc, 0, sizeof(nfunc));
+ memcpy(nfunc, func, sizeof(nfunc) - 1);
+
+ va_start(va, fmt);
+
+ vaf.fmt = fmt;
+ vaf.va = &va;
+
+ if (!(qedf_debug & level))
+ goto ret;
+
+ if (likely(qedf) && likely(qedf->pdev))
+ pr_info("[%s]:[%s:%d]:%d: %pV", dev_name(&(qedf->pdev->dev)),
+ nfunc, line, qedf->host_no, &vaf);
+ else
+ pr_info("[0000:00:00.0]:[%s:%d]: %pV", nfunc, line, &vaf);
+
+ret:
+ va_end(va);
+}
+
+int
+qedf_alloc_grc_dump_buf(u8 **buf, uint32_t len)
+{
+ *buf = vmalloc(len);
+ if (!(*buf))
+ return -ENOMEM;
+
+ memset(*buf, 0, len);
+ return 0;
+}
+
+void
+qedf_free_grc_dump_buf(uint8_t **buf)
+{
+ vfree(*buf);
+ *buf = NULL;
+}
+
+int
+qedf_get_grc_dump(struct qed_dev *cdev, const struct qed_common_ops *common,
+ u8 **buf, uint32_t *grcsize)
+{
+ if (!*buf)
+ return -EINVAL;
+
+ return common->dbg_grc(cdev, *buf, grcsize);
+}
+
+void
+qedf_uevent_emit(struct Scsi_Host *shost, u32 code, char *msg)
+{
+ char event_string[40];
+ char *envp[] = {event_string, NULL};
+
+ memset(event_string, 0, sizeof(event_string));
+ switch (code) {
+ case QEDF_UEVENT_CODE_GRCDUMP:
+ if (msg)
+ strncpy(event_string, msg, strlen(msg));
+ else
+ sprintf(event_string, "GRCDUMP=%u", shost->host_no);
+ break;
+ default:
+ /* do nothing */
+ break;
+ }
+
+ kobject_uevent_env(&shost->shost_gendev.kobj, KOBJ_CHANGE, envp);
+}
+
+int
+qedf_create_sysfs_attr(struct Scsi_Host *shost, struct sysfs_bin_attrs *iter)
+{
+ int ret = 0;
+
+ for (; iter->name; iter++) {
+ ret = sysfs_create_bin_file(&shost->shost_gendev.kobj,
+ iter->attr);
+ if (ret)
+ pr_err("Unable to create sysfs %s attr, err(%d).\n",
+ iter->name, ret);
+ }
+ return ret;
+}
+
+void
+qedf_remove_sysfs_attr(struct Scsi_Host *shost, struct sysfs_bin_attrs *iter)
+{
+ for (; iter->name; iter++)
+ sysfs_remove_bin_file(&shost->shost_gendev.kobj, iter->attr);
+}
diff --git a/drivers/scsi/qedf/qedf_dbg.h b/drivers/scsi/qedf/qedf_dbg.h
new file mode 100644
index 0000000..23bd706
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_dbg.h
@@ -0,0 +1,154 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#ifndef _QEDF_DBG_H_
+#define _QEDF_DBG_H_
+
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/compiler.h>
+#include <linux/string.h>
+#include <linux/version.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+#include <scsi/scsi_transport.h>
+#include <linux/fs.h>
+
+#include <linux/qed/common_hsi.h>
+#include <linux/qed/qed_if.h>
+
+extern uint qedf_debug;
+
+/* Debug print level definitions */
+#define QEDF_LOG_DEFAULT 0x1 /* Set default logging mask */
+#define QEDF_LOG_INFO 0x2 /*
+ * Informational logs,
+ * MAC address, WWPN, WWNN
+ */
+#define QEDF_LOG_DISC 0x4 /* Init, discovery, rport */
+#define QEDF_LOG_LL2 0x8 /* LL2, VLAN logs */
+#define QEDF_LOG_CONN 0x10 /* Connection setup, cleanup */
+#define QEDF_LOG_EVT 0x20 /* Events, link, mtu */
+#define QEDF_LOG_TIMER 0x40 /* Timer events */
+#define QEDF_LOG_MP_REQ 0x80 /* Middle Path (MP) logs */
+#define QEDF_LOG_SCSI_TM 0x100 /* SCSI Aborts, Task Mgmt */
+#define QEDF_LOG_UNSOL 0x200 /* unsolicited event logs */
+#define QEDF_LOG_IO 0x400 /* scsi cmd, completion */
+#define QEDF_LOG_MQ 0x800 /* Multi Queue logs */
+#define QEDF_LOG_BSG 0x1000 /* BSG logs */
+#define QEDF_LOG_DEBUGFS 0x2000 /* debugFS logs */
+#define QEDF_LOG_LPORT 0x4000 /* lport logs */
+#define QEDF_LOG_ELS 0x8000 /* ELS logs */
+#define QEDF_LOG_NPIV 0x10000 /* NPIV logs */
+#define QEDF_LOG_SESS 0x20000 /* Conection setup, cleanup */
+#define QEDF_LOG_TID 0x80000 /*
+ * FW TID context acquire
+ * free
+ */
+#define QEDF_TRACK_TID 0x100000 /*
+ * Track TID state. To be
+ * enabled only at module load
+ * and not run-time.
+ */
+#define QEDF_TRACK_CMD_LIST 0x300000 /*
+ * Track active cmd list nodes,
+ * done with reference to TID,
+ * hence TRACK_TID also enabled.
+ */
+#define QEDF_LOG_NOTICE 0x40000000 /* Notice logs */
+#define QEDF_LOG_WARN 0x80000000 /* Warning logs */
+
+/* Debug context structure */
+struct qedf_dbg_ctx {
+ unsigned int host_no;
+ struct pci_dev *pdev;
+#ifdef CONFIG_DEBUG_FS
+ struct dentry *bdf_dentry;
+#endif
+};
+
+#define QEDF_ERR(pdev, fmt, ...) \
+ qedf_dbg_err(pdev, __func__, __LINE__, fmt, ## __VA_ARGS__)
+#define QEDF_WARN(pdev, fmt, ...) \
+ qedf_dbg_warn(pdev, __func__, __LINE__, fmt, ## __VA_ARGS__)
+#define QEDF_NOTICE(pdev, fmt, ...) \
+ qedf_dbg_notice(pdev, __func__, __LINE__, fmt, ## __VA_ARGS__)
+#define QEDF_INFO(pdev, level, fmt, ...) \
+ qedf_dbg_info(pdev, __func__, __LINE__, level, fmt, \
+ ## __VA_ARGS__)
+
+extern void qedf_dbg_err(struct qedf_dbg_ctx *qedf, const char *func, u32 line,
+ const char *fmt, ...);
+extern void qedf_dbg_warn(struct qedf_dbg_ctx *qedf, const char *func, u32 line,
+ const char *, ...);
+extern void qedf_dbg_notice(struct qedf_dbg_ctx *qedf, const char *func,
+ u32 line, const char *, ...);
+extern void qedf_dbg_info(struct qedf_dbg_ctx *qedf, const char *func, u32 line,
+ u32 info, const char *fmt, ...);
+
+/* GRC Dump related defines */
+
+struct Scsi_Host;
+
+#define QEDF_UEVENT_CODE_GRCDUMP 0
+
+struct sysfs_bin_attrs {
+ char *name;
+ struct bin_attribute *attr;
+};
+
+extern int qedf_alloc_grc_dump_buf(uint8_t **buf, uint32_t len);
+extern void qedf_free_grc_dump_buf(uint8_t **buf);
+extern int qedf_get_grc_dump(struct qed_dev *cdev,
+ const struct qed_common_ops *common, uint8_t **buf,
+ uint32_t *grcsize);
+extern void qedf_uevent_emit(struct Scsi_Host *shost, u32 code, char *msg);
+extern int qedf_create_sysfs_attr(struct Scsi_Host *shost,
+ struct sysfs_bin_attrs *iter);
+extern void qedf_remove_sysfs_attr(struct Scsi_Host *shost,
+ struct sysfs_bin_attrs *iter);
+
+#ifdef CONFIG_DEBUG_FS
+/* DebugFS related code */
+struct qedf_list_of_funcs {
+ char *oper_str;
+ ssize_t (*oper_func)(struct qedf_dbg_ctx *qedf);
+};
+
+struct qedf_debugfs_ops {
+ char *name;
+ struct qedf_list_of_funcs *qedf_funcs;
+};
+
+#define qedf_dbg_fileops(drv, ops) \
+{ \
+ .owner = THIS_MODULE, \
+ .open = simple_open, \
+ .read = drv##_dbg_##ops##_cmd_read, \
+ .write = drv##_dbg_##ops##_cmd_write \
+}
+
+/* Used for debugfs sequential files */
+#define qedf_dbg_fileops_seq(drv, ops) \
+{ \
+ .owner = THIS_MODULE, \
+ .open = drv##_dbg_##ops##_open, \
+ .read = seq_read, \
+ .llseek = seq_lseek, \
+ .release = single_release, \
+}
+
+extern void qedf_dbg_host_init(struct qedf_dbg_ctx *qedf,
+ struct qedf_debugfs_ops *dops,
+ struct file_operations *fops);
+extern void qedf_dbg_host_exit(struct qedf_dbg_ctx *qedf);
+extern void qedf_dbg_init(char *drv_name);
+extern void qedf_dbg_exit(void);
+#endif /* CONFIG_DEBUG_FS */
+
+#endif /* _QEDF_DBG_H_ */
diff --git a/drivers/scsi/qedf/qedf_debugfs.c b/drivers/scsi/qedf/qedf_debugfs.c
new file mode 100644
index 0000000..8245b8c
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_debugfs.c
@@ -0,0 +1,471 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 QLogic Corporation
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#ifdef CONFIG_DEBUG_FS
+
+#include <linux/uaccess.h>
+#include <linux/debugfs.h>
+#include <linux/module.h>
+
+#include "qedf.h"
+#include "qedf_dbg.h"
+
+static struct dentry *qedf_dbg_root;
+
+/**
+ * qedf_dbg_host_init - setup the debugfs file for the pf
+ * @pf: the pf that is starting up
+ **/
+void
+qedf_dbg_host_init(struct qedf_dbg_ctx *qedf,
+ struct qedf_debugfs_ops *dops,
+ struct file_operations *fops)
+{
+ char host_dirname[32];
+ struct dentry *file_dentry = NULL;
+
+ QEDF_INFO(qedf, QEDF_LOG_DEBUGFS, "Creating debugfs host node\n");
+ /* create pf dir */
+ sprintf(host_dirname, "host%u", qedf->host_no);
+ qedf->bdf_dentry = debugfs_create_dir(host_dirname, qedf_dbg_root);
+ if (!qedf->bdf_dentry)
+ return;
+
+ /* create debugfs files */
+ while (dops) {
+ if (!(dops->name))
+ break;
+
+ file_dentry = debugfs_create_file(dops->name, 0600,
+ qedf->bdf_dentry, qedf,
+ fops);
+ if (!file_dentry) {
+ QEDF_INFO(qedf, QEDF_LOG_DEBUGFS,
+ "Debugfs entry %s creation failed\n",
+ dops->name);
+ debugfs_remove_recursive(qedf->bdf_dentry);
+ return;
+ }
+ dops++;
+ fops++;
+ }
+}
+
+/**
+ * qedf_dbg_host_exit - clear out the pf's debugfs entries
+ * @pf: the pf that is stopping
+ **/
+void
+qedf_dbg_host_exit(struct qedf_dbg_ctx *qedf)
+{
+ QEDF_INFO(qedf, QEDF_LOG_DEBUGFS, "Destroying debugfs host "
+ "entry\n");
+ /* remove debugfs entries of this PF */
+ debugfs_remove_recursive(qedf->bdf_dentry);
+ qedf->bdf_dentry = NULL;
+}
+
+/**
+ * qedf_dbg_init - start up debugfs for the driver
+ **/
+void
+qedf_dbg_init(char *drv_name)
+{
+ QEDF_INFO(NULL, QEDF_LOG_DEBUGFS, "Creating debugfs root node\n");
+
+ /* create qed dir in root of debugfs. NULL means debugfs root */
+ qedf_dbg_root = debugfs_create_dir(drv_name, NULL);
+ if (!qedf_dbg_root)
+ QEDF_INFO(NULL, QEDF_LOG_DEBUGFS, "Init of debugfs "
+ "failed\n");
+}
+
+/**
+ * qedf_dbg_exit - clean out the driver's debugfs entries
+ **/
+void
+qedf_dbg_exit(void)
+{
+ QEDF_INFO(NULL, QEDF_LOG_DEBUGFS, "Destroying debugfs root "
+ "entry\n");
+
+ /* remove qed dir in root of debugfs */
+ debugfs_remove_recursive(qedf_dbg_root);
+ qedf_dbg_root = NULL;
+}
+
+struct qedf_debugfs_ops qedf_debugfs_ops[] = {
+ { "fp_int", NULL },
+ { "io_trace", NULL },
+ { "debug", NULL },
+ { "stop_io_on_error", NULL},
+ { "driver_stats", NULL},
+ { "clear_stats", NULL},
+ { "offload_stats", NULL},
+ /* This must be last */
+ { NULL, NULL }
+};
+
+DECLARE_PER_CPU(struct qedf_percpu_iothread_s, qedf_percpu_iothreads);
+
+static ssize_t
+qedf_dbg_fp_int_cmd_read(struct file *filp, char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ size_t cnt = 0;
+ int id;
+ struct qedf_fastpath *fp = NULL;
+ struct qedf_dbg_ctx *qedf_dbg =
+ (struct qedf_dbg_ctx *)filp->private_data;
+ struct qedf_ctx *qedf = container_of(qedf_dbg,
+ struct qedf_ctx, dbg_ctx);
+ struct qedf_percpu_iothread_s *iothread;
+ unsigned int cpu;
+
+ QEDF_INFO(qedf_dbg, QEDF_LOG_DEBUGFS, "entered\n");
+
+ cnt = sprintf(buffer, "\nFastpath I/O completions\n\n");
+
+ for (id = 0; id < qedf->num_queues; id++) {
+ fp = &(qedf->fp_array[id]);
+ if (fp->sb_id == QEDF_SB_ID_NULL)
+ continue;
+ cnt += sprintf((buffer + cnt), "#%d: %lu\n", id,
+ fp->completions);
+ }
+
+ cnt += sprintf(buffer + cnt, "\nPer CPU Stats\n\n");
+
+ for_each_online_cpu(cpu) {
+ iothread = &per_cpu(qedf_percpu_iothreads, cpu);
+ cnt += sprintf((buffer + cnt),
+ "#%u: requests=%llu responses=%llu\n",
+ cpu, iothread->requests, iothread->responses);
+ }
+
+ cnt = min_t(int, count, cnt - *ppos);
+ *ppos += cnt;
+ return cnt;
+}
+
+static ssize_t
+qedf_dbg_fp_int_cmd_write(struct file *filp, const char __user *buffer,
+ size_t count, loff_t *ppos)
+{
+ if (!count || *ppos)
+ return 0;
+
+ return count;
+}
+
+static ssize_t
+qedf_dbg_debug_cmd_read(struct file *filp, char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ int cnt;
+ struct qedf_dbg_ctx *qedf =
+ (struct qedf_dbg_ctx *)filp->private_data;
+
+ QEDF_INFO(qedf, QEDF_LOG_DEBUGFS, "entered\n");
+ cnt = sprintf(buffer, "debug mask = 0x%x\n", qedf_debug);
+
+ cnt = min_t(int, count, cnt - *ppos);
+ *ppos += cnt;
+ return cnt;
+}
+
+static ssize_t
+qedf_dbg_debug_cmd_write(struct file *filp, const char __user *buffer,
+ size_t count, loff_t *ppos)
+{
+ uint32_t val;
+ void *kern_buf;
+ int rval;
+ struct qedf_dbg_ctx *qedf =
+ (struct qedf_dbg_ctx *)filp->private_data;
+
+ if (!count || *ppos)
+ return 0;
+
+ kern_buf = memdup_user(buffer, count);
+ if (IS_ERR(kern_buf))
+ return PTR_ERR(kern_buf);
+
+ rval = kstrtouint(kern_buf, 10, &val);
+ kfree(kern_buf);
+ if (rval)
+ return rval;
+
+ if (val == 1)
+ qedf_debug = QEDF_DEFAULT_LOG_MASK;
+ else
+ qedf_debug = val;
+
+ QEDF_INFO(qedf, QEDF_LOG_DEBUGFS, "Setting debug=0x%x.\n", val);
+ return count;
+}
+
+static ssize_t
+qedf_dbg_stop_io_on_error_cmd_read(struct file *filp, char __user *buffer,
+ size_t count, loff_t *ppos)
+{
+ int cnt;
+ struct qedf_dbg_ctx *qedf_dbg =
+ (struct qedf_dbg_ctx *)filp->private_data;
+ struct qedf_ctx *qedf = container_of(qedf_dbg,
+ struct qedf_ctx, dbg_ctx);
+
+ QEDF_INFO(qedf_dbg, QEDF_LOG_DEBUGFS, "entered\n");
+ cnt = sprintf(buffer, "%s\n",
+ qedf->stop_io_on_error ? "true" : "false");
+
+ cnt = min_t(int, count, cnt - *ppos);
+ *ppos += cnt;
+ return cnt;
+}
+
+static ssize_t
+qedf_dbg_stop_io_on_error_cmd_write(struct file *filp,
+ const char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ void *kern_buf;
+ struct qedf_dbg_ctx *qedf_dbg =
+ (struct qedf_dbg_ctx *)filp->private_data;
+ struct qedf_ctx *qedf = container_of(qedf_dbg, struct qedf_ctx,
+ dbg_ctx);
+
+ QEDF_INFO(qedf_dbg, QEDF_LOG_DEBUGFS, "entered\n");
+
+ if (!count || *ppos)
+ return 0;
+
+ kern_buf = memdup_user(buffer, 6);
+ if (IS_ERR(kern_buf))
+ return PTR_ERR(kern_buf);
+
+ if (strncmp(kern_buf, "false", 5) == 0)
+ qedf->stop_io_on_error = false;
+ else if (strncmp(kern_buf, "true", 4) == 0)
+ qedf->stop_io_on_error = true;
+ else if (strncmp(kern_buf, "now", 3) == 0)
+ /* Trigger from user to stop all I/O on this host */
+ set_bit(QEDF_UNLOADING, &qedf->flags);
+
+ kfree(kern_buf);
+ return count;
+}
+
+static int
+qedf_io_trace_show(struct seq_file *s, void *unused)
+{
+ int i, idx = 0;
+ struct qedf_ctx *qedf = s->private;
+ struct qedf_dbg_ctx *qedf_dbg = &qedf->dbg_ctx;
+ struct qedf_io_log *io_log;
+ unsigned long flags;
+
+ if (!qedf_io_tracing) {
+ seq_puts(s, "I/O tracing not enabled.\n");
+ goto out;
+ }
+
+ QEDF_INFO(qedf_dbg, QEDF_LOG_DEBUGFS, "entered\n");
+
+ spin_lock_irqsave(&qedf->io_trace_lock, flags);
+ idx = qedf->io_trace_idx;
+ for (i = 0; i < QEDF_IO_TRACE_SIZE; i++) {
+ io_log = &qedf->io_trace_buf[idx];
+ seq_printf(s, "%d:", io_log->direction);
+ seq_printf(s, "0x%x:", io_log->task_id);
+ seq_printf(s, "0x%06x:", io_log->port_id);
+ seq_printf(s, "%d:", io_log->lun);
+ seq_printf(s, "0x%02x:", io_log->op);
+ seq_printf(s, "0x%02x%02x%02x%02x:", io_log->lba[0],
+ io_log->lba[1], io_log->lba[2], io_log->lba[3]);
+ seq_printf(s, "%d:", io_log->bufflen);
+ seq_printf(s, "%d:", io_log->sg_count);
+ seq_printf(s, "0x%08x:", io_log->result);
+ seq_printf(s, "%lu:", io_log->jiffies);
+ seq_printf(s, "%d:", io_log->refcount);
+ seq_printf(s, "%d:", io_log->req_cpu);
+ seq_printf(s, "%d:", io_log->int_cpu);
+ seq_printf(s, "%d:", io_log->rsp_cpu);
+ seq_printf(s, "%d\n", io_log->sge_type);
+
+ idx++;
+ if (idx == QEDF_IO_TRACE_SIZE)
+ idx = 0;
+ }
+ spin_unlock_irqrestore(&qedf->io_trace_lock, flags);
+
+out:
+ return 0;
+}
+
+static int
+qedf_dbg_io_trace_open(struct inode *inode, struct file *file)
+{
+ struct qedf_dbg_ctx *qedf_dbg = inode->i_private;
+ struct qedf_ctx *qedf = container_of(qedf_dbg,
+ struct qedf_ctx, dbg_ctx);
+
+ return single_open(file, qedf_io_trace_show, qedf);
+}
+
+static int
+qedf_driver_stats_show(struct seq_file *s, void *unused)
+{
+ struct qedf_ctx *qedf = s->private;
+ struct qedf_rport *fcport;
+ struct fc_rport_priv *rdata;
+
+ seq_printf(s, "cmg_mgr free io_reqs: %d\n",
+ atomic_read(&qedf->cmd_mgr->free_list_cnt));
+ seq_printf(s, "slow SGEs: %d\n", qedf->slow_sge_ios);
+ seq_printf(s, "single SGEs: %d\n", qedf->single_sge_ios);
+ seq_printf(s, "fast SGEs: %d\n\n", qedf->fast_sge_ios);
+
+ seq_puts(s, "Offloaded ports:\n\n");
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(fcport, &qedf->fcports, peers) {
+ rdata = fcport->rdata;
+ if (rdata == NULL)
+ continue;
+ seq_printf(s, "%06x: free_sqes: %d, num_active_ios: %d\n",
+ rdata->ids.port_id, atomic_read(&fcport->free_sqes),
+ atomic_read(&fcport->num_active_ios));
+ }
+ rcu_read_unlock();
+
+ return 0;
+}
+
+static int
+qedf_dbg_driver_stats_open(struct inode *inode, struct file *file)
+{
+ struct qedf_dbg_ctx *qedf_dbg = inode->i_private;
+ struct qedf_ctx *qedf = container_of(qedf_dbg,
+ struct qedf_ctx, dbg_ctx);
+
+ return single_open(file, qedf_driver_stats_show, qedf);
+}
+
+static ssize_t
+qedf_dbg_clear_stats_cmd_read(struct file *filp, char __user *buffer,
+ size_t count, loff_t *ppos)
+{
+ int cnt = 0;
+
+ /* Essentially a read stub */
+ cnt = min_t(int, count, cnt - *ppos);
+ *ppos += cnt;
+ return cnt;
+}
+
+static ssize_t
+qedf_dbg_clear_stats_cmd_write(struct file *filp,
+ const char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ struct qedf_dbg_ctx *qedf_dbg =
+ (struct qedf_dbg_ctx *)filp->private_data;
+ struct qedf_ctx *qedf = container_of(qedf_dbg, struct qedf_ctx,
+ dbg_ctx);
+
+ QEDF_INFO(qedf_dbg, QEDF_LOG_DEBUGFS, "Clearing stat counters.\n");
+
+ if (!count || *ppos)
+ return 0;
+
+ /* Clear stat counters exposed by 'stats' node */
+ qedf->slow_sge_ios = 0;
+ qedf->single_sge_ios = 0;
+ qedf->fast_sge_ios = 0;
+
+ return count;
+}
+
+static int
+qedf_offload_stats_show(struct seq_file *s, void *unused)
+{
+ struct qedf_ctx *qedf = s->private;
+ struct qed_fcoe_stats *fw_fcoe_stats;
+
+ fw_fcoe_stats = kmalloc(sizeof(struct qed_fcoe_stats), GFP_KERNEL);
+ if (!fw_fcoe_stats) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not allocate memory for "
+ "fw_fcoe_stats.\n");
+ goto out;
+ }
+
+ /* Query firmware for offload stats */
+ qed_ops->get_stats(qedf->cdev, fw_fcoe_stats);
+
+ seq_printf(s, "fcoe_rx_byte_cnt=%llu\n"
+ "fcoe_rx_data_pkt_cnt=%llu\n"
+ "fcoe_rx_xfer_pkt_cnt=%llu\n"
+ "fcoe_rx_other_pkt_cnt=%llu\n"
+ "fcoe_silent_drop_pkt_cmdq_full_cnt=%u\n"
+ "fcoe_silent_drop_pkt_crc_error_cnt=%u\n"
+ "fcoe_silent_drop_pkt_task_invalid_cnt=%u\n"
+ "fcoe_silent_drop_total_pkt_cnt=%u\n"
+ "fcoe_silent_drop_pkt_rq_full_cnt=%u\n"
+ "fcoe_tx_byte_cnt=%llu\n"
+ "fcoe_tx_data_pkt_cnt=%llu\n"
+ "fcoe_tx_xfer_pkt_cnt=%llu\n"
+ "fcoe_tx_other_pkt_cnt=%llu\n",
+ fw_fcoe_stats->fcoe_rx_byte_cnt,
+ fw_fcoe_stats->fcoe_rx_data_pkt_cnt,
+ fw_fcoe_stats->fcoe_rx_xfer_pkt_cnt,
+ fw_fcoe_stats->fcoe_rx_other_pkt_cnt,
+ fw_fcoe_stats->fcoe_silent_drop_pkt_cmdq_full_cnt,
+ fw_fcoe_stats->fcoe_silent_drop_pkt_crc_error_cnt,
+ fw_fcoe_stats->fcoe_silent_drop_pkt_task_invalid_cnt,
+ fw_fcoe_stats->fcoe_silent_drop_total_pkt_cnt,
+ fw_fcoe_stats->fcoe_silent_drop_pkt_rq_full_cnt,
+ fw_fcoe_stats->fcoe_tx_byte_cnt,
+ fw_fcoe_stats->fcoe_tx_data_pkt_cnt,
+ fw_fcoe_stats->fcoe_tx_xfer_pkt_cnt,
+ fw_fcoe_stats->fcoe_tx_other_pkt_cnt);
+
+ kfree(fw_fcoe_stats);
+out:
+ return 0;
+}
+
+static int
+qedf_dbg_offload_stats_open(struct inode *inode, struct file *file)
+{
+ struct qedf_dbg_ctx *qedf_dbg = inode->i_private;
+ struct qedf_ctx *qedf = container_of(qedf_dbg,
+ struct qedf_ctx, dbg_ctx);
+
+ return single_open(file, qedf_offload_stats_show, qedf);
+}
+
+
+const struct file_operations qedf_dbg_fops[] = {
+ qedf_dbg_fileops(qedf, fp_int),
+ qedf_dbg_fileops_seq(qedf, io_trace),
+ qedf_dbg_fileops(qedf, debug),
+ qedf_dbg_fileops(qedf, stop_io_on_error),
+ qedf_dbg_fileops_seq(qedf, driver_stats),
+ qedf_dbg_fileops(qedf, clear_stats),
+ qedf_dbg_fileops_seq(qedf, offload_stats),
+ /* This must be last */
+ { NULL, NULL },
+};
+
+#else /* CONFIG_DEBUG_FS */
+void qedf_dbg_host_init(struct qedf_dbg_ctx *);
+void qedf_dbg_host_exit(struct qedf_dbg_ctx *);
+void qedf_dbg_init(char *);
+void qedf_dbg_exit(void);
+#endif /* CONFIG_DEBUG_FS */
diff --git a/drivers/scsi/qedf/qedf_els.c b/drivers/scsi/qedf/qedf_els.c
new file mode 100644
index 0000000..7b9d8d0
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_els.c
@@ -0,0 +1,984 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#include "qedf.h"
+
+/* It's assumed that the lock is held when calling this function. */
+static int qedf_initiate_els(struct qedf_rport *fcport, unsigned int op,
+ void *data, uint32_t data_len,
+ void (*cb_func)(struct qedf_els_cb_arg *cb_arg),
+ struct qedf_els_cb_arg *cb_arg, uint32_t timer_msec)
+{
+ struct qedf_ctx *qedf = fcport->qedf;
+ struct fc_lport *lport = qedf->lport;
+ struct qedf_ioreq *els_req;
+ struct qedf_mp_req *mp_req;
+ struct fc_frame_header *fc_hdr;
+ struct fcoe_task_context *task;
+ int rc = 0;
+ uint32_t did, sid;
+ uint16_t xid;
+ uint32_t start_time = jiffies / HZ;
+ uint32_t current_time;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Sending ELS\n");
+
+ rc = fc_remote_port_chkready(fcport->rport);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "els 0x%x: rport not ready\n", op);
+ rc = -EAGAIN;
+ goto els_err;
+ }
+ if (lport->state != LPORT_ST_READY || !(lport->link_up)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "els 0x%x: link is not ready\n",
+ op);
+ rc = -EAGAIN;
+ goto els_err;
+ }
+
+ if (!(test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags))) {
+ QEDF_ERR(&(qedf->dbg_ctx), "els 0x%x: fcport not ready\n", op);
+ rc = -EINVAL;
+ goto els_err;
+ }
+
+retry_els:
+ els_req = qedf_alloc_cmd(fcport, QEDF_ELS);
+ if (!els_req) {
+ current_time = jiffies / HZ;
+ if ((current_time - start_time) > 10) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "els: Failed els 0x%x\n", op);
+ rc = -ENOMEM;
+ goto els_err;
+ }
+ mdelay(20 * USEC_PER_MSEC);
+ goto retry_els;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "initiate_els els_req = "
+ "0x%p cb_arg = %p xid = %x\n", els_req, cb_arg,
+ els_req->xid);
+ els_req->sc_cmd = NULL;
+ els_req->cmd_type = QEDF_ELS;
+ els_req->fcport = fcport;
+ els_req->cb_func = cb_func;
+ cb_arg->io_req = els_req;
+ cb_arg->op = op;
+ els_req->cb_arg = cb_arg;
+ els_req->data_xfer_len = data_len;
+
+ /* Record which cpu this request is associated with */
+ els_req->cpu = smp_processor_id();
+ qedf_inc_percpu_requests(els_req->cpu);
+
+ mp_req = (struct qedf_mp_req *)&(els_req->mp_req);
+ rc = qedf_init_mp_req(els_req);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "ELS MP request init failed\n");
+ kref_put(&els_req->refcount, qedf_release_cmd);
+ goto els_err;
+ } else {
+ rc = 0;
+ }
+
+ /* Fill ELS Payload */
+ if ((op >= ELS_LS_RJT) && (op <= ELS_AUTH_ELS)) {
+ memcpy(mp_req->req_buf, data, data_len);
+ } else {
+ QEDF_ERR(&(qedf->dbg_ctx), "Invalid ELS op 0x%x\n", op);
+ els_req->cb_func = NULL;
+ els_req->cb_arg = NULL;
+ kref_put(&els_req->refcount, qedf_release_cmd);
+ rc = -EINVAL;
+ }
+
+ if (rc)
+ goto els_err;
+
+ /* Fill FC header */
+ fc_hdr = &(mp_req->req_fc_hdr);
+
+ did = fcport->rdata->ids.port_id;
+ sid = fcport->sid;
+
+ __fc_fill_fc_hdr(fc_hdr, FC_RCTL_ELS_REQ, sid, did,
+ FC_TYPE_ELS, FC_FC_FIRST_SEQ | FC_FC_END_SEQ |
+ FC_FC_SEQ_INIT, 0);
+
+ /* Obtain exchange id */
+ xid = els_req->xid;
+
+ /* Initialize task context for this IO request */
+ task = qedf_get_task_mem(&qedf->tasks, xid);
+ qedf_init_mp_task(els_req, task);
+
+ /* Put timer on original I/O request */
+ if (timer_msec)
+ qedf_cmd_timer_set(qedf, els_req, timer_msec);
+
+ qedf_add_to_sq(fcport, xid, 0, FCOE_TASK_TYPE_MIDPATH, 0);
+
+ /* Ring doorbell */
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Ringing doorbell for ELS "
+ "req\n");
+ qedf_ring_doorbell(fcport);
+els_err:
+ return rc;
+}
+
+void qedf_process_els_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *els_req)
+{
+ struct fcoe_task_context *task_ctx;
+ struct scsi_cmnd *sc_cmd;
+ uint16_t xid;
+ struct fcoe_cqe_midpath_info *mp_info;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Entered with xid = 0x%x"
+ " cmd_type = %d.\n", els_req->xid, els_req->cmd_type);
+
+ /* Kill the ELS timer */
+ cancel_delayed_work(&els_req->timeout_work);
+
+ xid = els_req->xid;
+ task_ctx = qedf_get_task_mem(&qedf->tasks, xid);
+ sc_cmd = els_req->sc_cmd;
+
+ /* Get ELS response length from CQE */
+ mp_info = &cqe->cqe_info.midpath_info;
+ els_req->mp_req.resp_len = mp_info->data_placement_size;
+
+ /* Parse ELS response */
+ if ((els_req->cb_func) && (els_req->cb_arg)) {
+ els_req->cb_func(els_req->cb_arg);
+ els_req->cb_arg = NULL;
+ }
+
+ kref_put(&els_req->refcount, qedf_release_cmd);
+}
+
+static void qedf_rrq_compl(struct qedf_els_cb_arg *cb_arg)
+{
+ struct qedf_ioreq *orig_io_req;
+ struct qedf_ioreq *rrq_req;
+ struct qedf_ctx *qedf;
+ int refcount;
+
+ rrq_req = cb_arg->io_req;
+ qedf = rrq_req->fcport->qedf;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Entered.\n");
+
+ orig_io_req = cb_arg->aborted_io_req;
+
+ if (!orig_io_req)
+ goto out_free;
+
+ if (rrq_req->event != QEDF_IOREQ_EV_ELS_TMO &&
+ rrq_req->event != QEDF_IOREQ_EV_ELS_ERR_DETECT)
+ cancel_delayed_work_sync(&orig_io_req->timeout_work);
+
+ refcount = atomic_read(&orig_io_req->refcount.refcount);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "rrq_compl: orig io = %p,"
+ " orig xid = 0x%x, rrq_xid = 0x%x, refcount=%d\n",
+ orig_io_req, orig_io_req->xid, rrq_req->xid, refcount);
+
+ /* This should return the aborted io_req to the command pool */
+ if (orig_io_req)
+ kref_put(&orig_io_req->refcount, qedf_release_cmd);
+
+out_free:
+ kfree(cb_arg);
+}
+
+/* Assumes kref is already held by caller */
+int qedf_send_rrq(struct qedf_ioreq *aborted_io_req)
+{
+
+ struct fc_els_rrq rrq;
+ struct qedf_rport *fcport;
+ struct fc_lport *lport;
+ struct qedf_els_cb_arg *cb_arg = NULL;
+ struct qedf_ctx *qedf;
+ uint32_t sid;
+ uint32_t r_a_tov;
+ int rc;
+
+ if (!aborted_io_req) {
+ QEDF_ERR(NULL, "abort_io_req is NULL.\n");
+ return -EINVAL;
+ }
+
+ fcport = aborted_io_req->fcport;
+
+ /* Check that fcport is still offloaded */
+ if (!(test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags))) {
+ QEDF_ERR(NULL, "fcport is no longer offloaded.\n");
+ return -EINVAL;
+ }
+
+ if (!fcport->qedf) {
+ QEDF_ERR(NULL, "fcport->qedf is NULL.\n");
+ return -EINVAL;
+ }
+
+ qedf = fcport->qedf;
+ lport = qedf->lport;
+ sid = fcport->sid;
+ r_a_tov = lport->r_a_tov;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Sending RRQ orig "
+ "io = %p, orig_xid = 0x%x\n", aborted_io_req,
+ aborted_io_req->xid);
+ memset(&rrq, 0, sizeof(rrq));
+
+ cb_arg = kzalloc(sizeof(struct qedf_els_cb_arg), GFP_NOIO);
+ if (!cb_arg) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to allocate cb_arg for "
+ "RRQ\n");
+ rc = -ENOMEM;
+ goto rrq_err;
+ }
+
+ cb_arg->aborted_io_req = aborted_io_req;
+
+ rrq.rrq_cmd = ELS_RRQ;
+ hton24(rrq.rrq_s_id, sid);
+ rrq.rrq_ox_id = htons(aborted_io_req->xid);
+ rrq.rrq_rx_id =
+ htons(aborted_io_req->task->tstorm_st_context.read_write.rx_id);
+
+ rc = qedf_initiate_els(fcport, ELS_RRQ, &rrq, sizeof(rrq),
+ qedf_rrq_compl, cb_arg, r_a_tov);
+
+rrq_err:
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "RRQ failed - release orig io "
+ "req 0x%x\n", aborted_io_req->xid);
+ kfree(cb_arg);
+ kref_put(&aborted_io_req->refcount, qedf_release_cmd);
+ }
+ return rc;
+}
+
+static void qedf_process_l2_frame_compl(struct qedf_rport *fcport,
+ unsigned char *buf,
+ u32 frame_len, u16 l2_oxid)
+{
+ struct fc_lport *lport = fcport->qedf->lport;
+ struct fc_frame_header *fh;
+ struct fc_frame *fp;
+ u32 payload_len;
+ u32 crc;
+
+ payload_len = frame_len - sizeof(struct fc_frame_header);
+
+ fp = fc_frame_alloc(lport, payload_len);
+ if (!fp) {
+ QEDF_ERR(&(fcport->qedf->dbg_ctx),
+ "fc_frame_alloc failure.\n");
+ return;
+ }
+
+ /* Copy FC Frame header and payload into the frame */
+ fh = (struct fc_frame_header *)fc_frame_header_get(fp);
+ memcpy(fh, buf, frame_len);
+
+ /* Set the OXID we return to what libfc used */
+ if (l2_oxid != FC_XID_UNKNOWN)
+ fh->fh_ox_id = htons(l2_oxid);
+
+ /* Setup header fields */
+ fh->fh_r_ctl = FC_RCTL_ELS_REP;
+ fh->fh_type = FC_TYPE_ELS;
+ /* Last sequence, end sequence */
+ fh->fh_f_ctl[0] = 0x98;
+ hton24(fh->fh_d_id, lport->port_id);
+ hton24(fh->fh_s_id, fcport->rdata->ids.port_id);
+ fh->fh_rx_id = 0xffff;
+
+ /* Set frame attributes */
+ crc = fcoe_fc_crc(fp);
+ fc_frame_init(fp);
+ fr_dev(fp) = lport;
+ fr_sof(fp) = FC_SOF_I3;
+ fr_eof(fp) = FC_EOF_T;
+ fr_crc(fp) = cpu_to_le32(~crc);
+
+ /* Send completed request to libfc */
+ fc_exch_recv(lport, fp);
+}
+
+/*
+ * In instances where an ELS command times out we may need to restart the
+ * rport by logging out and then logging back in.
+ */
+void qedf_restart_rport(struct qedf_rport *fcport)
+{
+ struct fc_lport *lport;
+ struct fc_rport_priv *rdata;
+ u32 port_id;
+
+ if (!fcport)
+ return;
+
+ rdata = fcport->rdata;
+ if (rdata) {
+ lport = fcport->qedf->lport;
+ port_id = rdata->ids.port_id;
+ QEDF_ERR(&(fcport->qedf->dbg_ctx),
+ "LOGO port_id=%x.\n", port_id);
+ mutex_lock(&lport->disc.disc_mutex);
+ fc_rport_logoff(rdata);
+ /* Recreate the rport and log back in */
+ rdata = fc_rport_create(lport, port_id);
+ if (rdata)
+ fc_rport_login(rdata);
+ mutex_unlock(&lport->disc.disc_mutex);
+ }
+}
+
+static void qedf_l2_els_compl(struct qedf_els_cb_arg *cb_arg)
+{
+ struct qedf_ioreq *els_req;
+ struct qedf_rport *fcport;
+ struct qedf_mp_req *mp_req;
+ struct fc_frame_header *fc_hdr;
+ unsigned char *buf;
+ void *resp_buf;
+ u32 resp_len, hdr_len;
+ u16 l2_oxid;
+ int frame_len;
+
+ l2_oxid = cb_arg->l2_oxid;
+ els_req = cb_arg->io_req;
+
+ if (!els_req) {
+ QEDF_ERR(NULL, "els_req is NULL.\n");
+ goto free_arg;
+ }
+
+ /*
+ * If we are flushing the command just free the cb_arg as none of the
+ * response data will be valid.
+ */
+ if (els_req->event == QEDF_IOREQ_EV_ELS_FLUSH)
+ goto free_arg;
+
+ fcport = els_req->fcport;
+ mp_req = &(els_req->mp_req);
+ fc_hdr = &(mp_req->resp_fc_hdr);
+ resp_len = mp_req->resp_len;
+ resp_buf = mp_req->resp_buf;
+
+ /*
+ * If a middle path ELS command times out, don't try to return
+ * the command but rather do any internal cleanup and then libfc
+ * timeout the command and clean up its internal resources.
+ */
+ if (els_req->event == QEDF_IOREQ_EV_ELS_TMO) {
+ /*
+ * If ADISC times out, libfc will timeout the exchange and then
+ * try to send a PLOGI which will timeout since the session is
+ * still offloaded. Force libfc to logout the session which
+ * will offload the connection and allow the PLOGI response to
+ * flow over the LL2 path.
+ */
+ if (cb_arg->op == ELS_ADISC)
+ qedf_restart_rport(fcport);
+ return;
+ }
+
+ buf = kzalloc(QEDF_PAGE_SIZE, GFP_ATOMIC);
+ if (!buf) {
+ QEDF_ERR(&(fcport->qedf->dbg_ctx),
+ "Unable to alloc mp buf.\n");
+ goto free_arg;
+ }
+ hdr_len = sizeof(*fc_hdr);
+ if (hdr_len + resp_len > QEDF_PAGE_SIZE) {
+ QEDF_ERR(&(fcport->qedf->dbg_ctx), "resp_len is "
+ "beyond page size.\n");
+ goto free_buf;
+ }
+ memcpy(buf, fc_hdr, hdr_len);
+ memcpy(buf + hdr_len, resp_buf, resp_len);
+ frame_len = hdr_len + resp_len;
+
+ QEDF_INFO(&(fcport->qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Completing OX_ID 0x%x back to libfc.\n", l2_oxid);
+ qedf_process_l2_frame_compl(fcport, buf, frame_len, l2_oxid);
+
+free_buf:
+ kfree(buf);
+free_arg:
+ kfree(cb_arg);
+}
+
+int qedf_send_adisc(struct qedf_rport *fcport, struct fc_frame *fp)
+{
+ struct fc_els_adisc *adisc;
+ struct fc_frame_header *fh;
+ struct fc_lport *lport = fcport->qedf->lport;
+ struct qedf_els_cb_arg *cb_arg = NULL;
+ struct qedf_ctx *qedf;
+ uint32_t r_a_tov = lport->r_a_tov;
+ int rc;
+
+ qedf = fcport->qedf;
+ fh = fc_frame_header_get(fp);
+
+ cb_arg = kzalloc(sizeof(struct qedf_els_cb_arg), GFP_NOIO);
+ if (!cb_arg) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to allocate cb_arg for "
+ "ADISC\n");
+ rc = -ENOMEM;
+ goto adisc_err;
+ }
+ cb_arg->l2_oxid = ntohs(fh->fh_ox_id);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Sending ADISC ox_id=0x%x.\n", cb_arg->l2_oxid);
+
+ adisc = fc_frame_payload_get(fp, sizeof(*adisc));
+
+ rc = qedf_initiate_els(fcport, ELS_ADISC, adisc, sizeof(*adisc),
+ qedf_l2_els_compl, cb_arg, r_a_tov);
+
+adisc_err:
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "ADISC failed.\n");
+ kfree(cb_arg);
+ }
+ return rc;
+}
+
+static void qedf_srr_compl(struct qedf_els_cb_arg *cb_arg)
+{
+ struct qedf_ioreq *orig_io_req;
+ struct qedf_ioreq *srr_req;
+ struct qedf_mp_req *mp_req;
+ struct fc_frame_header *fc_hdr, *fh;
+ struct fc_frame *fp;
+ unsigned char *buf;
+ void *resp_buf;
+ u32 resp_len, hdr_len;
+ struct fc_lport *lport;
+ struct qedf_ctx *qedf;
+ int refcount;
+ u8 opcode;
+
+ srr_req = cb_arg->io_req;
+ qedf = srr_req->fcport->qedf;
+ lport = qedf->lport;
+
+ orig_io_req = cb_arg->aborted_io_req;
+
+ if (!orig_io_req)
+ goto out_free;
+
+ clear_bit(QEDF_CMD_SRR_SENT, &orig_io_req->flags);
+
+ if (srr_req->event != QEDF_IOREQ_EV_ELS_TMO &&
+ srr_req->event != QEDF_IOREQ_EV_ELS_ERR_DETECT)
+ cancel_delayed_work_sync(&orig_io_req->timeout_work);
+
+ refcount = atomic_read(&orig_io_req->refcount.refcount);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Entered: orig_io=%p,"
+ " orig_io_xid=0x%x, rec_xid=0x%x, refcount=%d\n",
+ orig_io_req, orig_io_req->xid, srr_req->xid, refcount);
+
+ /* If a SRR times out, simply free resources */
+ if (srr_req->event == QEDF_IOREQ_EV_ELS_TMO)
+ goto out_free;
+
+ /* Normalize response data into struct fc_frame */
+ mp_req = &(srr_req->mp_req);
+ fc_hdr = &(mp_req->resp_fc_hdr);
+ resp_len = mp_req->resp_len;
+ resp_buf = mp_req->resp_buf;
+ hdr_len = sizeof(*fc_hdr);
+
+ buf = kzalloc(QEDF_PAGE_SIZE, GFP_ATOMIC);
+ if (!buf) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "Unable to alloc mp buf.\n");
+ goto out_free;
+ }
+
+ memcpy(buf, fc_hdr, hdr_len);
+ memcpy(buf + hdr_len, resp_buf, resp_len);
+
+ fp = fc_frame_alloc(lport, resp_len);
+ if (!fp) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "fc_frame_alloc failure.\n");
+ goto out_buf;
+ }
+
+ /* Copy FC Frame header and payload into the frame */
+ fh = (struct fc_frame_header *)fc_frame_header_get(fp);
+ memcpy(fh, buf, hdr_len + resp_len);
+
+ opcode = fc_frame_payload_op(fp);
+ switch (opcode) {
+ case ELS_LS_ACC:
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "SRR success.\n");
+ break;
+ case ELS_LS_RJT:
+ QEDF_INFO(&qedf->dbg_ctx, QEDF_LOG_ELS,
+ "SRR rejected.\n");
+ qedf_initiate_abts(orig_io_req, true);
+ break;
+ }
+
+ fc_frame_free(fp);
+out_buf:
+ kfree(buf);
+out_free:
+ /* Put reference for original command since SRR completed */
+ kref_put(&orig_io_req->refcount, qedf_release_cmd);
+ kfree(cb_arg);
+}
+
+static int qedf_send_srr(struct qedf_ioreq *orig_io_req, u32 offset, u8 r_ctl)
+{
+ struct fcp_srr srr;
+ struct qedf_ctx *qedf;
+ struct qedf_rport *fcport;
+ struct fc_lport *lport;
+ struct qedf_els_cb_arg *cb_arg = NULL;
+ u32 sid, r_a_tov;
+ int rc;
+
+ if (!orig_io_req) {
+ QEDF_ERR(NULL, "orig_io_req is NULL.\n");
+ return -EINVAL;
+ }
+
+ fcport = orig_io_req->fcport;
+
+ /* Check that fcport is still offloaded */
+ if (!(test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags))) {
+ QEDF_ERR(NULL, "fcport is no longer offloaded.\n");
+ return -EINVAL;
+ }
+
+ if (!fcport->qedf) {
+ QEDF_ERR(NULL, "fcport->qedf is NULL.\n");
+ return -EINVAL;
+ }
+
+ /* Take reference until SRR command completion */
+ kref_get(&orig_io_req->refcount);
+
+ qedf = fcport->qedf;
+ lport = qedf->lport;
+ sid = fcport->sid;
+ r_a_tov = lport->r_a_tov;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Sending SRR orig_io=%p, "
+ "orig_xid=0x%x\n", orig_io_req, orig_io_req->xid);
+ memset(&srr, 0, sizeof(srr));
+
+ cb_arg = kzalloc(sizeof(struct qedf_els_cb_arg), GFP_NOIO);
+ if (!cb_arg) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to allocate cb_arg for "
+ "SRR\n");
+ rc = -ENOMEM;
+ goto srr_err;
+ }
+
+ cb_arg->aborted_io_req = orig_io_req;
+
+ srr.srr_op = ELS_SRR;
+ srr.srr_ox_id = htons(orig_io_req->xid);
+ srr.srr_rx_id = htons(orig_io_req->rx_id);
+ srr.srr_rel_off = htonl(offset);
+ srr.srr_r_ctl = r_ctl;
+
+ rc = qedf_initiate_els(fcport, ELS_SRR, &srr, sizeof(srr),
+ qedf_srr_compl, cb_arg, r_a_tov);
+
+srr_err:
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "SRR failed - release orig_io_req"
+ "=0x%x\n", orig_io_req->xid);
+ kfree(cb_arg);
+ /* If we fail to queue SRR, send ABTS to orig_io */
+ qedf_initiate_abts(orig_io_req, true);
+ kref_put(&orig_io_req->refcount, qedf_release_cmd);
+ } else
+ /* Tell other threads that SRR is in progress */
+ set_bit(QEDF_CMD_SRR_SENT, &orig_io_req->flags);
+
+ return rc;
+}
+
+static void qedf_initiate_seq_cleanup(struct qedf_ioreq *orig_io_req,
+ u32 offset, u8 r_ctl)
+{
+ struct qedf_rport *fcport;
+ unsigned long flags;
+ struct qedf_els_cb_arg *cb_arg;
+
+ fcport = orig_io_req->fcport;
+
+ QEDF_INFO(&(fcport->qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Doing sequence cleanup for xid=0x%x offset=%u.\n",
+ orig_io_req->xid, offset);
+
+ cb_arg = kzalloc(sizeof(struct qedf_els_cb_arg), GFP_NOIO);
+ if (!cb_arg) {
+ QEDF_ERR(&(fcport->qedf->dbg_ctx), "Unable to allocate cb_arg "
+ "for sequence cleanup\n");
+ return;
+ }
+
+ /* Get reference for cleanup request */
+ kref_get(&orig_io_req->refcount);
+
+ orig_io_req->cmd_type = QEDF_SEQ_CLEANUP;
+ cb_arg->offset = offset;
+ cb_arg->r_ctl = r_ctl;
+ orig_io_req->cb_arg = cb_arg;
+
+ qedf_cmd_timer_set(fcport->qedf, orig_io_req,
+ QEDF_CLEANUP_TIMEOUT * HZ);
+
+ spin_lock_irqsave(&fcport->rport_lock, flags);
+
+ qedf_add_to_sq(fcport, orig_io_req->xid, 0,
+ FCOE_TASK_TYPE_SEQUENCE_CLEANUP, offset);
+ qedf_ring_doorbell(fcport);
+
+ spin_unlock_irqrestore(&fcport->rport_lock, flags);
+}
+
+void qedf_process_seq_cleanup_compl(struct qedf_ctx *qedf,
+ struct fcoe_cqe *cqe, struct qedf_ioreq *io_req)
+{
+ int rc;
+ struct qedf_els_cb_arg *cb_arg;
+
+ cb_arg = io_req->cb_arg;
+
+ /* If we timed out just free resources */
+ if (io_req->event == QEDF_IOREQ_EV_ELS_TMO || !cqe)
+ goto free;
+
+ /* Kill the timer we put on the request */
+ cancel_delayed_work_sync(&io_req->timeout_work);
+
+ rc = qedf_send_srr(io_req, cb_arg->offset, cb_arg->r_ctl);
+ if (rc)
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to send SRR, I/O will "
+ "abort, xid=0x%x.\n", io_req->xid);
+free:
+ kfree(cb_arg);
+ kref_put(&io_req->refcount, qedf_release_cmd);
+}
+
+static bool qedf_requeue_io_req(struct qedf_ioreq *orig_io_req)
+{
+ struct qedf_rport *fcport;
+ struct qedf_ioreq *new_io_req;
+ unsigned long flags;
+ bool rc = false;
+
+ fcport = orig_io_req->fcport;
+ if (!fcport) {
+ QEDF_ERR(NULL, "fcport is NULL.\n");
+ goto out;
+ }
+
+ if (!orig_io_req->sc_cmd) {
+ QEDF_ERR(&(fcport->qedf->dbg_ctx), "sc_cmd is NULL for "
+ "xid=0x%x.\n", orig_io_req->xid);
+ goto out;
+ }
+
+ new_io_req = qedf_alloc_cmd(fcport, QEDF_SCSI_CMD);
+ if (!new_io_req) {
+ QEDF_ERR(&(fcport->qedf->dbg_ctx), "Could not allocate new "
+ "io_req.\n");
+ goto out;
+ }
+
+ new_io_req->sc_cmd = orig_io_req->sc_cmd;
+
+ /*
+ * This keeps the sc_cmd struct from being returned to the tape
+ * driver and being requeued twice. We do need to put a reference
+ * for the original I/O request since we will not do a SCSI completion
+ * for it.
+ */
+ orig_io_req->sc_cmd = NULL;
+ kref_put(&orig_io_req->refcount, qedf_release_cmd);
+
+ spin_lock_irqsave(&fcport->rport_lock, flags);
+
+ /* kref for new command released in qedf_post_io_req on error */
+ if (qedf_post_io_req(fcport, new_io_req)) {
+ QEDF_ERR(&(fcport->qedf->dbg_ctx), "Unable to post io_req\n");
+ /* Return SQE to pool */
+ atomic_inc(&fcport->free_sqes);
+ } else {
+ QEDF_INFO(&(fcport->qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Reissued SCSI command from orig_xid=0x%x on "
+ "new_xid=0x%x.\n", orig_io_req->xid, new_io_req->xid);
+ /*
+ * Abort the original I/O but do not return SCSI command as
+ * it has been reissued on another OX_ID.
+ */
+ spin_unlock_irqrestore(&fcport->rport_lock, flags);
+ qedf_initiate_abts(orig_io_req, false);
+ goto out;
+ }
+
+ spin_unlock_irqrestore(&fcport->rport_lock, flags);
+out:
+ return rc;
+}
+
+
+static void qedf_rec_compl(struct qedf_els_cb_arg *cb_arg)
+{
+ struct qedf_ioreq *orig_io_req;
+ struct qedf_ioreq *rec_req;
+ struct qedf_mp_req *mp_req;
+ struct fc_frame_header *fc_hdr, *fh;
+ struct fc_frame *fp;
+ unsigned char *buf;
+ void *resp_buf;
+ u32 resp_len, hdr_len;
+ struct fc_lport *lport;
+ struct qedf_ctx *qedf;
+ int refcount;
+ enum fc_rctl r_ctl;
+ struct fc_els_ls_rjt *rjt;
+ struct fc_els_rec_acc *acc;
+ u8 opcode;
+ u32 offset, e_stat;
+ struct scsi_cmnd *sc_cmd;
+ bool srr_needed = false;
+
+ rec_req = cb_arg->io_req;
+ qedf = rec_req->fcport->qedf;
+ lport = qedf->lport;
+
+ orig_io_req = cb_arg->aborted_io_req;
+
+ if (!orig_io_req)
+ goto out_free;
+
+ if (rec_req->event != QEDF_IOREQ_EV_ELS_TMO &&
+ rec_req->event != QEDF_IOREQ_EV_ELS_ERR_DETECT)
+ cancel_delayed_work_sync(&orig_io_req->timeout_work);
+
+ refcount = atomic_read(&orig_io_req->refcount.refcount);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Entered: orig_io=%p,"
+ " orig_io_xid=0x%x, rec_xid=0x%x, refcount=%d\n",
+ orig_io_req, orig_io_req->xid, rec_req->xid, refcount);
+
+ /* If a REC times out, free resources */
+ if (rec_req->event == QEDF_IOREQ_EV_ELS_TMO)
+ goto out_free;
+
+ /* Normalize response data into struct fc_frame */
+ mp_req = &(rec_req->mp_req);
+ fc_hdr = &(mp_req->resp_fc_hdr);
+ resp_len = mp_req->resp_len;
+ acc = resp_buf = mp_req->resp_buf;
+ hdr_len = sizeof(*fc_hdr);
+
+ buf = kzalloc(QEDF_PAGE_SIZE, GFP_ATOMIC);
+ if (!buf) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "Unable to alloc mp buf.\n");
+ goto out_free;
+ }
+
+ memcpy(buf, fc_hdr, hdr_len);
+ memcpy(buf + hdr_len, resp_buf, resp_len);
+
+ fp = fc_frame_alloc(lport, resp_len);
+ if (!fp) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "fc_frame_alloc failure.\n");
+ goto out_buf;
+ }
+
+ /* Copy FC Frame header and payload into the frame */
+ fh = (struct fc_frame_header *)fc_frame_header_get(fp);
+ memcpy(fh, buf, hdr_len + resp_len);
+
+ opcode = fc_frame_payload_op(fp);
+
+ if (opcode == ELS_LS_RJT) {
+ rjt = fc_frame_payload_get(fp, sizeof(*rjt));
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Received LS_RJT for REC: er_reason=0x%x, "
+ "er_explan=0x%x.\n", rjt->er_reason, rjt->er_explan);
+ /*
+ * The following response(s) mean that we need to reissue the
+ * request on another exchange. We need to do this without
+ * informing the upper layers lest it cause an application
+ * error.
+ */
+ if ((rjt->er_reason == ELS_RJT_LOGIC ||
+ rjt->er_reason == ELS_RJT_UNAB) &&
+ rjt->er_explan == ELS_EXPL_OXID_RXID) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Handle CMD LOST case.\n");
+ qedf_requeue_io_req(orig_io_req);
+ }
+ } else if (opcode == ELS_LS_ACC) {
+ offset = ntohl(acc->reca_fc4value);
+ e_stat = ntohl(acc->reca_e_stat);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Received LS_ACC for REC: offset=0x%x, e_stat=0x%x.\n",
+ offset, e_stat);
+ if (e_stat & ESB_ST_SEQ_INIT) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Target has the seq init\n");
+ goto out_free_frame;
+ }
+ sc_cmd = orig_io_req->sc_cmd;
+ if (!sc_cmd) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "sc_cmd is NULL for xid=0x%x.\n",
+ orig_io_req->xid);
+ goto out_free_frame;
+ }
+ /* SCSI write case */
+ if (sc_cmd->sc_data_direction == DMA_TO_DEVICE) {
+ if (offset == orig_io_req->data_xfer_len) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "WRITE - response lost.\n");
+ r_ctl = FC_RCTL_DD_CMD_STATUS;
+ srr_needed = true;
+ offset = 0;
+ } else {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "WRITE - XFER_RDY/DATA lost.\n");
+ r_ctl = FC_RCTL_DD_DATA_DESC;
+ /* Use data from warning CQE instead of REC */
+ offset = orig_io_req->tx_buf_off;
+ }
+ /* SCSI read case */
+ } else {
+ if (orig_io_req->rx_buf_off ==
+ orig_io_req->data_xfer_len) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "READ - response lost.\n");
+ srr_needed = true;
+ r_ctl = FC_RCTL_DD_CMD_STATUS;
+ offset = 0;
+ } else {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "READ - DATA lost.\n");
+ /*
+ * For read case we always set the offset to 0
+ * for sequence recovery task.
+ */
+ offset = 0;
+ r_ctl = FC_RCTL_DD_SOL_DATA;
+ }
+ }
+
+ if (srr_needed)
+ qedf_send_srr(orig_io_req, offset, r_ctl);
+ else
+ qedf_initiate_seq_cleanup(orig_io_req, offset, r_ctl);
+ }
+
+out_free_frame:
+ fc_frame_free(fp);
+out_buf:
+ kfree(buf);
+out_free:
+ /* Put reference for original command since REC completed */
+ kref_put(&orig_io_req->refcount, qedf_release_cmd);
+ kfree(cb_arg);
+}
+
+/* Assumes kref is already held by caller */
+int qedf_send_rec(struct qedf_ioreq *orig_io_req)
+{
+
+ struct fc_els_rec rec;
+ struct qedf_rport *fcport;
+ struct fc_lport *lport;
+ struct qedf_els_cb_arg *cb_arg = NULL;
+ struct qedf_ctx *qedf;
+ uint32_t sid;
+ uint32_t r_a_tov;
+ int rc;
+
+ if (!orig_io_req) {
+ QEDF_ERR(NULL, "orig_io_req is NULL.\n");
+ return -EINVAL;
+ }
+
+ fcport = orig_io_req->fcport;
+
+ /* Check that fcport is still offloaded */
+ if (!test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ QEDF_ERR(NULL, "fcport is no longer offloaded.\n");
+ return -EINVAL;
+ }
+
+ if (!fcport->qedf) {
+ QEDF_ERR(NULL, "fcport->qedf is NULL.\n");
+ return -EINVAL;
+ }
+
+ /* Take reference until REC command completion */
+ kref_get(&orig_io_req->refcount);
+
+ qedf = fcport->qedf;
+ lport = qedf->lport;
+ sid = fcport->sid;
+ r_a_tov = lport->r_a_tov;
+
+ memset(&rec, 0, sizeof(rec));
+
+ cb_arg = kzalloc(sizeof(struct qedf_els_cb_arg), GFP_NOIO);
+ if (!cb_arg) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to allocate cb_arg for "
+ "REC\n");
+ rc = -ENOMEM;
+ goto rec_err;
+ }
+
+ cb_arg->aborted_io_req = orig_io_req;
+
+ rec.rec_cmd = ELS_REC;
+ hton24(rec.rec_s_id, sid);
+ rec.rec_ox_id = htons(orig_io_req->xid);
+ rec.rec_rx_id =
+ htons(orig_io_req->task->tstorm_st_context.read_write.rx_id);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS, "Sending REC orig_io=%p, "
+ "orig_xid=0x%x rx_id=0x%x\n", orig_io_req,
+ orig_io_req->xid, rec.rec_rx_id);
+ rc = qedf_initiate_els(fcport, ELS_REC, &rec, sizeof(rec),
+ qedf_rec_compl, cb_arg, r_a_tov);
+
+rec_err:
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "REC failed - release orig_io_req"
+ "=0x%x\n", orig_io_req->xid);
+ kfree(cb_arg);
+ kref_put(&orig_io_req->refcount, qedf_release_cmd);
+ }
+ return rc;
+}
diff --git a/drivers/scsi/qedf/qedf_fip.c b/drivers/scsi/qedf/qedf_fip.c
new file mode 100644
index 0000000..4f185c6
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_fip.c
@@ -0,0 +1,267 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#include <linux/if_ether.h>
+#include <linux/if_vlan.h>
+#include "qedf.h"
+
+extern const struct qed_fcoe_ops *qed_ops;
+/*
+ * FIP VLAN functions that will eventually move to libfcoe.
+ */
+
+void qedf_fcoe_send_vlan_req(struct qedf_ctx *qedf)
+{
+ struct sk_buff *skb;
+ char *eth_fr;
+ int fr_len;
+ struct fip_vlan *vlan;
+#define MY_FIP_ALL_FCF_MACS ((__u8[6]) { 1, 0x10, 0x18, 1, 0, 2 })
+ static u8 my_fcoe_all_fcfs[ETH_ALEN] = MY_FIP_ALL_FCF_MACS;
+
+ skb = dev_alloc_skb(sizeof(struct fip_vlan));
+ if (!skb)
+ return;
+
+ fr_len = sizeof(*vlan);
+ eth_fr = (char *)skb->data;
+ vlan = (struct fip_vlan *)eth_fr;
+
+ memset(vlan, 0, sizeof(*vlan));
+ ether_addr_copy(vlan->eth.h_source, qedf->mac);
+ ether_addr_copy(vlan->eth.h_dest, my_fcoe_all_fcfs);
+ vlan->eth.h_proto = htons(ETH_P_FIP);
+
+ vlan->fip.fip_ver = FIP_VER_ENCAPS(FIP_VER);
+ vlan->fip.fip_op = htons(FIP_OP_VLAN);
+ vlan->fip.fip_subcode = FIP_SC_VL_REQ;
+ vlan->fip.fip_dl_len = htons(sizeof(vlan->desc) / FIP_BPW);
+
+ vlan->desc.mac.fd_desc.fip_dtype = FIP_DT_MAC;
+ vlan->desc.mac.fd_desc.fip_dlen = sizeof(vlan->desc.mac) / FIP_BPW;
+ ether_addr_copy(vlan->desc.mac.fd_mac, qedf->mac);
+
+ vlan->desc.wwnn.fd_desc.fip_dtype = FIP_DT_NAME;
+ vlan->desc.wwnn.fd_desc.fip_dlen = sizeof(vlan->desc.wwnn) / FIP_BPW;
+ put_unaligned_be64(qedf->lport->wwnn, &vlan->desc.wwnn.fd_wwn);
+
+ skb_put(skb, sizeof(*vlan));
+ skb->protocol = htons(ETH_P_FIP);
+ skb_reset_mac_header(skb);
+ skb_reset_network_header(skb);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Sending FIP VLAN "
+ "request.");
+
+ if (atomic_read(&qedf->link_state) != QEDF_LINK_UP) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Cannot send vlan request "
+ "because link is not up.\n");
+
+ kfree_skb(skb);
+ return;
+ }
+ qed_ops->ll2->start_xmit(qedf->cdev, skb);
+}
+
+static void qedf_fcoe_process_vlan_resp(struct qedf_ctx *qedf,
+ struct sk_buff *skb)
+{
+ struct fip_header *fiph;
+ struct fip_desc *desc;
+ u16 vid = 0;
+ ssize_t rlen;
+ size_t dlen;
+
+ fiph = (struct fip_header *)(((void *)skb->data) + 2 * ETH_ALEN + 2);
+
+ rlen = ntohs(fiph->fip_dl_len) * 4;
+ desc = (struct fip_desc *)(fiph + 1);
+ while (rlen > 0) {
+ dlen = desc->fip_dlen * FIP_BPW;
+ switch (desc->fip_dtype) {
+ case FIP_DT_VLAN:
+ vid = ntohs(((struct fip_vlan_desc *)desc)->fd_vlan);
+ break;
+ }
+ desc = (struct fip_desc *)((char *)desc + dlen);
+ rlen -= dlen;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "VLAN response, "
+ "vid=0x%x.\n", vid);
+
+ if (vid > 0 && qedf->vlan_id != vid) {
+ qedf_set_vlan_id(qedf, vid);
+
+ /* Inform waiter that it's ok to call fcoe_ctlr_link up() */
+ complete(&qedf->fipvlan_compl);
+ }
+}
+
+void qedf_fip_send(struct fcoe_ctlr *fip, struct sk_buff *skb)
+{
+ struct qedf_ctx *qedf = container_of(fip, struct qedf_ctx, ctlr);
+ struct ethhdr *eth_hdr;
+ struct vlan_ethhdr *vlan_hdr;
+ struct fip_header *fiph;
+ u16 op, vlan_tci = 0;
+ u8 sub;
+
+ if (!test_bit(QEDF_LL2_STARTED, &qedf->flags)) {
+ QEDF_WARN(&(qedf->dbg_ctx), "LL2 not started\n");
+ kfree_skb(skb);
+ return;
+ }
+
+ fiph = (struct fip_header *) ((void *)skb->data + 2 * ETH_ALEN + 2);
+ eth_hdr = (struct ethhdr *)skb_mac_header(skb);
+ op = ntohs(fiph->fip_op);
+ sub = fiph->fip_subcode;
+
+ if (!qedf->vlan_hw_insert) {
+ vlan_hdr = (struct vlan_ethhdr *)skb_push(skb, sizeof(*vlan_hdr)
+ - sizeof(*eth_hdr));
+ memcpy(vlan_hdr, eth_hdr, 2 * ETH_ALEN);
+ vlan_hdr->h_vlan_proto = htons(ETH_P_8021Q);
+ vlan_hdr->h_vlan_encapsulated_proto = eth_hdr->h_proto;
+ vlan_hdr->h_vlan_TCI = vlan_tci = htons(qedf->vlan_id);
+ }
+
+ /* Update eth_hdr since we added a VLAN tag */
+ eth_hdr = (struct ethhdr *)skb_mac_header(skb);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2, "FIP frame send: "
+ "dest=%pM op=%x sub=%x vlan=%04x.", eth_hdr->h_dest, op, sub,
+ ntohs(vlan_tci));
+ if (qedf_dump_frames)
+ print_hex_dump(KERN_WARNING, "fip ", DUMP_PREFIX_OFFSET, 16, 1,
+ skb->data, skb->len, false);
+
+ qed_ops->ll2->start_xmit(qedf->cdev, skb);
+}
+
+/* Process incoming FIP frames. */
+void qedf_fip_recv(struct qedf_ctx *qedf, struct sk_buff *skb)
+{
+ struct ethhdr *eth_hdr;
+ struct fip_header *fiph;
+ struct fip_desc *desc;
+ struct fip_mac_desc *mp;
+ struct fip_wwn_desc *wp;
+ struct fip_vn_desc *vp;
+ size_t rlen, dlen;
+ uint32_t cvl_port_id;
+ __u8 cvl_mac[ETH_ALEN];
+ u16 op;
+ u8 sub;
+
+ eth_hdr = (struct ethhdr *)skb_mac_header(skb);
+ fiph = (struct fip_header *) ((void *)skb->data + 2 * ETH_ALEN + 2);
+ op = ntohs(fiph->fip_op);
+ sub = fiph->fip_subcode;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2, "FIP frame received: "
+ "skb=%p fiph=%p source=%pM op=%x sub=%x", skb, fiph,
+ eth_hdr->h_source, op, sub);
+ if (qedf_dump_frames)
+ print_hex_dump(KERN_WARNING, "fip ", DUMP_PREFIX_OFFSET, 16, 1,
+ skb->data, skb->len, false);
+
+ /* Handle FIP VLAN resp in the driver */
+ if (op == FIP_OP_VLAN && sub == FIP_SC_VL_NOTE) {
+ qedf_fcoe_process_vlan_resp(qedf, skb);
+ qedf->vlan_hw_insert = 0;
+ } else if (op == FIP_OP_CTRL && sub == FIP_SC_CLR_VLINK) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Clear virtual "
+ "link received.\n");
+
+ /* Check that an FCF has been selected by fcoe */
+ if (qedf->ctlr.sel_fcf == NULL) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Dropping CVL since FCF has not been selected "
+ "yet.");
+ return;
+ }
+
+ cvl_port_id = 0;
+ memset(cvl_mac, 0, ETH_ALEN);
+ /*
+ * We need to loop through the CVL descriptors to determine
+ * if we want to reset the fcoe link
+ */
+ rlen = ntohs(fiph->fip_dl_len) * FIP_BPW;
+ desc = (struct fip_desc *)(fiph + 1);
+ while (rlen >= sizeof(*desc)) {
+ dlen = desc->fip_dlen * FIP_BPW;
+ switch (desc->fip_dtype) {
+ case FIP_DT_MAC:
+ mp = (struct fip_mac_desc *)desc;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2,
+ "fd_mac=%pM.\n", __func__, mp->fd_mac);
+ ether_addr_copy(cvl_mac, mp->fd_mac);
+ break;
+ case FIP_DT_NAME:
+ wp = (struct fip_wwn_desc *)desc;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2,
+ "fc_wwpn=%016llx.\n",
+ get_unaligned_be64(&wp->fd_wwn));
+ break;
+ case FIP_DT_VN_ID:
+ vp = (struct fip_vn_desc *)desc;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2,
+ "fd_fc_id=%x.\n", ntoh24(vp->fd_fc_id));
+ cvl_port_id = ntoh24(vp->fd_fc_id);
+ break;
+ default:
+ /* Ignore anything else */
+ break;
+ }
+ desc = (struct fip_desc *)((char *)desc + dlen);
+ rlen -= dlen;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2,
+ "cvl_port_id=%06x cvl_mac=%pM.\n", cvl_port_id,
+ cvl_mac);
+ if (cvl_port_id == qedf->lport->port_id &&
+ ether_addr_equal(cvl_mac,
+ qedf->ctlr.sel_fcf->fcf_mac)) {
+ fcoe_ctlr_link_down(&qedf->ctlr);
+ qedf_wait_for_upload(qedf);
+ fcoe_ctlr_link_up(&qedf->ctlr);
+ }
+ } else {
+ /* Everything else is handled by libfcoe */
+ __skb_pull(skb, ETH_HLEN);
+ fcoe_ctlr_recv(&qedf->ctlr, skb);
+ }
+}
+
+void qedf_update_src_mac(struct fc_lport *lport, u8 *addr)
+{
+ struct qedf_ctx *qedf = lport_priv(lport);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Setting data_src_addr=%pM.\n", addr);
+ ether_addr_copy(qedf->data_src_addr, addr);
+}
+
+u8 *qedf_get_src_mac(struct fc_lport *lport)
+{
+ u8 mac[ETH_ALEN];
+ u8 port_id[3];
+ struct qedf_ctx *qedf = lport_priv(lport);
+
+ /* We need to use the lport port_id to create the data_src_addr */
+ if (is_zero_ether_addr(qedf->data_src_addr)) {
+ hton24(port_id, lport->port_id);
+ fc_fcoe_set_mac(mac, port_id);
+ qedf->ctlr.update_mac(lport, mac);
+ }
+ return qedf->data_src_addr;
+}
diff --git a/drivers/scsi/qedf/qedf_hsi.h b/drivers/scsi/qedf/qedf_hsi.h
new file mode 100644
index 0000000..953aa5e
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_hsi.h
@@ -0,0 +1,427 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#ifndef __QEDF_HSI__
+#define __QEDF_HSI__
+/*
+ * Add include to common target
+ */
+#include <linux/qed/common_hsi.h>
+
+/*
+ * Add include to common storage target
+ */
+#include <linux/qed/storage_common.h>
+
+/*
+ * Add include to common fcoe target for both eCore and protocol driver
+ */
+#include <linux/qed/fcoe_common.h>
+
+
+/*
+ * FCoE CQ element ABTS information
+ */
+struct fcoe_abts_info {
+ u8 r_ctl /* R_CTL in the ABTS response frame */;
+ u8 reserved0;
+ __le16 rx_id;
+ __le32 reserved2[2];
+ __le32 fc_payload[3] /* ABTS FC payload response frame */;
+};
+
+
+/*
+ * FCoE class type
+ */
+enum fcoe_class_type {
+ FCOE_TASK_CLASS_TYPE_3,
+ FCOE_TASK_CLASS_TYPE_2,
+ MAX_FCOE_CLASS_TYPE
+};
+
+
+/*
+ * FCoE CMDQ element control information
+ */
+struct fcoe_cmdqe_control {
+ __le16 conn_id;
+ u8 num_additional_cmdqes;
+ u8 cmdType;
+ /* true for ABTS request cmdqe. used in Target mode */
+#define FCOE_CMDQE_CONTROL_ABTSREQCMD_MASK 0x1
+#define FCOE_CMDQE_CONTROL_ABTSREQCMD_SHIFT 0
+#define FCOE_CMDQE_CONTROL_RESERVED1_MASK 0x7F
+#define FCOE_CMDQE_CONTROL_RESERVED1_SHIFT 1
+ u8 reserved2[4];
+};
+
+/*
+ * FCoE control + payload CMDQ element
+ */
+struct fcoe_cmdqe {
+ struct fcoe_cmdqe_control hdr;
+ u8 fc_header[24];
+ __le32 fcp_cmd_payload[8];
+};
+
+
+
+/*
+ * FCP RSP flags
+ */
+struct fcoe_fcp_rsp_flags {
+ u8 flags;
+#define FCOE_FCP_RSP_FLAGS_FCP_RSP_LEN_VALID_MASK 0x1
+#define FCOE_FCP_RSP_FLAGS_FCP_RSP_LEN_VALID_SHIFT 0
+#define FCOE_FCP_RSP_FLAGS_FCP_SNS_LEN_VALID_MASK 0x1
+#define FCOE_FCP_RSP_FLAGS_FCP_SNS_LEN_VALID_SHIFT 1
+#define FCOE_FCP_RSP_FLAGS_FCP_RESID_OVER_MASK 0x1
+#define FCOE_FCP_RSP_FLAGS_FCP_RESID_OVER_SHIFT 2
+#define FCOE_FCP_RSP_FLAGS_FCP_RESID_UNDER_MASK 0x1
+#define FCOE_FCP_RSP_FLAGS_FCP_RESID_UNDER_SHIFT 3
+#define FCOE_FCP_RSP_FLAGS_FCP_CONF_REQ_MASK 0x1
+#define FCOE_FCP_RSP_FLAGS_FCP_CONF_REQ_SHIFT 4
+#define FCOE_FCP_RSP_FLAGS_FCP_BIDI_FLAGS_MASK 0x7
+#define FCOE_FCP_RSP_FLAGS_FCP_BIDI_FLAGS_SHIFT 5
+};
+
+/*
+ * FCoE CQ element response information
+ */
+struct fcoe_cqe_rsp_info {
+ struct fcoe_fcp_rsp_flags rsp_flags;
+ u8 scsi_status_code;
+ __le16 retry_delay_timer;
+ __le32 fcp_resid;
+ __le32 fcp_sns_len;
+ __le32 fcp_rsp_len;
+ __le16 rx_id;
+ u8 fw_error_flags;
+#define FCOE_CQE_RSP_INFO_FW_UNDERRUN_MASK 0x1 /* FW detected underrun */
+#define FCOE_CQE_RSP_INFO_FW_UNDERRUN_SHIFT 0
+#define FCOE_CQE_RSP_INFO_RESREVED_MASK 0x7F
+#define FCOE_CQE_RSP_INFO_RESREVED_SHIFT 1
+ u8 reserved;
+ __le32 fw_residual /* Residual bytes calculated by FW */;
+};
+
+/*
+ * FCoE CQ element Target completion information
+ */
+struct fcoe_cqe_target_info {
+ __le16 rx_id;
+ __le16 reserved0;
+ __le32 reserved1[5];
+};
+
+/*
+ * FCoE error/warning reporting entry
+ */
+struct fcoe_err_report_entry {
+ __le32 err_warn_bitmap_lo /* Error bitmap lower 32 bits */;
+ __le32 err_warn_bitmap_hi /* Error bitmap higher 32 bits */;
+ /* Buffer offset the beginning of the Sequence last transmitted */
+ __le32 tx_buf_off;
+ /* Buffer offset from the beginning of the Sequence last received */
+ __le32 rx_buf_off;
+ __le16 rx_id /* RX_ID of the associated task */;
+ __le16 reserved1;
+ __le32 reserved2;
+};
+
+/*
+ * FCoE CQ element middle path information
+ */
+struct fcoe_cqe_midpath_info {
+ __le32 data_placement_size;
+ __le16 rx_id;
+ __le16 reserved0;
+ __le32 reserved1[4];
+};
+
+/*
+ * FCoE CQ element unsolicited information
+ */
+struct fcoe_unsolic_info {
+ /* BD information: Physical address and opaque data */
+ struct scsi_bd bd_info;
+ __le16 conn_id /* Connection ID the frame is associated to */;
+ __le16 pkt_len /* Packet length */;
+ u8 reserved1[4];
+};
+
+/*
+ * FCoE warning reporting entry
+ */
+struct fcoe_warning_report_entry {
+ /* BD information: Physical address and opaque data */
+ struct scsi_bd bd_info;
+ /* Buffer offset the beginning of the Sequence last transmitted */
+ __le32 buf_off;
+ __le16 rx_id /* RX_ID of the associated task */;
+ __le16 reserved1;
+};
+
+/*
+ * FCoE CQ element information
+ */
+union fcoe_cqe_info {
+ struct fcoe_cqe_rsp_info rsp_info /* Response completion information */;
+ /* Target completion information */
+ struct fcoe_cqe_target_info target_info;
+ /* Error completion information */
+ struct fcoe_err_report_entry err_info;
+ struct fcoe_abts_info abts_info /* ABTS completion information */;
+ /* Middle path completion information */
+ struct fcoe_cqe_midpath_info midpath_info;
+ /* Unsolicited packet completion information */
+ struct fcoe_unsolic_info unsolic_info;
+ /* Warning completion information (Rec Tov expiration) */
+ struct fcoe_warning_report_entry warn_info;
+};
+
+/*
+ * FCoE CQ element
+ */
+struct fcoe_cqe {
+ __le32 cqe_data;
+ /* The task identifier (OX_ID) to be completed */
+#define FCOE_CQE_TASK_ID_MASK 0xFFFF
+#define FCOE_CQE_TASK_ID_SHIFT 0
+ /*
+ * The CQE type: 0x0 Indicating on a pending work request completion.
+ * 0x1 - Indicating on an unsolicited event notification. use enum
+ * fcoe_cqe_type (use enum fcoe_cqe_type)
+ */
+#define FCOE_CQE_CQE_TYPE_MASK 0xF
+#define FCOE_CQE_CQE_TYPE_SHIFT 16
+#define FCOE_CQE_RESERVED0_MASK 0xFFF
+#define FCOE_CQE_RESERVED0_SHIFT 20
+ __le16 reserved1;
+ __le16 fw_cq_prod;
+ union fcoe_cqe_info cqe_info;
+};
+
+
+
+
+
+
+/*
+ * FCoE CQE type
+ */
+enum fcoe_cqe_type {
+ /* solicited response on a R/W or middle-path SQE */
+ FCOE_GOOD_COMPLETION_CQE_TYPE,
+ FCOE_UNSOLIC_CQE_TYPE /* unsolicited packet, RQ consumed */,
+ FCOE_ERROR_DETECTION_CQE_TYPE /* timer expiration, validation error */,
+ FCOE_WARNING_CQE_TYPE /* rec_tov or rr_tov timer expiration */,
+ FCOE_EXCH_CLEANUP_CQE_TYPE /* task cleanup completed */,
+ FCOE_ABTS_CQE_TYPE /* ABTS received and task cleaned */,
+ FCOE_DUMMY_CQE_TYPE /* just increment SQ CONS */,
+ /* Task was completed wight after sending a pkt to the target */
+ FCOE_LOCAL_COMP_CQE_TYPE,
+ MAX_FCOE_CQE_TYPE
+};
+
+
+/*
+ * FCoE device type
+ */
+enum fcoe_device_type {
+ FCOE_TASK_DEV_TYPE_DISK,
+ FCOE_TASK_DEV_TYPE_TAPE,
+ MAX_FCOE_DEVICE_TYPE
+};
+
+
+
+
+/*
+ * FCoE fast path error codes
+ */
+enum fcoe_fp_error_warning_code {
+ FCOE_ERROR_CODE_XFER_OOO_RO /* XFER error codes */,
+ FCOE_ERROR_CODE_XFER_RO_NOT_ALIGNED,
+ FCOE_ERROR_CODE_XFER_NULL_BURST_LEN,
+ FCOE_ERROR_CODE_XFER_RO_GREATER_THAN_DATA2TRNS,
+ FCOE_ERROR_CODE_XFER_INVALID_PAYLOAD_SIZE,
+ FCOE_ERROR_CODE_XFER_TASK_TYPE_NOT_WRITE,
+ FCOE_ERROR_CODE_XFER_PEND_XFER_SET,
+ FCOE_ERROR_CODE_XFER_OPENED_SEQ,
+ FCOE_ERROR_CODE_XFER_FCTL,
+ FCOE_ERROR_CODE_FCP_RSP_BIDI_FLAGS_SET /* FCP RSP error codes */,
+ FCOE_ERROR_CODE_FCP_RSP_INVALID_LENGTH_FIELD,
+ FCOE_ERROR_CODE_FCP_RSP_INVALID_SNS_FIELD,
+ FCOE_ERROR_CODE_FCP_RSP_INVALID_PAYLOAD_SIZE,
+ FCOE_ERROR_CODE_FCP_RSP_PEND_XFER_SET,
+ FCOE_ERROR_CODE_FCP_RSP_OPENED_SEQ,
+ FCOE_ERROR_CODE_FCP_RSP_FCTL,
+ FCOE_ERROR_CODE_FCP_RSP_LAST_SEQ_RESET,
+ FCOE_ERROR_CODE_FCP_RSP_CONF_REQ_NOT_SUPPORTED_YET,
+ FCOE_ERROR_CODE_DATA_OOO_RO /* FCP DATA error codes */,
+ FCOE_ERROR_CODE_DATA_EXCEEDS_DEFINED_MAX_FRAME_SIZE,
+ FCOE_ERROR_CODE_DATA_EXCEEDS_DATA2TRNS,
+ FCOE_ERROR_CODE_DATA_SOFI3_SEQ_ACTIVE_SET,
+ FCOE_ERROR_CODE_DATA_SOFN_SEQ_ACTIVE_RESET,
+ FCOE_ERROR_CODE_DATA_EOFN_END_SEQ_SET,
+ FCOE_ERROR_CODE_DATA_EOFT_END_SEQ_RESET,
+ FCOE_ERROR_CODE_DATA_TASK_TYPE_NOT_READ,
+ FCOE_ERROR_CODE_DATA_FCTL_INITIATIR,
+ FCOE_ERROR_CODE_MIDPATH_INVALID_TYPE /* Middle path error codes */,
+ FCOE_ERROR_CODE_MIDPATH_SOFI3_SEQ_ACTIVE_SET,
+ FCOE_ERROR_CODE_MIDPATH_SOFN_SEQ_ACTIVE_RESET,
+ FCOE_ERROR_CODE_MIDPATH_EOFN_END_SEQ_SET,
+ FCOE_ERROR_CODE_MIDPATH_EOFT_END_SEQ_RESET,
+ FCOE_ERROR_CODE_MIDPATH_REPLY_FCTL,
+ FCOE_ERROR_CODE_MIDPATH_INVALID_REPLY,
+ FCOE_ERROR_CODE_MIDPATH_ELS_REPLY_RCTL,
+ FCOE_ERROR_CODE_COMMON_MIDDLE_FRAME_WITH_PAD /* Common error codes */,
+ FCOE_ERROR_CODE_COMMON_SEQ_INIT_IN_TCE,
+ FCOE_ERROR_CODE_COMMON_FC_HDR_RX_ID_MISMATCH,
+ FCOE_ERROR_CODE_COMMON_INCORRECT_SEQ_CNT,
+ FCOE_ERROR_CODE_COMMON_DATA_FC_HDR_FCP_TYPE_MISMATCH,
+ FCOE_ERROR_CODE_COMMON_DATA_NO_MORE_SGES,
+ FCOE_ERROR_CODE_COMMON_OPTIONAL_FC_HDR,
+ FCOE_ERROR_CODE_COMMON_READ_TCE_OX_ID_TOO_BIG,
+ FCOE_ERROR_CODE_COMMON_DATA_WAS_NOT_TRANSMITTED,
+ FCOE_ERROR_CODE_COMMON_TASK_DDF_RCTL_INFO_FIELD,
+ FCOE_ERROR_CODE_COMMON_TASK_INVALID_RCTL,
+ FCOE_ERROR_CODE_COMMON_TASK_RCTL_GENERAL_MISMATCH,
+ FCOE_ERROR_CODE_E_D_TOV_TIMER_EXPIRATION /* Timer error codes */,
+ FCOE_WARNING_CODE_REC_TOV_TIMER_EXPIRATION /* Timer error codes */,
+ FCOE_ERROR_CODE_RR_TOV_TIMER_EXPIRATION /* Timer error codes */,
+ /* ABTSrsp pckt arrived unexpected */
+ FCOE_ERROR_CODE_ABTS_REPLY_UNEXPECTED,
+ FCOE_ERROR_CODE_TARGET_MODE_FCP_RSP,
+ FCOE_ERROR_CODE_TARGET_MODE_FCP_XFER,
+ FCOE_ERROR_CODE_TARGET_MODE_DATA_TASK_TYPE_NOT_WRITE,
+ FCOE_ERROR_CODE_DATA_FCTL_TARGET,
+ FCOE_ERROR_CODE_TARGET_DATA_SIZE_NO_MATCH_XFER,
+ FCOE_ERROR_CODE_TARGET_DIF_CRC_CHECKSUM_ERROR,
+ FCOE_ERROR_CODE_TARGET_DIF_REF_TAG_ERROR,
+ FCOE_ERROR_CODE_TARGET_DIF_APP_TAG_ERROR,
+ MAX_FCOE_FP_ERROR_WARNING_CODE
+};
+
+
+/*
+ * FCoE RESPQ element
+ */
+struct fcoe_respqe {
+ __le16 ox_id /* OX_ID that is located in the FCP_RSP FC header */;
+ __le16 rx_id /* RX_ID that is located in the FCP_RSP FC header */;
+ __le32 additional_info;
+/* PARAM that is located in the FCP_RSP FC header */
+#define FCOE_RESPQE_PARAM_MASK 0xFFFFFF
+#define FCOE_RESPQE_PARAM_SHIFT 0
+/* Indication whther its Target-auto-rsp mode or not */
+#define FCOE_RESPQE_TARGET_AUTO_RSP_MASK 0xFF
+#define FCOE_RESPQE_TARGET_AUTO_RSP_SHIFT 24
+};
+
+
+/*
+ * FCoE slow path error codes
+ */
+enum fcoe_sp_error_code {
+ /* Error codes for Error Reporting in slow path flows */
+ FCOE_ERROR_CODE_SLOW_PATH_TOO_MANY_FUNCS,
+ FCOE_ERROR_SLOW_PATH_CODE_NO_LICENSE,
+ MAX_FCOE_SP_ERROR_CODE
+};
+
+
+/*
+ * FCoE SQE request type
+ */
+enum fcoe_sqe_request_type {
+ SEND_FCOE_CMD,
+ SEND_FCOE_MIDPATH,
+ SEND_FCOE_ABTS_REQUEST,
+ FCOE_EXCHANGE_CLEANUP,
+ FCOE_SEQUENCE_RECOVERY,
+ SEND_FCOE_XFER_RDY,
+ SEND_FCOE_RSP,
+ SEND_FCOE_RSP_WITH_SENSE_DATA,
+ SEND_FCOE_TARGET_DATA,
+ SEND_FCOE_INITIATOR_DATA,
+ /*
+ * Xfer Continuation (==1) ready to be sent. Previous XFERs data
+ * received successfully.
+ */
+ SEND_FCOE_XFER_CONTINUATION_RDY,
+ SEND_FCOE_TARGET_ABTS_RSP,
+ MAX_FCOE_SQE_REQUEST_TYPE
+};
+
+
+/*
+ * FCoE task TX state
+ */
+enum fcoe_task_tx_state {
+ /* Initiate state after driver has initialized the task */
+ FCOE_TASK_TX_STATE_NORMAL,
+ /* Updated by TX path after complete transmitting unsolicited packet */
+ FCOE_TASK_TX_STATE_UNSOLICITED_COMPLETED,
+ /*
+ * Updated by TX path after start processing the task requesting the
+ * cleanup/abort operation
+ */
+ FCOE_TASK_TX_STATE_CLEAN_REQ,
+ FCOE_TASK_TX_STATE_ABTS /* Updated by TX path during abort procedure */,
+ /* Updated by TX path during exchange cleanup procedure */
+ FCOE_TASK_TX_STATE_EXCLEANUP,
+ /*
+ * Updated by TX path during exchange cleanup continuation task
+ * procedure
+ */
+ FCOE_TASK_TX_STATE_EXCLEANUP_TARGET_WRITE_CONT,
+ /* Updated by TX path during exchange cleanup first xfer procedure */
+ FCOE_TASK_TX_STATE_EXCLEANUP_TARGET_WRITE,
+ /* Updated by TX path during exchange cleanup read task in Target */
+ FCOE_TASK_TX_STATE_EXCLEANUP_TARGET_READ_OR_RSP,
+ /* Updated by TX path during target exchange cleanup procedure */
+ FCOE_TASK_TX_STATE_EXCLEANUP_TARGET_WRITE_LAST_CYCLE,
+ /* Updated by TX path during sequence recovery procedure */
+ FCOE_TASK_TX_STATE_SEQRECOVERY,
+ MAX_FCOE_TASK_TX_STATE
+};
+
+
+/*
+ * FCoE task type
+ */
+enum fcoe_task_type {
+ FCOE_TASK_TYPE_WRITE_INITIATOR,
+ FCOE_TASK_TYPE_READ_INITIATOR,
+ FCOE_TASK_TYPE_MIDPATH,
+ FCOE_TASK_TYPE_UNSOLICITED,
+ FCOE_TASK_TYPE_ABTS,
+ FCOE_TASK_TYPE_EXCHANGE_CLEANUP,
+ FCOE_TASK_TYPE_SEQUENCE_CLEANUP,
+ FCOE_TASK_TYPE_WRITE_TARGET,
+ FCOE_TASK_TYPE_READ_TARGET,
+ FCOE_TASK_TYPE_RSP,
+ FCOE_TASK_TYPE_RSP_SENSE_DATA,
+ FCOE_TASK_TYPE_ABTS_TARGET,
+ FCOE_TASK_TYPE_ENUM_SIZE,
+ MAX_FCOE_TASK_TYPE
+};
+
+struct scsi_glbl_queue_entry {
+ /* Start physical address for the RQ (receive queue) PBL. */
+ struct regpair rq_pbl_addr;
+ /* Start physical address for the CQ (completion queue) PBL. */
+ struct regpair cq_pbl_addr;
+ /* Start physical address for the CMDQ (command queue) PBL. */
+ struct regpair cmdq_pbl_addr;
+};
+
+#endif /* __QEDF_HSI__ */
diff --git a/drivers/scsi/qedf/qedf_io.c b/drivers/scsi/qedf/qedf_io.c
new file mode 100644
index 0000000..4b971bb
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_io.c
@@ -0,0 +1,2301 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#include <linux/spinlock.h>
+#include <linux/vmalloc.h>
+#include "qedf.h"
+#include <scsi/scsi_tcq.h>
+
+DECLARE_PER_CPU(struct qedf_percpu_iothread_s, qedf_percpu_iothreads);
+
+void qedf_cmd_timer_set(struct qedf_ctx *qedf, struct qedf_ioreq *io_req,
+ unsigned int timer_msec)
+{
+ queue_delayed_work(qedf->timer_work_queue, &io_req->timeout_work,
+ msecs_to_jiffies(timer_msec));
+}
+
+static void qedf_cmd_timeout(struct work_struct *work)
+{
+
+ struct qedf_ioreq *io_req =
+ container_of(work, struct qedf_ioreq, timeout_work.work);
+ struct qedf_ctx *qedf = io_req->fcport->qedf;
+ struct qedf_rport *fcport = io_req->fcport;
+ u8 op = 0;
+
+ switch (io_req->cmd_type) {
+ case QEDF_ABTS:
+ QEDF_ERR((&qedf->dbg_ctx), "ABTS timeout, xid=0x%x.\n",
+ io_req->xid);
+ /* Cleanup timed out ABTS */
+ qedf_initiate_cleanup(io_req, true);
+ complete(&io_req->abts_done);
+
+ /*
+ * Need to call kref_put for reference taken when initiate_abts
+ * was called since abts_compl won't be called now that we've
+ * cleaned up the task.
+ */
+ kref_put(&io_req->refcount, qedf_release_cmd);
+
+ /*
+ * Now that the original I/O and the ABTS are complete see
+ * if we need to reconnect to the target.
+ */
+ qedf_restart_rport(fcport);
+ break;
+ case QEDF_ELS:
+ kref_get(&io_req->refcount);
+ /*
+ * Don't attempt to clean an ELS timeout as any subseqeunt
+ * ABTS or cleanup requests just hang. For now just free
+ * the resources of the original I/O and the RRQ
+ */
+ QEDF_ERR(&(qedf->dbg_ctx), "ELS timeout, xid=0x%x.\n",
+ io_req->xid);
+ io_req->event = QEDF_IOREQ_EV_ELS_TMO;
+ /* Call callback function to complete command */
+ if (io_req->cb_func && io_req->cb_arg) {
+ op = io_req->cb_arg->op;
+ io_req->cb_func(io_req->cb_arg);
+ io_req->cb_arg = NULL;
+ }
+ qedf_initiate_cleanup(io_req, true);
+ kref_put(&io_req->refcount, qedf_release_cmd);
+ break;
+ case QEDF_SEQ_CLEANUP:
+ QEDF_ERR(&(qedf->dbg_ctx), "Sequence cleanup timeout, "
+ "xid=0x%x.\n", io_req->xid);
+ qedf_initiate_cleanup(io_req, true);
+ io_req->event = QEDF_IOREQ_EV_ELS_TMO;
+ qedf_process_seq_cleanup_compl(qedf, NULL, io_req);
+ break;
+ default:
+ break;
+ }
+}
+
+void qedf_cmd_mgr_free(struct qedf_cmd_mgr *cmgr)
+{
+ struct io_bdt *bdt_info;
+ struct qedf_ctx *qedf = cmgr->qedf;
+ size_t bd_tbl_sz;
+ u16 min_xid = QEDF_MIN_XID;
+ u16 max_xid = (FCOE_PARAMS_NUM_TASKS - 1);
+ int num_ios;
+ int i;
+ struct qedf_ioreq *io_req;
+
+ num_ios = max_xid - min_xid + 1;
+
+ /* Free fcoe_bdt_ctx structures */
+ if (!cmgr->io_bdt_pool)
+ goto free_cmd_pool;
+
+ bd_tbl_sz = QEDF_MAX_BDS_PER_CMD * sizeof(struct fcoe_sge);
+ for (i = 0; i < num_ios; i++) {
+ bdt_info = cmgr->io_bdt_pool[i];
+ if (bdt_info->bd_tbl) {
+ dma_free_coherent(&qedf->pdev->dev, bd_tbl_sz,
+ bdt_info->bd_tbl, bdt_info->bd_tbl_dma);
+ bdt_info->bd_tbl = NULL;
+ }
+ }
+
+ /* Destroy io_bdt pool */
+ for (i = 0; i < num_ios; i++) {
+ kfree(cmgr->io_bdt_pool[i]);
+ cmgr->io_bdt_pool[i] = NULL;
+ }
+
+ kfree(cmgr->io_bdt_pool);
+ cmgr->io_bdt_pool = NULL;
+
+free_cmd_pool:
+
+ for (i = 0; i < num_ios; i++) {
+ io_req = &cmgr->cmds[i];
+ /* Make sure we free per command sense buffer */
+ if (io_req->sense_buffer)
+ dma_free_coherent(&qedf->pdev->dev,
+ QEDF_SCSI_SENSE_BUFFERSIZE, io_req->sense_buffer,
+ io_req->sense_buffer_dma);
+ cancel_delayed_work_sync(&io_req->rrq_work);
+ }
+
+ /* Free command manager itself */
+ vfree(cmgr);
+}
+
+static void qedf_handle_rrq(struct work_struct *work)
+{
+ struct qedf_ioreq *io_req =
+ container_of(work, struct qedf_ioreq, rrq_work.work);
+
+ qedf_send_rrq(io_req);
+
+}
+
+struct qedf_cmd_mgr *qedf_cmd_mgr_alloc(struct qedf_ctx *qedf)
+{
+ struct qedf_cmd_mgr *cmgr;
+ struct io_bdt *bdt_info;
+ struct qedf_ioreq *io_req;
+ u16 xid;
+ int i;
+ int num_ios;
+ u16 min_xid = QEDF_MIN_XID;
+ u16 max_xid = (FCOE_PARAMS_NUM_TASKS - 1);
+
+ /* Make sure num_queues is already set before calling this function */
+ if (!qedf->num_queues) {
+ QEDF_ERR(&(qedf->dbg_ctx), "num_queues is not set.\n");
+ return NULL;
+ }
+
+ if (max_xid <= min_xid || max_xid == FC_XID_UNKNOWN) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Invalid min_xid 0x%x and "
+ "max_xid 0x%x.\n", min_xid, max_xid);
+ return NULL;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "min xid 0x%x, max xid "
+ "0x%x.\n", min_xid, max_xid);
+
+ num_ios = max_xid - min_xid + 1;
+
+ cmgr = vzalloc(sizeof(struct qedf_cmd_mgr));
+ if (!cmgr) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Failed to alloc cmd mgr.\n");
+ return NULL;
+ }
+
+ cmgr->qedf = qedf;
+ spin_lock_init(&cmgr->lock);
+
+ /*
+ * Initialize list of qedf_ioreq.
+ */
+ xid = QEDF_MIN_XID;
+
+ for (i = 0; i < num_ios; i++) {
+ io_req = &cmgr->cmds[i];
+ INIT_DELAYED_WORK(&io_req->timeout_work, qedf_cmd_timeout);
+
+ io_req->xid = xid++;
+
+ INIT_DELAYED_WORK(&io_req->rrq_work, qedf_handle_rrq);
+
+ /* Allocate DMA memory to hold sense buffer */
+ io_req->sense_buffer = dma_alloc_coherent(&qedf->pdev->dev,
+ QEDF_SCSI_SENSE_BUFFERSIZE, &io_req->sense_buffer_dma,
+ GFP_KERNEL);
+ if (!io_req->sense_buffer)
+ goto mem_err;
+ }
+
+ /* Allocate pool of io_bdts - one for each qedf_ioreq */
+ cmgr->io_bdt_pool = kmalloc_array(num_ios, sizeof(struct io_bdt *),
+ GFP_KERNEL);
+
+ if (!cmgr->io_bdt_pool) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Failed to alloc io_bdt_pool.\n");
+ goto mem_err;
+ }
+
+ for (i = 0; i < num_ios; i++) {
+ cmgr->io_bdt_pool[i] = kmalloc(sizeof(struct io_bdt),
+ GFP_KERNEL);
+ if (!cmgr->io_bdt_pool[i]) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Failed to alloc "
+ "io_bdt_pool[%d].\n", i);
+ goto mem_err;
+ }
+ }
+
+ for (i = 0; i < num_ios; i++) {
+ bdt_info = cmgr->io_bdt_pool[i];
+ bdt_info->bd_tbl = dma_alloc_coherent(&qedf->pdev->dev,
+ QEDF_MAX_BDS_PER_CMD * sizeof(struct fcoe_sge),
+ &bdt_info->bd_tbl_dma, GFP_KERNEL);
+ if (!bdt_info->bd_tbl) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Failed to alloc "
+ "bdt_tbl[%d].\n", i);
+ goto mem_err;
+ }
+ }
+ atomic_set(&cmgr->free_list_cnt, num_ios);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "cmgr->free_list_cnt=%d.\n",
+ atomic_read(&cmgr->free_list_cnt));
+
+ return cmgr;
+
+mem_err:
+ qedf_cmd_mgr_free(cmgr);
+ return NULL;
+}
+
+struct qedf_ioreq *qedf_alloc_cmd(struct qedf_rport *fcport, u8 cmd_type)
+{
+ struct qedf_ctx *qedf = fcport->qedf;
+ struct qedf_cmd_mgr *cmd_mgr = qedf->cmd_mgr;
+ struct qedf_ioreq *io_req = NULL;
+ struct io_bdt *bd_tbl;
+ u16 xid;
+ uint32_t free_sqes;
+ int i;
+ unsigned long flags;
+
+ free_sqes = atomic_read(&fcport->free_sqes);
+
+ if (!free_sqes) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Returning NULL, free_sqes=%d.\n ",
+ free_sqes);
+ goto out_failed;
+ }
+
+ /* Limit the number of outstanding R/W tasks */
+ if ((atomic_read(&fcport->num_active_ios) >=
+ NUM_RW_TASKS_PER_CONNECTION)) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Returning NULL, num_active_ios=%d.\n",
+ atomic_read(&fcport->num_active_ios));
+ goto out_failed;
+ }
+
+ /* Limit global TIDs certain tasks */
+ if (atomic_read(&cmd_mgr->free_list_cnt) <= GBL_RSVD_TASKS) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Returning NULL, free_list_cnt=%d.\n",
+ atomic_read(&cmd_mgr->free_list_cnt));
+ goto out_failed;
+ }
+
+ spin_lock_irqsave(&cmd_mgr->lock, flags);
+ for (i = 0; i < FCOE_PARAMS_NUM_TASKS; i++) {
+ io_req = &cmd_mgr->cmds[cmd_mgr->idx];
+ cmd_mgr->idx++;
+ if (cmd_mgr->idx == FCOE_PARAMS_NUM_TASKS)
+ cmd_mgr->idx = 0;
+
+ /* Check to make sure command was previously freed */
+ if (!test_bit(QEDF_CMD_OUTSTANDING, &io_req->flags))
+ break;
+ }
+
+ if (i == FCOE_PARAMS_NUM_TASKS) {
+ spin_unlock_irqrestore(&cmd_mgr->lock, flags);
+ goto out_failed;
+ }
+
+ set_bit(QEDF_CMD_OUTSTANDING, &io_req->flags);
+ spin_unlock_irqrestore(&cmd_mgr->lock, flags);
+
+ atomic_inc(&fcport->num_active_ios);
+ atomic_dec(&fcport->free_sqes);
+ xid = io_req->xid;
+ atomic_dec(&cmd_mgr->free_list_cnt);
+
+ io_req->cmd_mgr = cmd_mgr;
+ io_req->fcport = fcport;
+
+ /* Hold the io_req against deletion */
+ kref_init(&io_req->refcount);
+
+ /* Bind io_bdt for this io_req */
+ /* Have a static link between io_req and io_bdt_pool */
+ bd_tbl = io_req->bd_tbl = cmd_mgr->io_bdt_pool[xid];
+ if (bd_tbl == NULL) {
+ QEDF_ERR(&(qedf->dbg_ctx), "bd_tbl is NULL, xid=%x.\n", xid);
+ kref_put(&io_req->refcount, qedf_release_cmd);
+ goto out_failed;
+ }
+ bd_tbl->io_req = io_req;
+ io_req->cmd_type = cmd_type;
+
+ /* Reset sequence offset data */
+ io_req->rx_buf_off = 0;
+ io_req->tx_buf_off = 0;
+ io_req->rx_id = 0xffff; /* No OX_ID */
+
+ return io_req;
+
+out_failed:
+ /* Record failure for stats and return NULL to caller */
+ qedf->alloc_failures++;
+ return NULL;
+}
+
+static void qedf_free_mp_resc(struct qedf_ioreq *io_req)
+{
+ struct qedf_mp_req *mp_req = &(io_req->mp_req);
+ struct qedf_ctx *qedf = io_req->fcport->qedf;
+ uint64_t sz = sizeof(struct fcoe_sge);
+
+ /* clear tm flags */
+ mp_req->tm_flags = 0;
+ if (mp_req->mp_req_bd) {
+ dma_free_coherent(&qedf->pdev->dev, sz,
+ mp_req->mp_req_bd, mp_req->mp_req_bd_dma);
+ mp_req->mp_req_bd = NULL;
+ }
+ if (mp_req->mp_resp_bd) {
+ dma_free_coherent(&qedf->pdev->dev, sz,
+ mp_req->mp_resp_bd, mp_req->mp_resp_bd_dma);
+ mp_req->mp_resp_bd = NULL;
+ }
+ if (mp_req->req_buf) {
+ dma_free_coherent(&qedf->pdev->dev, QEDF_PAGE_SIZE,
+ mp_req->req_buf, mp_req->req_buf_dma);
+ mp_req->req_buf = NULL;
+ }
+ if (mp_req->resp_buf) {
+ dma_free_coherent(&qedf->pdev->dev, QEDF_PAGE_SIZE,
+ mp_req->resp_buf, mp_req->resp_buf_dma);
+ mp_req->resp_buf = NULL;
+ }
+}
+
+void qedf_release_cmd(struct kref *ref)
+{
+ struct qedf_ioreq *io_req =
+ container_of(ref, struct qedf_ioreq, refcount);
+ struct qedf_cmd_mgr *cmd_mgr = io_req->cmd_mgr;
+ struct qedf_rport *fcport = io_req->fcport;
+
+ if (io_req->cmd_type == QEDF_ELS ||
+ io_req->cmd_type == QEDF_TASK_MGMT_CMD)
+ qedf_free_mp_resc(io_req);
+
+ atomic_inc(&cmd_mgr->free_list_cnt);
+ atomic_dec(&fcport->num_active_ios);
+ if (atomic_read(&fcport->num_active_ios) < 0)
+ QEDF_WARN(&(fcport->qedf->dbg_ctx), "active_ios < 0.\n");
+
+ /* Increment task retry identifier now that the request is released */
+ io_req->task_retry_identifier++;
+
+ clear_bit(QEDF_CMD_OUTSTANDING, &io_req->flags);
+}
+
+static int qedf_split_bd(struct qedf_ioreq *io_req, u64 addr, int sg_len,
+ int bd_index)
+{
+ struct fcoe_sge *bd = io_req->bd_tbl->bd_tbl;
+ int frag_size, sg_frags;
+
+ sg_frags = 0;
+ while (sg_len) {
+ if (sg_len > QEDF_BD_SPLIT_SZ)
+ frag_size = QEDF_BD_SPLIT_SZ;
+ else
+ frag_size = sg_len;
+ bd[bd_index + sg_frags].sge_addr.lo = U64_LO(addr);
+ bd[bd_index + sg_frags].sge_addr.hi = U64_HI(addr);
+ bd[bd_index + sg_frags].size = (uint16_t)frag_size;
+
+ addr += (u64)frag_size;
+ sg_frags++;
+ sg_len -= frag_size;
+ }
+ return sg_frags;
+}
+
+static int qedf_map_sg(struct qedf_ioreq *io_req)
+{
+ struct scsi_cmnd *sc = io_req->sc_cmd;
+ struct Scsi_Host *host = sc->device->host;
+ struct fc_lport *lport = shost_priv(host);
+ struct qedf_ctx *qedf = lport_priv(lport);
+ struct fcoe_sge *bd = io_req->bd_tbl->bd_tbl;
+ struct scatterlist *sg;
+ int byte_count = 0;
+ int sg_count = 0;
+ int bd_count = 0;
+ int sg_frags;
+ unsigned int sg_len;
+ u64 addr, end_addr;
+ int i;
+
+ sg_count = dma_map_sg(&qedf->pdev->dev, scsi_sglist(sc),
+ scsi_sg_count(sc), sc->sc_data_direction);
+
+ sg = scsi_sglist(sc);
+
+ /*
+ * New condition to send single SGE as cached-SGL with length less
+ * than 64k.
+ */
+ if ((sg_count == 1) && (sg_dma_len(sg) <=
+ QEDF_MAX_SGLEN_FOR_CACHESGL)) {
+ sg_len = sg_dma_len(sg);
+ addr = (u64)sg_dma_address(sg);
+
+ bd[bd_count].sge_addr.lo = (addr & 0xffffffff);
+ bd[bd_count].sge_addr.hi = (addr >> 32);
+ bd[bd_count].size = (u16)sg_len;
+
+ return ++bd_count;
+ }
+
+ scsi_for_each_sg(sc, sg, sg_count, i) {
+ sg_len = sg_dma_len(sg);
+ addr = (u64)sg_dma_address(sg);
+ end_addr = (u64)(addr + sg_len);
+
+ /*
+ * First s/g element in the list so check if the end_addr
+ * is paged aligned. Also check to make sure the length is
+ * at least page size.
+ */
+ if ((i == 0) && (sg_count > 1) &&
+ ((end_addr % QEDF_PAGE_SIZE) ||
+ sg_len < QEDF_PAGE_SIZE))
+ io_req->use_slowpath = true;
+ /*
+ * Last s/g element so check if the start address is paged
+ * aligned.
+ */
+ else if ((i == (sg_count - 1)) && (sg_count > 1) &&
+ (addr % QEDF_PAGE_SIZE))
+ io_req->use_slowpath = true;
+ /*
+ * Intermediate s/g element so check if start and end address
+ * is page aligned.
+ */
+ else if ((i != 0) && (i != (sg_count - 1)) &&
+ ((addr % QEDF_PAGE_SIZE) || (end_addr % QEDF_PAGE_SIZE)))
+ io_req->use_slowpath = true;
+
+ if (sg_len > QEDF_MAX_BD_LEN) {
+ sg_frags = qedf_split_bd(io_req, addr, sg_len,
+ bd_count);
+ } else {
+ sg_frags = 1;
+ bd[bd_count].sge_addr.lo = U64_LO(addr);
+ bd[bd_count].sge_addr.hi = U64_HI(addr);
+ bd[bd_count].size = (uint16_t)sg_len;
+ }
+
+ bd_count += sg_frags;
+ byte_count += sg_len;
+ }
+
+ if (byte_count != scsi_bufflen(sc))
+ QEDF_ERR(&(qedf->dbg_ctx), "byte_count = %d != "
+ "scsi_bufflen = %d, task_id = 0x%x.\n", byte_count,
+ scsi_bufflen(sc), io_req->xid);
+
+ return bd_count;
+}
+
+static int qedf_build_bd_list_from_sg(struct qedf_ioreq *io_req)
+{
+ struct scsi_cmnd *sc = io_req->sc_cmd;
+ struct fcoe_sge *bd = io_req->bd_tbl->bd_tbl;
+ int bd_count;
+
+ if (scsi_sg_count(sc)) {
+ bd_count = qedf_map_sg(io_req);
+ if (bd_count == 0)
+ return -ENOMEM;
+ } else {
+ bd_count = 0;
+ bd[0].sge_addr.lo = bd[0].sge_addr.hi = 0;
+ bd[0].size = 0;
+ }
+ io_req->bd_tbl->bd_valid = bd_count;
+
+ return 0;
+}
+
+static void qedf_build_fcp_cmnd(struct qedf_ioreq *io_req,
+ struct fcp_cmnd *fcp_cmnd)
+{
+ struct scsi_cmnd *sc_cmd = io_req->sc_cmd;
+
+ /* fcp_cmnd is 32 bytes */
+ memset(fcp_cmnd, 0, FCP_CMND_LEN);
+
+ /* 8 bytes: SCSI LUN info */
+ int_to_scsilun(sc_cmd->device->lun,
+ (struct scsi_lun *)&fcp_cmnd->fc_lun);
+
+ /* 4 bytes: flag info */
+ fcp_cmnd->fc_pri_ta = 0;
+ fcp_cmnd->fc_tm_flags = io_req->mp_req.tm_flags;
+ fcp_cmnd->fc_flags = io_req->io_req_flags;
+ fcp_cmnd->fc_cmdref = 0;
+
+ /* Populate data direction */
+ if (sc_cmd->sc_data_direction == DMA_TO_DEVICE)
+ fcp_cmnd->fc_flags |= FCP_CFL_WRDATA;
+ else if (sc_cmd->sc_data_direction == DMA_FROM_DEVICE)
+ fcp_cmnd->fc_flags |= FCP_CFL_RDDATA;
+
+ fcp_cmnd->fc_pri_ta = FCP_PTA_SIMPLE;
+
+ /* 16 bytes: CDB information */
+ memcpy(fcp_cmnd->fc_cdb, sc_cmd->cmnd, sc_cmd->cmd_len);
+
+ /* 4 bytes: FCP data length */
+ fcp_cmnd->fc_dl = htonl(io_req->data_xfer_len);
+
+}
+
+static void qedf_init_task(struct qedf_rport *fcport, struct fc_lport *lport,
+ struct qedf_ioreq *io_req, u32 *ptu_invalidate,
+ struct fcoe_task_context *task_ctx)
+{
+ enum fcoe_task_type task_type;
+ struct scsi_cmnd *sc_cmd = io_req->sc_cmd;
+ struct io_bdt *bd_tbl = io_req->bd_tbl;
+ union fcoe_data_desc_ctx *data_desc;
+ u32 *fcp_cmnd;
+ u32 tmp_fcp_cmnd[8];
+ int cnt, i;
+ int bd_count;
+ struct qedf_ctx *qedf = fcport->qedf;
+ uint16_t cq_idx = smp_processor_id() % qedf->num_queues;
+ u8 tmp_sgl_mode = 0;
+ u8 mst_sgl_mode = 0;
+
+ memset(task_ctx, 0, sizeof(struct fcoe_task_context));
+ io_req->task = task_ctx;
+
+ if (sc_cmd->sc_data_direction == DMA_TO_DEVICE)
+ task_type = FCOE_TASK_TYPE_WRITE_INITIATOR;
+ else
+ task_type = FCOE_TASK_TYPE_READ_INITIATOR;
+
+ /* Y Storm context */
+ task_ctx->ystorm_st_context.expect_first_xfer = 1;
+ task_ctx->ystorm_st_context.data_2_trns_rem = io_req->data_xfer_len;
+ /* Check if this is required */
+ task_ctx->ystorm_st_context.ox_id = io_req->xid;
+ task_ctx->ystorm_st_context.task_rety_identifier =
+ io_req->task_retry_identifier;
+
+ /* T Storm ag context */
+ SET_FIELD(task_ctx->tstorm_ag_context.flags0,
+ TSTORM_FCOE_TASK_AG_CTX_CONNECTION_TYPE, PROTOCOLID_FCOE);
+ task_ctx->tstorm_ag_context.icid = (u16)fcport->fw_cid;
+
+ /* T Storm st context */
+ SET_FIELD(task_ctx->tstorm_st_context.read_write.flags,
+ FCOE_TSTORM_FCOE_TASK_ST_CTX_READ_WRITE_EXP_FIRST_FRAME,
+ 1);
+ task_ctx->tstorm_st_context.read_write.rx_id = 0xffff;
+
+ task_ctx->tstorm_st_context.read_only.dev_type =
+ FCOE_TASK_DEV_TYPE_DISK;
+ task_ctx->tstorm_st_context.read_only.conf_supported = 0;
+ task_ctx->tstorm_st_context.read_only.cid = fcport->fw_cid;
+
+ /* Completion queue for response. */
+ task_ctx->tstorm_st_context.read_only.glbl_q_num = cq_idx;
+ task_ctx->tstorm_st_context.read_only.fcp_cmd_trns_size =
+ io_req->data_xfer_len;
+ task_ctx->tstorm_st_context.read_write.e_d_tov_exp_timeout_val =
+ lport->e_d_tov;
+
+ task_ctx->ustorm_ag_context.global_cq_num = cq_idx;
+ io_req->fp_idx = cq_idx;
+
+ bd_count = bd_tbl->bd_valid;
+ if (task_type == FCOE_TASK_TYPE_WRITE_INITIATOR) {
+ /* Setup WRITE task */
+ struct fcoe_sge *fcoe_bd_tbl = bd_tbl->bd_tbl;
+
+ task_ctx->ystorm_st_context.task_type =
+ FCOE_TASK_TYPE_WRITE_INITIATOR;
+ data_desc = &task_ctx->ystorm_st_context.data_desc;
+
+ if (io_req->use_slowpath) {
+ SET_FIELD(task_ctx->ystorm_st_context.sgl_mode,
+ YSTORM_FCOE_TASK_ST_CTX_TX_SGL_MODE,
+ FCOE_SLOW_SGL);
+ data_desc->slow.base_sgl_addr.lo =
+ U64_LO(bd_tbl->bd_tbl_dma);
+ data_desc->slow.base_sgl_addr.hi =
+ U64_HI(bd_tbl->bd_tbl_dma);
+ data_desc->slow.remainder_num_sges = bd_count;
+ data_desc->slow.curr_sge_off = 0;
+ data_desc->slow.curr_sgl_index = 0;
+ qedf->slow_sge_ios++;
+ io_req->sge_type = QEDF_IOREQ_SLOW_SGE;
+ } else {
+ SET_FIELD(task_ctx->ystorm_st_context.sgl_mode,
+ YSTORM_FCOE_TASK_ST_CTX_TX_SGL_MODE,
+ (bd_count <= 4) ? (enum fcoe_sgl_mode)bd_count :
+ FCOE_MUL_FAST_SGES);
+
+ if (bd_count == 1) {
+ data_desc->single_sge.sge_addr.lo =
+ fcoe_bd_tbl->sge_addr.lo;
+ data_desc->single_sge.sge_addr.hi =
+ fcoe_bd_tbl->sge_addr.hi;
+ data_desc->single_sge.size =
+ fcoe_bd_tbl->size;
+ data_desc->single_sge.is_valid_sge = 0;
+ qedf->single_sge_ios++;
+ io_req->sge_type = QEDF_IOREQ_SINGLE_SGE;
+ } else {
+ data_desc->fast.sgl_start_addr.lo =
+ U64_LO(bd_tbl->bd_tbl_dma);
+ data_desc->fast.sgl_start_addr.hi =
+ U64_HI(bd_tbl->bd_tbl_dma);
+ data_desc->fast.sgl_byte_offset =
+ data_desc->fast.sgl_start_addr.lo &
+ (QEDF_PAGE_SIZE - 1);
+ if (data_desc->fast.sgl_byte_offset > 0)
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "byte_offset=%u for xid=0x%x.\n",
+ io_req->xid,
+ data_desc->fast.sgl_byte_offset);
+ data_desc->fast.task_reuse_cnt =
+ io_req->reuse_count;
+ io_req->reuse_count++;
+ if (io_req->reuse_count == QEDF_MAX_REUSE) {
+ *ptu_invalidate = 1;
+ io_req->reuse_count = 0;
+ }
+ qedf->fast_sge_ios++;
+ io_req->sge_type = QEDF_IOREQ_FAST_SGE;
+ }
+ }
+
+ /* T Storm context */
+ task_ctx->tstorm_st_context.read_only.task_type =
+ FCOE_TASK_TYPE_WRITE_INITIATOR;
+
+ /* M Storm context */
+ tmp_sgl_mode = GET_FIELD(task_ctx->ystorm_st_context.sgl_mode,
+ YSTORM_FCOE_TASK_ST_CTX_TX_SGL_MODE);
+ SET_FIELD(task_ctx->mstorm_st_context.non_fp.tx_rx_sgl_mode,
+ FCOE_MSTORM_FCOE_TASK_ST_CTX_NON_FP_TX_SGL_MODE,
+ tmp_sgl_mode);
+
+ } else {
+ /* Setup READ task */
+
+ /* M Storm context */
+ struct fcoe_sge *fcoe_bd_tbl = bd_tbl->bd_tbl;
+
+ data_desc = &task_ctx->mstorm_st_context.fp.data_desc;
+ task_ctx->mstorm_st_context.fp.data_2_trns_rem =
+ io_req->data_xfer_len;
+
+ if (io_req->use_slowpath) {
+ SET_FIELD(
+ task_ctx->mstorm_st_context.non_fp.tx_rx_sgl_mode,
+ FCOE_MSTORM_FCOE_TASK_ST_CTX_NON_FP_RX_SGL_MODE,
+ FCOE_SLOW_SGL);
+ data_desc->slow.base_sgl_addr.lo =
+ U64_LO(bd_tbl->bd_tbl_dma);
+ data_desc->slow.base_sgl_addr.hi =
+ U64_HI(bd_tbl->bd_tbl_dma);
+ data_desc->slow.remainder_num_sges =
+ bd_count;
+ data_desc->slow.curr_sge_off = 0;
+ data_desc->slow.curr_sgl_index = 0;
+ qedf->slow_sge_ios++;
+ io_req->sge_type = QEDF_IOREQ_SLOW_SGE;
+ } else {
+ SET_FIELD(
+ task_ctx->mstorm_st_context.non_fp.tx_rx_sgl_mode,
+ FCOE_MSTORM_FCOE_TASK_ST_CTX_NON_FP_RX_SGL_MODE,
+ (bd_count <= 4) ? (enum fcoe_sgl_mode)bd_count :
+ FCOE_MUL_FAST_SGES);
+
+ if (bd_count == 1) {
+ data_desc->single_sge.sge_addr.lo =
+ fcoe_bd_tbl->sge_addr.lo;
+ data_desc->single_sge.sge_addr.hi =
+ fcoe_bd_tbl->sge_addr.hi;
+ data_desc->single_sge.size =
+ fcoe_bd_tbl->size;
+ data_desc->single_sge.is_valid_sge = 0;
+ qedf->single_sge_ios++;
+ io_req->sge_type = QEDF_IOREQ_SINGLE_SGE;
+ } else {
+ data_desc->fast.sgl_start_addr.lo =
+ U64_LO(bd_tbl->bd_tbl_dma);
+ data_desc->fast.sgl_start_addr.hi =
+ U64_HI(bd_tbl->bd_tbl_dma);
+ data_desc->fast.sgl_byte_offset = 0;
+ data_desc->fast.task_reuse_cnt =
+ io_req->reuse_count;
+ io_req->reuse_count++;
+ if (io_req->reuse_count == QEDF_MAX_REUSE) {
+ *ptu_invalidate = 1;
+ io_req->reuse_count = 0;
+ }
+ qedf->fast_sge_ios++;
+ io_req->sge_type = QEDF_IOREQ_FAST_SGE;
+ }
+ }
+
+ /* Y Storm context */
+ task_ctx->ystorm_st_context.expect_first_xfer = 0;
+ task_ctx->ystorm_st_context.task_type =
+ FCOE_TASK_TYPE_READ_INITIATOR;
+
+ /* T Storm context */
+ task_ctx->tstorm_st_context.read_only.task_type =
+ FCOE_TASK_TYPE_READ_INITIATOR;
+ mst_sgl_mode = GET_FIELD(
+ task_ctx->mstorm_st_context.non_fp.tx_rx_sgl_mode,
+ FCOE_MSTORM_FCOE_TASK_ST_CTX_NON_FP_RX_SGL_MODE);
+ SET_FIELD(task_ctx->tstorm_st_context.read_write.flags,
+ FCOE_TSTORM_FCOE_TASK_ST_CTX_READ_WRITE_RX_SGL_MODE,
+ mst_sgl_mode);
+ }
+
+ /* fill FCP_CMND IU */
+ fcp_cmnd = (u32 *)task_ctx->ystorm_st_context.tx_info_union.fcp_cmd_payload.opaque;
+ qedf_build_fcp_cmnd(io_req, (struct fcp_cmnd *)&tmp_fcp_cmnd);
+
+ /* Swap fcp_cmnd since FC is big endian */
+ cnt = sizeof(struct fcp_cmnd) / sizeof(u32);
+
+ for (i = 0; i < cnt; i++) {
+ *fcp_cmnd = cpu_to_be32(tmp_fcp_cmnd[i]);
+ fcp_cmnd++;
+ }
+
+ /* M Storm context - Sense buffer */
+ task_ctx->mstorm_st_context.non_fp.rsp_buf_addr.lo =
+ U64_LO(io_req->sense_buffer_dma);
+ task_ctx->mstorm_st_context.non_fp.rsp_buf_addr.hi =
+ U64_HI(io_req->sense_buffer_dma);
+}
+
+void qedf_init_mp_task(struct qedf_ioreq *io_req,
+ struct fcoe_task_context *task_ctx)
+{
+ struct qedf_mp_req *mp_req = &(io_req->mp_req);
+ struct qedf_rport *fcport = io_req->fcport;
+ struct qedf_ctx *qedf = io_req->fcport->qedf;
+ struct fc_frame_header *fc_hdr;
+ enum fcoe_task_type task_type = 0;
+ union fcoe_data_desc_ctx *data_desc;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Initializing MP task "
+ "for cmd_type = %d\n", io_req->cmd_type);
+
+ qedf->control_requests++;
+
+ /* Obtain task_type */
+ if ((io_req->cmd_type == QEDF_TASK_MGMT_CMD) ||
+ (io_req->cmd_type == QEDF_ELS)) {
+ task_type = FCOE_TASK_TYPE_MIDPATH;
+ } else if (io_req->cmd_type == QEDF_ABTS) {
+ task_type = FCOE_TASK_TYPE_ABTS;
+ }
+
+ memset(task_ctx, 0, sizeof(struct fcoe_task_context));
+
+ /* Setup the task from io_req for easy reference */
+ io_req->task = task_ctx;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "task type = %d\n",
+ task_type);
+
+ /* YSTORM only */
+ {
+ /* Initialize YSTORM task context */
+ struct fcoe_tx_mid_path_params *task_fc_hdr =
+ &task_ctx->ystorm_st_context.tx_info_union.tx_params.mid_path;
+ memset(task_fc_hdr, 0, sizeof(struct fcoe_tx_mid_path_params));
+ task_ctx->ystorm_st_context.task_rety_identifier =
+ io_req->task_retry_identifier;
+
+ /* Init SGL parameters */
+ if ((task_type == FCOE_TASK_TYPE_MIDPATH) ||
+ (task_type == FCOE_TASK_TYPE_UNSOLICITED)) {
+ data_desc = &task_ctx->ystorm_st_context.data_desc;
+ data_desc->slow.base_sgl_addr.lo =
+ U64_LO(mp_req->mp_req_bd_dma);
+ data_desc->slow.base_sgl_addr.hi =
+ U64_HI(mp_req->mp_req_bd_dma);
+ data_desc->slow.remainder_num_sges = 1;
+ data_desc->slow.curr_sge_off = 0;
+ data_desc->slow.curr_sgl_index = 0;
+ }
+
+ fc_hdr = &(mp_req->req_fc_hdr);
+ if (task_type == FCOE_TASK_TYPE_MIDPATH) {
+ fc_hdr->fh_ox_id = io_req->xid;
+ fc_hdr->fh_rx_id = htons(0xffff);
+ } else if (task_type == FCOE_TASK_TYPE_UNSOLICITED) {
+ fc_hdr->fh_rx_id = io_req->xid;
+ }
+
+ /* Fill FC Header into middle path buffer */
+ task_fc_hdr->parameter = fc_hdr->fh_parm_offset;
+ task_fc_hdr->r_ctl = fc_hdr->fh_r_ctl;
+ task_fc_hdr->type = fc_hdr->fh_type;
+ task_fc_hdr->cs_ctl = fc_hdr->fh_cs_ctl;
+ task_fc_hdr->df_ctl = fc_hdr->fh_df_ctl;
+ task_fc_hdr->rx_id = fc_hdr->fh_rx_id;
+ task_fc_hdr->ox_id = fc_hdr->fh_ox_id;
+
+ task_ctx->ystorm_st_context.data_2_trns_rem =
+ io_req->data_xfer_len;
+ task_ctx->ystorm_st_context.task_type = task_type;
+ }
+
+ /* TSTORM ONLY */
+ {
+ task_ctx->tstorm_ag_context.icid = (u16)fcport->fw_cid;
+ task_ctx->tstorm_st_context.read_only.cid = fcport->fw_cid;
+ /* Always send middle-path repsonses on CQ #0 */
+ task_ctx->tstorm_st_context.read_only.glbl_q_num = 0;
+ io_req->fp_idx = 0;
+ SET_FIELD(task_ctx->tstorm_ag_context.flags0,
+ TSTORM_FCOE_TASK_AG_CTX_CONNECTION_TYPE,
+ PROTOCOLID_FCOE);
+ task_ctx->tstorm_st_context.read_only.task_type = task_type;
+ SET_FIELD(task_ctx->tstorm_st_context.read_write.flags,
+ FCOE_TSTORM_FCOE_TASK_ST_CTX_READ_WRITE_EXP_FIRST_FRAME,
+ 1);
+ task_ctx->tstorm_st_context.read_write.rx_id = 0xffff;
+ }
+
+ /* MSTORM only */
+ {
+ if (task_type == FCOE_TASK_TYPE_MIDPATH) {
+ /* Initialize task context */
+ data_desc = &task_ctx->mstorm_st_context.fp.data_desc;
+
+ /* Set cache sges address and length */
+ data_desc->slow.base_sgl_addr.lo =
+ U64_LO(mp_req->mp_resp_bd_dma);
+ data_desc->slow.base_sgl_addr.hi =
+ U64_HI(mp_req->mp_resp_bd_dma);
+ data_desc->slow.remainder_num_sges = 1;
+ data_desc->slow.curr_sge_off = 0;
+ data_desc->slow.curr_sgl_index = 0;
+
+ /*
+ * Also need to fil in non-fastpath response address
+ * for middle path commands.
+ */
+ task_ctx->mstorm_st_context.non_fp.rsp_buf_addr.lo =
+ U64_LO(mp_req->mp_resp_bd_dma);
+ task_ctx->mstorm_st_context.non_fp.rsp_buf_addr.hi =
+ U64_HI(mp_req->mp_resp_bd_dma);
+ }
+ }
+
+ /* USTORM ONLY */
+ {
+ task_ctx->ustorm_ag_context.global_cq_num = 0;
+ }
+
+ /* I/O stats. Middle path commands always use slow SGEs */
+ qedf->slow_sge_ios++;
+ io_req->sge_type = QEDF_IOREQ_SLOW_SGE;
+}
+
+void qedf_add_to_sq(struct qedf_rport *fcport, u16 xid, u32 ptu_invalidate,
+ enum fcoe_task_type req_type, u32 offset)
+{
+ struct fcoe_wqe *sqe;
+ uint16_t total_sqe = (fcport->sq_mem_size)/(sizeof(struct fcoe_wqe));
+
+ sqe = &fcport->sq[fcport->sq_prod_idx];
+
+ fcport->sq_prod_idx++;
+ fcport->fw_sq_prod_idx++;
+ if (fcport->sq_prod_idx == total_sqe)
+ fcport->sq_prod_idx = 0;
+
+ switch (req_type) {
+ case FCOE_TASK_TYPE_WRITE_INITIATOR:
+ case FCOE_TASK_TYPE_READ_INITIATOR:
+ SET_FIELD(sqe->flags, FCOE_WQE_REQ_TYPE, SEND_FCOE_CMD);
+ if (ptu_invalidate)
+ SET_FIELD(sqe->flags, FCOE_WQE_INVALIDATE_PTU, 1);
+ break;
+ case FCOE_TASK_TYPE_MIDPATH:
+ SET_FIELD(sqe->flags, FCOE_WQE_REQ_TYPE, SEND_FCOE_MIDPATH);
+ break;
+ case FCOE_TASK_TYPE_ABTS:
+ SET_FIELD(sqe->flags, FCOE_WQE_REQ_TYPE,
+ SEND_FCOE_ABTS_REQUEST);
+ break;
+ case FCOE_TASK_TYPE_EXCHANGE_CLEANUP:
+ SET_FIELD(sqe->flags, FCOE_WQE_REQ_TYPE,
+ FCOE_EXCHANGE_CLEANUP);
+ break;
+ case FCOE_TASK_TYPE_SEQUENCE_CLEANUP:
+ SET_FIELD(sqe->flags, FCOE_WQE_REQ_TYPE,
+ FCOE_SEQUENCE_RECOVERY);
+ /* NOTE: offset param only used for sequence recovery */
+ sqe->additional_info_union.seq_rec_updated_offset = offset;
+ break;
+ case FCOE_TASK_TYPE_UNSOLICITED:
+ break;
+ default:
+ break;
+ }
+
+ sqe->task_id = xid;
+
+ /* Make sure SQ data is coherent */
+ wmb();
+
+}
+
+void qedf_ring_doorbell(struct qedf_rport *fcport)
+{
+ struct fcoe_db_data dbell = { 0 };
+
+ dbell.agg_flags = 0;
+
+ dbell.params |= DB_DEST_XCM << FCOE_DB_DATA_DEST_SHIFT;
+ dbell.params |= DB_AGG_CMD_SET << FCOE_DB_DATA_AGG_CMD_SHIFT;
+ dbell.params |= DQ_XCM_FCOE_SQ_PROD_CMD <<
+ FCOE_DB_DATA_AGG_VAL_SEL_SHIFT;
+
+ dbell.sq_prod = fcport->fw_sq_prod_idx;
+ writel(*(u32 *)&dbell, fcport->p_doorbell);
+ /* Make sure SQ index is updated so f/w prcesses requests in order */
+ wmb();
+ mmiowb();
+}
+
+static void qedf_trace_io(struct qedf_rport *fcport, struct qedf_ioreq *io_req,
+ int8_t direction)
+{
+ struct qedf_ctx *qedf = fcport->qedf;
+ struct qedf_io_log *io_log;
+ struct scsi_cmnd *sc_cmd = io_req->sc_cmd;
+ unsigned long flags;
+ uint8_t op;
+
+ spin_lock_irqsave(&qedf->io_trace_lock, flags);
+
+ io_log = &qedf->io_trace_buf[qedf->io_trace_idx];
+ io_log->direction = direction;
+ io_log->task_id = io_req->xid;
+ io_log->port_id = fcport->rdata->ids.port_id;
+ io_log->lun = sc_cmd->device->lun;
+ io_log->op = op = sc_cmd->cmnd[0];
+ io_log->lba[0] = sc_cmd->cmnd[2];
+ io_log->lba[1] = sc_cmd->cmnd[3];
+ io_log->lba[2] = sc_cmd->cmnd[4];
+ io_log->lba[3] = sc_cmd->cmnd[5];
+ io_log->bufflen = scsi_bufflen(sc_cmd);
+ io_log->sg_count = scsi_sg_count(sc_cmd);
+ io_log->result = sc_cmd->result;
+ io_log->jiffies = jiffies;
+ io_log->refcount = atomic_read(&io_req->refcount.refcount);
+
+ if (direction == QEDF_IO_TRACE_REQ) {
+ /* For requests we only care abot the submission CPU */
+ io_log->req_cpu = io_req->cpu;
+ io_log->int_cpu = 0;
+ io_log->rsp_cpu = 0;
+ } else if (direction == QEDF_IO_TRACE_RSP) {
+ io_log->req_cpu = io_req->cpu;
+ io_log->int_cpu = io_req->int_cpu;
+ io_log->rsp_cpu = smp_processor_id();
+ }
+
+ io_log->sge_type = io_req->sge_type;
+
+ qedf->io_trace_idx++;
+ if (qedf->io_trace_idx == QEDF_IO_TRACE_SIZE)
+ qedf->io_trace_idx = 0;
+
+ spin_unlock_irqrestore(&qedf->io_trace_lock, flags);
+}
+
+void qedf_inc_percpu_requests(unsigned long cpu)
+{
+ struct qedf_percpu_iothread_s *iothread;
+
+ iothread = &per_cpu(qedf_percpu_iothreads, cpu);
+ iothread->requests++;
+}
+
+int qedf_post_io_req(struct qedf_rport *fcport, struct qedf_ioreq *io_req)
+{
+ struct scsi_cmnd *sc_cmd = io_req->sc_cmd;
+ struct Scsi_Host *host = sc_cmd->device->host;
+ struct fc_lport *lport = shost_priv(host);
+ struct qedf_ctx *qedf = lport_priv(lport);
+ struct fcoe_task_context *task_ctx;
+ u16 xid;
+ enum fcoe_task_type req_type = 0;
+ u32 ptu_invalidate = 0;
+
+ /* Initialize rest of io_req fileds */
+ io_req->data_xfer_len = scsi_bufflen(sc_cmd);
+ sc_cmd->SCp.ptr = (char *)io_req;
+ io_req->use_slowpath = false; /* Assume fast SGL by default */
+
+ /* Record which cpu this request is associated with */
+ io_req->cpu = smp_processor_id();
+ qedf_inc_percpu_requests(io_req->cpu);
+
+ if (sc_cmd->sc_data_direction == DMA_FROM_DEVICE) {
+ req_type = FCOE_TASK_TYPE_READ_INITIATOR;
+ io_req->io_req_flags = QEDF_READ;
+ qedf->input_requests++;
+ } else if (sc_cmd->sc_data_direction == DMA_TO_DEVICE) {
+ req_type = FCOE_TASK_TYPE_WRITE_INITIATOR;
+ io_req->io_req_flags = QEDF_WRITE;
+ qedf->output_requests++;
+ } else {
+ io_req->io_req_flags = 0;
+ qedf->control_requests++;
+ }
+
+ xid = io_req->xid;
+
+ /* Build buffer descriptor list for firmware from sg list */
+ if (qedf_build_bd_list_from_sg(io_req)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "BD list creation failed.\n");
+ kref_put(&io_req->refcount, qedf_release_cmd);
+ return -EAGAIN;
+ }
+
+ /* Get the task context */
+ task_ctx = qedf_get_task_mem(&qedf->tasks, xid);
+ if (!task_ctx) {
+ QEDF_WARN(&(qedf->dbg_ctx), "task_ctx is NULL, xid=%d.\n",
+ xid);
+ kref_put(&io_req->refcount, qedf_release_cmd);
+ return -EINVAL;
+ }
+
+ qedf_init_task(fcport, lport, io_req, &ptu_invalidate, task_ctx);
+
+ if (!test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Session not offloaded yet.\n");
+ kref_put(&io_req->refcount, qedf_release_cmd);
+ }
+
+ /* Obtain free SQ entry */
+ qedf_add_to_sq(fcport, xid, ptu_invalidate, req_type, 0);
+
+ /* Ring doorbell */
+ qedf_ring_doorbell(fcport);
+
+ if (qedf_io_tracing && io_req->sc_cmd)
+ qedf_trace_io(fcport, io_req, QEDF_IO_TRACE_REQ);
+
+ return false;
+}
+
+int
+qedf_queuecommand(struct Scsi_Host *host, struct scsi_cmnd *sc_cmd)
+{
+ struct fc_lport *lport = shost_priv(host);
+ struct qedf_ctx *qedf = lport_priv(lport);
+ struct fc_rport *rport = starget_to_rport(scsi_target(sc_cmd->device));
+ struct fc_rport_libfc_priv *rp = rport->dd_data;
+ struct qedf_rport *fcport = rport->dd_data;
+ struct qedf_ioreq *io_req;
+ int rc = 0;
+ int rval;
+ unsigned long flags = 0;
+
+
+ if (test_bit(QEDF_UNLOADING, &qedf->flags)) {
+ sc_cmd->result = DID_NO_CONNECT << 16;
+ sc_cmd->scsi_done(sc_cmd);
+ return 0;
+ }
+
+ rval = fc_remote_port_chkready(rport);
+ if (rval) {
+ sc_cmd->result = rval;
+ sc_cmd->scsi_done(sc_cmd);
+ return 0;
+ }
+
+ /* Retry command if we are doing a qed drain operation */
+ if (test_bit(QEDF_DRAIN_ACTIVE, &qedf->flags)) {
+ rc = SCSI_MLQUEUE_HOST_BUSY;
+ goto exit_qcmd;
+ }
+
+ if (lport->state != LPORT_ST_READY ||
+ atomic_read(&qedf->link_state) != QEDF_LINK_UP) {
+ rc = SCSI_MLQUEUE_HOST_BUSY;
+ goto exit_qcmd;
+ }
+
+ /* rport and tgt are allocated together, so tgt should be non-NULL */
+ fcport = (struct qedf_rport *)&rp[1];
+
+ if (!test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ /*
+ * Session is not offloaded yet. Let SCSI-ml retry
+ * the command.
+ */
+ rc = SCSI_MLQUEUE_TARGET_BUSY;
+ goto exit_qcmd;
+ }
+ if (fcport->retry_delay_timestamp) {
+ if (time_after(jiffies, fcport->retry_delay_timestamp)) {
+ fcport->retry_delay_timestamp = 0;
+ } else {
+ /* If retry_delay timer is active, flow off the ML */
+ rc = SCSI_MLQUEUE_TARGET_BUSY;
+ goto exit_qcmd;
+ }
+ }
+
+ io_req = qedf_alloc_cmd(fcport, QEDF_SCSI_CMD);
+ if (!io_req) {
+ rc = SCSI_MLQUEUE_HOST_BUSY;
+ goto exit_qcmd;
+ }
+
+ io_req->sc_cmd = sc_cmd;
+
+ /* Take fcport->rport_lock for posting to fcport send queue */
+ spin_lock_irqsave(&fcport->rport_lock, flags);
+ if (qedf_post_io_req(fcport, io_req)) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Unable to post io_req\n");
+ /* Return SQE to pool */
+ atomic_inc(&fcport->free_sqes);
+ rc = SCSI_MLQUEUE_HOST_BUSY;
+ }
+ spin_unlock_irqrestore(&fcport->rport_lock, flags);
+
+exit_qcmd:
+ return rc;
+}
+
+static void qedf_parse_fcp_rsp(struct qedf_ioreq *io_req,
+ struct fcoe_cqe_rsp_info *fcp_rsp)
+{
+ struct scsi_cmnd *sc_cmd = io_req->sc_cmd;
+ struct qedf_ctx *qedf = io_req->fcport->qedf;
+ u8 rsp_flags = fcp_rsp->rsp_flags.flags;
+ int fcp_sns_len = 0;
+ int fcp_rsp_len = 0;
+ uint8_t *rsp_info, *sense_data;
+
+ io_req->fcp_status = FC_GOOD;
+ io_req->fcp_resid = 0;
+ if (rsp_flags & (FCOE_FCP_RSP_FLAGS_FCP_RESID_OVER |
+ FCOE_FCP_RSP_FLAGS_FCP_RESID_UNDER))
+ io_req->fcp_resid = fcp_rsp->fcp_resid;
+
+ io_req->scsi_comp_flags = rsp_flags;
+ CMD_SCSI_STATUS(sc_cmd) = io_req->cdb_status =
+ fcp_rsp->scsi_status_code;
+
+ if (rsp_flags &
+ FCOE_FCP_RSP_FLAGS_FCP_RSP_LEN_VALID)
+ fcp_rsp_len = fcp_rsp->fcp_rsp_len;
+
+ if (rsp_flags &
+ FCOE_FCP_RSP_FLAGS_FCP_SNS_LEN_VALID)
+ fcp_sns_len = fcp_rsp->fcp_sns_len;
+
+ io_req->fcp_rsp_len = fcp_rsp_len;
+ io_req->fcp_sns_len = fcp_sns_len;
+ rsp_info = sense_data = io_req->sense_buffer;
+
+ /* fetch fcp_rsp_code */
+ if ((fcp_rsp_len == 4) || (fcp_rsp_len == 8)) {
+ /* Only for task management function */
+ io_req->fcp_rsp_code = rsp_info[3];
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "fcp_rsp_code = %d\n", io_req->fcp_rsp_code);
+ /* Adjust sense-data location. */
+ sense_data += fcp_rsp_len;
+ }
+
+ if (fcp_sns_len > SCSI_SENSE_BUFFERSIZE) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Truncating sense buffer\n");
+ fcp_sns_len = SCSI_SENSE_BUFFERSIZE;
+ }
+
+ memset(sc_cmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
+ if (fcp_sns_len)
+ memcpy(sc_cmd->sense_buffer, sense_data,
+ fcp_sns_len);
+}
+
+static void qedf_unmap_sg_list(struct qedf_ctx *qedf, struct qedf_ioreq *io_req)
+{
+ struct scsi_cmnd *sc = io_req->sc_cmd;
+
+ if (io_req->bd_tbl->bd_valid && sc && scsi_sg_count(sc)) {
+ dma_unmap_sg(&qedf->pdev->dev, scsi_sglist(sc),
+ scsi_sg_count(sc), sc->sc_data_direction);
+ io_req->bd_tbl->bd_valid = 0;
+ }
+}
+
+void qedf_scsi_completion(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req)
+{
+ u16 xid, rval;
+ struct fcoe_task_context *task_ctx;
+ struct scsi_cmnd *sc_cmd;
+ struct fcoe_cqe_rsp_info *fcp_rsp;
+ struct qedf_rport *fcport;
+ int refcount;
+ u16 scope, qualifier = 0;
+ u8 fw_residual_flag = 0;
+
+ if (!io_req)
+ return;
+ if (!cqe)
+ return;
+
+ xid = io_req->xid;
+ task_ctx = qedf_get_task_mem(&qedf->tasks, xid);
+ sc_cmd = io_req->sc_cmd;
+ fcp_rsp = &cqe->cqe_info.rsp_info;
+
+ if (!sc_cmd) {
+ QEDF_WARN(&(qedf->dbg_ctx), "sc_cmd is NULL!\n");
+ return;
+ }
+
+ if (!sc_cmd->SCp.ptr) {
+ QEDF_WARN(&(qedf->dbg_ctx), "SCp.ptr is NULL, returned in "
+ "another context.\n");
+ return;
+ }
+
+ if (!sc_cmd->request) {
+ QEDF_WARN(&(qedf->dbg_ctx), "sc_cmd->request is NULL, "
+ "sc_cmd=%p.\n", sc_cmd);
+ return;
+ }
+
+ if (!sc_cmd->request->special) {
+ QEDF_WARN(&(qedf->dbg_ctx), "request->special is NULL so "
+ "request not valid, sc_cmd=%p.\n", sc_cmd);
+ return;
+ }
+
+ if (!sc_cmd->request->q) {
+ QEDF_WARN(&(qedf->dbg_ctx), "request->q is NULL so request "
+ "is not valid, sc_cmd=%p.\n", sc_cmd);
+ return;
+ }
+
+ fcport = io_req->fcport;
+
+ qedf_parse_fcp_rsp(io_req, fcp_rsp);
+
+ qedf_unmap_sg_list(qedf, io_req);
+
+ /* Check for FCP transport error */
+ if (io_req->fcp_rsp_len > 3 && io_req->fcp_rsp_code) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "FCP I/O protocol failure xid=0x%x fcp_rsp_len=%d "
+ "fcp_rsp_code=%d.\n", io_req->xid, io_req->fcp_rsp_len,
+ io_req->fcp_rsp_code);
+ sc_cmd->result = DID_BUS_BUSY << 16;
+ goto out;
+ }
+
+ fw_residual_flag = GET_FIELD(cqe->cqe_info.rsp_info.fw_error_flags,
+ FCOE_CQE_RSP_INFO_FW_UNDERRUN);
+ if (fw_residual_flag) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "Firmware detected underrun: xid=0x%x fcp_rsp.flags=0x%02x "
+ "fcp_resid=%d fw_residual=0x%x.\n", io_req->xid,
+ fcp_rsp->rsp_flags.flags, io_req->fcp_resid,
+ cqe->cqe_info.rsp_info.fw_residual);
+
+ if (io_req->cdb_status == 0)
+ sc_cmd->result = (DID_ERROR << 16) | io_req->cdb_status;
+ else
+ sc_cmd->result = (DID_OK << 16) | io_req->cdb_status;
+
+ /* Abort the command since we did not get all the data */
+ init_completion(&io_req->abts_done);
+ rval = qedf_initiate_abts(io_req, true);
+ if (rval) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed to queue ABTS.\n");
+ sc_cmd->result = (DID_ERROR << 16) | io_req->cdb_status;
+ }
+
+ /*
+ * Set resid to the whole buffer length so we won't try to resue
+ * any previously data.
+ */
+ scsi_set_resid(sc_cmd, scsi_bufflen(sc_cmd));
+ goto out;
+ }
+
+ switch (io_req->fcp_status) {
+ case FC_GOOD:
+ if (io_req->cdb_status == 0) {
+ /* Good I/O completion */
+ sc_cmd->result = DID_OK << 16;
+ } else {
+ refcount = atomic_read(&io_req->refcount.refcount);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "%d:0:%d:%d xid=0x%0x op=0x%02x "
+ "lba=%02x%02x%02x%02x cdb_status=%d "
+ "fcp_resid=0x%x refcount=%d.\n",
+ qedf->lport->host->host_no, sc_cmd->device->id,
+ sc_cmd->device->lun, io_req->xid,
+ sc_cmd->cmnd[0], sc_cmd->cmnd[2], sc_cmd->cmnd[3],
+ sc_cmd->cmnd[4], sc_cmd->cmnd[5],
+ io_req->cdb_status, io_req->fcp_resid,
+ refcount);
+ sc_cmd->result = (DID_OK << 16) | io_req->cdb_status;
+
+ if (io_req->cdb_status == SAM_STAT_TASK_SET_FULL ||
+ io_req->cdb_status == SAM_STAT_BUSY) {
+ /*
+ * Check whether we need to set retry_delay at
+ * all based on retry_delay module parameter
+ * and the status qualifier.
+ */
+
+ /* Upper 2 bits */
+ scope = fcp_rsp->retry_delay_timer & 0xC000;
+ /* Lower 14 bits */
+ qualifier = fcp_rsp->retry_delay_timer & 0x3FFF;
+
+ if (qedf_retry_delay &&
+ scope > 0 && qualifier > 0 &&
+ qualifier <= 0x3FEF) {
+ /* Check we don't go over the max */
+ if (qualifier > QEDF_RETRY_DELAY_MAX)
+ qualifier =
+ QEDF_RETRY_DELAY_MAX;
+ fcport->retry_delay_timestamp =
+ jiffies + (qualifier * HZ / 10);
+ }
+ }
+ }
+ if (io_req->fcp_resid)
+ scsi_set_resid(sc_cmd, io_req->fcp_resid);
+ break;
+ default:
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO, "fcp_status=%d.\n",
+ io_req->fcp_status);
+ break;
+ }
+
+out:
+ if (qedf_io_tracing)
+ qedf_trace_io(fcport, io_req, QEDF_IO_TRACE_RSP);
+
+ io_req->sc_cmd = NULL;
+ sc_cmd->SCp.ptr = NULL;
+ sc_cmd->scsi_done(sc_cmd);
+ kref_put(&io_req->refcount, qedf_release_cmd);
+}
+
+/* Return a SCSI command in some other context besides a normal completion */
+void qedf_scsi_done(struct qedf_ctx *qedf, struct qedf_ioreq *io_req,
+ int result)
+{
+ u16 xid;
+ struct scsi_cmnd *sc_cmd;
+ int refcount;
+
+ if (!io_req)
+ return;
+
+ xid = io_req->xid;
+ sc_cmd = io_req->sc_cmd;
+
+ if (!sc_cmd) {
+ QEDF_WARN(&(qedf->dbg_ctx), "sc_cmd is NULL!\n");
+ return;
+ }
+
+ if (!sc_cmd->SCp.ptr) {
+ QEDF_WARN(&(qedf->dbg_ctx), "SCp.ptr is NULL, returned in "
+ "another context.\n");
+ return;
+ }
+
+ qedf_unmap_sg_list(qedf, io_req);
+
+ sc_cmd->result = result << 16;
+ refcount = atomic_read(&io_req->refcount.refcount);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO, "%d:0:%d:%d: Completing "
+ "sc_cmd=%p result=0x%08x op=0x%02x lba=0x%02x%02x%02x%02x, "
+ "allowed=%d retries=%d refcount=%d.\n",
+ qedf->lport->host->host_no, sc_cmd->device->id,
+ sc_cmd->device->lun, sc_cmd, sc_cmd->result, sc_cmd->cmnd[0],
+ sc_cmd->cmnd[2], sc_cmd->cmnd[3], sc_cmd->cmnd[4],
+ sc_cmd->cmnd[5], sc_cmd->allowed, sc_cmd->retries,
+ refcount);
+
+ /*
+ * Set resid to the whole buffer length so we won't try to resue any
+ * previously read data
+ */
+ scsi_set_resid(sc_cmd, scsi_bufflen(sc_cmd));
+
+ if (qedf_io_tracing)
+ qedf_trace_io(io_req->fcport, io_req, QEDF_IO_TRACE_RSP);
+
+ io_req->sc_cmd = NULL;
+ sc_cmd->SCp.ptr = NULL;
+ sc_cmd->scsi_done(sc_cmd);
+ kref_put(&io_req->refcount, qedf_release_cmd);
+}
+
+/*
+ * Handle warning type CQE completions. This is mainly used for REC timer
+ * popping.
+ */
+void qedf_process_warning_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req)
+{
+ int rval, i;
+ struct qedf_rport *fcport = io_req->fcport;
+ u64 err_warn_bit_map;
+ u8 err_warn = 0xff;
+
+ if (!cqe)
+ return;
+
+ QEDF_ERR(&(io_req->fcport->qedf->dbg_ctx), "Warning CQE, "
+ "xid=0x%x\n", io_req->xid);
+ QEDF_ERR(&(io_req->fcport->qedf->dbg_ctx),
+ "err_warn_bitmap=%08x:%08x\n",
+ le32_to_cpu(cqe->cqe_info.err_info.err_warn_bitmap_hi),
+ le32_to_cpu(cqe->cqe_info.err_info.err_warn_bitmap_lo));
+ QEDF_ERR(&(io_req->fcport->qedf->dbg_ctx), "tx_buff_off=%08x, "
+ "rx_buff_off=%08x, rx_id=%04x\n",
+ le32_to_cpu(cqe->cqe_info.err_info.tx_buf_off),
+ le32_to_cpu(cqe->cqe_info.err_info.rx_buf_off),
+ le32_to_cpu(cqe->cqe_info.err_info.rx_id));
+
+ /* Normalize the error bitmap value to an just an unsigned int */
+ err_warn_bit_map = (u64)
+ ((u64)cqe->cqe_info.err_info.err_warn_bitmap_hi << 32) |
+ (u64)cqe->cqe_info.err_info.err_warn_bitmap_lo;
+ for (i = 0; i < 64; i++) {
+ if (err_warn_bit_map & (u64)((u64)1 << i)) {
+ err_warn = i;
+ break;
+ }
+ }
+
+ /* Check if REC TOV expired if this is a tape device */
+ if (fcport->dev_type == QEDF_RPORT_TYPE_TAPE) {
+ if (err_warn ==
+ FCOE_WARNING_CODE_REC_TOV_TIMER_EXPIRATION) {
+ QEDF_ERR(&(qedf->dbg_ctx), "REC timer expired.\n");
+ if (!test_bit(QEDF_CMD_SRR_SENT, &io_req->flags)) {
+ io_req->rx_buf_off =
+ cqe->cqe_info.err_info.rx_buf_off;
+ io_req->tx_buf_off =
+ cqe->cqe_info.err_info.tx_buf_off;
+ io_req->rx_id = cqe->cqe_info.err_info.rx_id;
+ rval = qedf_send_rec(io_req);
+ /*
+ * We only want to abort the io_req if we
+ * can't queue the REC command as we want to
+ * keep the exchange open for recovery.
+ */
+ if (rval)
+ goto send_abort;
+ }
+ return;
+ }
+ }
+
+send_abort:
+ init_completion(&io_req->abts_done);
+ rval = qedf_initiate_abts(io_req, true);
+ if (rval)
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed to queue ABTS.\n");
+}
+
+/* Cleanup a command when we receive an error detection completion */
+void qedf_process_error_detect(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req)
+{
+ int rval;
+
+ if (!cqe)
+ return;
+
+ QEDF_ERR(&(io_req->fcport->qedf->dbg_ctx), "Error detection CQE, "
+ "xid=0x%x\n", io_req->xid);
+ QEDF_ERR(&(io_req->fcport->qedf->dbg_ctx),
+ "err_warn_bitmap=%08x:%08x\n",
+ le32_to_cpu(cqe->cqe_info.err_info.err_warn_bitmap_hi),
+ le32_to_cpu(cqe->cqe_info.err_info.err_warn_bitmap_lo));
+ QEDF_ERR(&(io_req->fcport->qedf->dbg_ctx), "tx_buff_off=%08x, "
+ "rx_buff_off=%08x, rx_id=%04x\n",
+ le32_to_cpu(cqe->cqe_info.err_info.tx_buf_off),
+ le32_to_cpu(cqe->cqe_info.err_info.rx_buf_off),
+ le32_to_cpu(cqe->cqe_info.err_info.rx_id));
+
+ if (qedf->stop_io_on_error) {
+ qedf_stop_all_io(qedf);
+ return;
+ }
+
+ init_completion(&io_req->abts_done);
+ rval = qedf_initiate_abts(io_req, true);
+ if (rval)
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed to queue ABTS.\n");
+}
+
+static void qedf_flush_els_req(struct qedf_ctx *qedf,
+ struct qedf_ioreq *els_req)
+{
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Flushing ELS request xid=0x%x refcount=%d.\n", els_req->xid,
+ atomic_read(&els_req->refcount.refcount));
+
+ /*
+ * Need to distinguish this from a timeout when calling the
+ * els_req->cb_func.
+ */
+ els_req->event = QEDF_IOREQ_EV_ELS_FLUSH;
+
+ /* Cancel the timer */
+ cancel_delayed_work_sync(&els_req->timeout_work);
+
+ /* Call callback function to complete command */
+ if (els_req->cb_func && els_req->cb_arg) {
+ els_req->cb_func(els_req->cb_arg);
+ els_req->cb_arg = NULL;
+ }
+
+ /* Release kref for original initiate_els */
+ kref_put(&els_req->refcount, qedf_release_cmd);
+}
+
+/* A value of -1 for lun is a wild card that means flush all
+ * active SCSI I/Os for the target.
+ */
+void qedf_flush_active_ios(struct qedf_rport *fcport, int lun)
+{
+ struct qedf_ioreq *io_req;
+ struct qedf_ctx *qedf;
+ struct qedf_cmd_mgr *cmd_mgr;
+ int i, rc;
+
+ if (!fcport)
+ return;
+
+ qedf = fcport->qedf;
+ cmd_mgr = qedf->cmd_mgr;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO, "Flush active i/o's.\n");
+
+ for (i = 0; i < FCOE_PARAMS_NUM_TASKS; i++) {
+ io_req = &cmd_mgr->cmds[i];
+
+ if (!io_req)
+ continue;
+ if (io_req->fcport != fcport)
+ continue;
+ if (io_req->cmd_type == QEDF_ELS) {
+ rc = kref_get_unless_zero(&io_req->refcount);
+ if (!rc) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "Could not get kref for io_req=0x%p.\n",
+ io_req);
+ continue;
+ }
+ qedf_flush_els_req(qedf, io_req);
+ /*
+ * Release the kref and go back to the top of the
+ * loop.
+ */
+ goto free_cmd;
+ }
+
+ if (!io_req->sc_cmd)
+ continue;
+ if (lun > 0) {
+ if (io_req->sc_cmd->device->lun !=
+ (u64)lun)
+ continue;
+ }
+
+ /*
+ * Use kref_get_unless_zero in the unlikely case the command
+ * we're about to flush was completed in the normal SCSI path
+ */
+ rc = kref_get_unless_zero(&io_req->refcount);
+ if (!rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not get kref for "
+ "io_req=0x%p\n", io_req);
+ continue;
+ }
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Cleanup xid=0x%x.\n", io_req->xid);
+
+ /* Cleanup task and return I/O mid-layer */
+ qedf_initiate_cleanup(io_req, true);
+
+free_cmd:
+ kref_put(&io_req->refcount, qedf_release_cmd);
+ }
+}
+
+/*
+ * Initiate a ABTS middle path command. Note that we don't have to initialize
+ * the task context for an ABTS task.
+ */
+int qedf_initiate_abts(struct qedf_ioreq *io_req, bool return_scsi_cmd_on_abts)
+{
+ struct fc_lport *lport;
+ struct qedf_rport *fcport = io_req->fcport;
+ struct fc_rport_priv *rdata = fcport->rdata;
+ struct qedf_ctx *qedf = fcport->qedf;
+ u16 xid;
+ u32 r_a_tov = 0;
+ int rc = 0;
+ unsigned long flags;
+
+ r_a_tov = rdata->r_a_tov;
+ lport = qedf->lport;
+
+ if (!test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "tgt not offloaded\n");
+ rc = 1;
+ goto abts_err;
+ }
+
+ if (lport->state != LPORT_ST_READY || !(lport->link_up)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "link is not ready\n");
+ rc = 1;
+ goto abts_err;
+ }
+
+ if (atomic_read(&qedf->link_down_tmo_valid) > 0) {
+ QEDF_ERR(&(qedf->dbg_ctx), "link_down_tmo active.\n");
+ rc = 1;
+ goto abts_err;
+ }
+
+ /* Ensure room on SQ */
+ if (!atomic_read(&fcport->free_sqes)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "No SQ entries available\n");
+ rc = 1;
+ goto abts_err;
+ }
+
+
+ kref_get(&io_req->refcount);
+
+ xid = io_req->xid;
+ qedf->control_requests++;
+ qedf->packet_aborts++;
+
+ /* Set the return CPU to be the same as the request one */
+ io_req->cpu = smp_processor_id();
+ qedf_inc_percpu_requests(io_req->cpu);
+
+ /* Set the command type to abort */
+ io_req->cmd_type = QEDF_ABTS;
+ io_req->return_scsi_cmd_on_abts = return_scsi_cmd_on_abts;
+
+ set_bit(QEDF_CMD_IN_ABORT, &io_req->flags);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_SCSI_TM, "ABTS io_req xid = "
+ "0x%x\n", xid);
+
+ qedf_cmd_timer_set(qedf, io_req, QEDF_ABORT_TIMEOUT * HZ);
+
+ spin_lock_irqsave(&fcport->rport_lock, flags);
+
+ /* Add ABTS to send queue */
+ qedf_add_to_sq(fcport, xid, 0, FCOE_TASK_TYPE_ABTS, 0);
+
+ /* Ring doorbell */
+ qedf_ring_doorbell(fcport);
+
+ spin_unlock_irqrestore(&fcport->rport_lock, flags);
+
+ return rc;
+abts_err:
+ /*
+ * If the ABTS task fails to queue then we need to cleanup the
+ * task at the firmware.
+ */
+ qedf_initiate_cleanup(io_req, return_scsi_cmd_on_abts);
+ return rc;
+}
+
+void qedf_process_abts_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req)
+{
+ uint32_t r_ctl;
+ uint16_t xid;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_SCSI_TM, "Entered with xid = "
+ "0x%x cmd_type = %d\n", io_req->xid, io_req->cmd_type);
+
+ cancel_delayed_work(&io_req->timeout_work);
+
+ xid = io_req->xid;
+ r_ctl = cqe->cqe_info.abts_info.r_ctl;
+
+ switch (r_ctl) {
+ case FC_RCTL_BA_ACC:
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_SCSI_TM,
+ "ABTS response - ACC Send RRQ after R_A_TOV\n");
+ io_req->event = QEDF_IOREQ_EV_ABORT_SUCCESS;
+ /*
+ * Dont release this cmd yet. It will be relesed
+ * after we get RRQ response
+ */
+ kref_get(&io_req->refcount);
+ queue_delayed_work(qedf->dpc_wq, &io_req->rrq_work,
+ msecs_to_jiffies(qedf->lport->r_a_tov));
+ break;
+ /* For error cases let the cleanup return the command */
+ case FC_RCTL_BA_RJT:
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_SCSI_TM,
+ "ABTS response - RJT\n");
+ io_req->event = QEDF_IOREQ_EV_ABORT_FAILED;
+ break;
+ default:
+ QEDF_ERR(&(qedf->dbg_ctx), "Unknown ABTS response\n");
+ break;
+ }
+
+ clear_bit(QEDF_CMD_IN_ABORT, &io_req->flags);
+
+ if (io_req->sc_cmd) {
+ if (io_req->return_scsi_cmd_on_abts)
+ qedf_scsi_done(qedf, io_req, DID_ERROR);
+ }
+
+ /* Notify eh_abort handler that ABTS is complete */
+ complete(&io_req->abts_done);
+
+ kref_put(&io_req->refcount, qedf_release_cmd);
+}
+
+int qedf_init_mp_req(struct qedf_ioreq *io_req)
+{
+ struct qedf_mp_req *mp_req;
+ struct fcoe_sge *mp_req_bd;
+ struct fcoe_sge *mp_resp_bd;
+ struct qedf_ctx *qedf = io_req->fcport->qedf;
+ dma_addr_t addr;
+ uint64_t sz;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_MP_REQ, "Entered.\n");
+
+ mp_req = (struct qedf_mp_req *)&(io_req->mp_req);
+ memset(mp_req, 0, sizeof(struct qedf_mp_req));
+
+ if (io_req->cmd_type != QEDF_ELS) {
+ mp_req->req_len = sizeof(struct fcp_cmnd);
+ io_req->data_xfer_len = mp_req->req_len;
+ } else
+ mp_req->req_len = io_req->data_xfer_len;
+
+ mp_req->req_buf = dma_alloc_coherent(&qedf->pdev->dev, QEDF_PAGE_SIZE,
+ &mp_req->req_buf_dma, GFP_KERNEL);
+ if (!mp_req->req_buf) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to alloc MP req buffer\n");
+ qedf_free_mp_resc(io_req);
+ return -ENOMEM;
+ }
+
+ mp_req->resp_buf = dma_alloc_coherent(&qedf->pdev->dev,
+ QEDF_PAGE_SIZE, &mp_req->resp_buf_dma, GFP_KERNEL);
+ if (!mp_req->resp_buf) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to alloc TM resp "
+ "buffer\n");
+ qedf_free_mp_resc(io_req);
+ return -ENOMEM;
+ }
+
+ /* Allocate and map mp_req_bd and mp_resp_bd */
+ sz = sizeof(struct fcoe_sge);
+ mp_req->mp_req_bd = dma_alloc_coherent(&qedf->pdev->dev, sz,
+ &mp_req->mp_req_bd_dma, GFP_KERNEL);
+ if (!mp_req->mp_req_bd) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to alloc MP req bd\n");
+ qedf_free_mp_resc(io_req);
+ return -ENOMEM;
+ }
+
+ mp_req->mp_resp_bd = dma_alloc_coherent(&qedf->pdev->dev, sz,
+ &mp_req->mp_resp_bd_dma, GFP_KERNEL);
+ if (!mp_req->mp_resp_bd) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to alloc MP resp bd\n");
+ qedf_free_mp_resc(io_req);
+ return -ENOMEM;
+ }
+
+ /* Fill bd table */
+ addr = mp_req->req_buf_dma;
+ mp_req_bd = mp_req->mp_req_bd;
+ mp_req_bd->sge_addr.lo = U64_LO(addr);
+ mp_req_bd->sge_addr.hi = U64_HI(addr);
+ mp_req_bd->size = QEDF_PAGE_SIZE;
+
+ /*
+ * MP buffer is either a task mgmt command or an ELS.
+ * So the assumption is that it consumes a single bd
+ * entry in the bd table
+ */
+ mp_resp_bd = mp_req->mp_resp_bd;
+ addr = mp_req->resp_buf_dma;
+ mp_resp_bd->sge_addr.lo = U64_LO(addr);
+ mp_resp_bd->sge_addr.hi = U64_HI(addr);
+ mp_resp_bd->size = QEDF_PAGE_SIZE;
+
+ return 0;
+}
+
+/*
+ * Last ditch effort to clear the port if it's stuck. Used only after a
+ * cleanup task times out.
+ */
+static void qedf_drain_request(struct qedf_ctx *qedf)
+{
+ if (test_bit(QEDF_DRAIN_ACTIVE, &qedf->flags)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "MCP drain already active.\n");
+ return;
+ }
+
+ /* Set bit to return all queuecommand requests as busy */
+ set_bit(QEDF_DRAIN_ACTIVE, &qedf->flags);
+
+ /* Call qed drain request for function. Should be synchronous */
+ qed_ops->common->drain(qedf->cdev);
+
+ /* Settle time for CQEs to be returned */
+ msleep(100);
+
+ /* Unplug and continue */
+ clear_bit(QEDF_DRAIN_ACTIVE, &qedf->flags);
+}
+
+/*
+ * Returns SUCCESS if the cleanup task does not timeout, otherwise return
+ * FAILURE.
+ */
+int qedf_initiate_cleanup(struct qedf_ioreq *io_req,
+ bool return_scsi_cmd_on_abts)
+{
+ struct qedf_rport *fcport;
+ struct qedf_ctx *qedf;
+ uint16_t xid;
+ struct fcoe_task_context *task;
+ int tmo = 0;
+ int rc = SUCCESS;
+ unsigned long flags;
+
+ fcport = io_req->fcport;
+ if (!fcport) {
+ QEDF_ERR(NULL, "fcport is NULL.\n");
+ return SUCCESS;
+ }
+
+ qedf = fcport->qedf;
+ if (!qedf) {
+ QEDF_ERR(NULL, "qedf is NULL.\n");
+ return SUCCESS;
+ }
+
+ if (!test_bit(QEDF_CMD_OUTSTANDING, &io_req->flags) ||
+ test_bit(QEDF_CMD_IN_CLEANUP, &io_req->flags)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "io_req xid=0x%x already in "
+ "cleanup processing or already completed.\n",
+ io_req->xid);
+ return SUCCESS;
+ }
+
+ /* Ensure room on SQ */
+ if (!atomic_read(&fcport->free_sqes)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "No SQ entries available\n");
+ return FAILED;
+ }
+
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO, "Entered xid=0x%x\n",
+ io_req->xid);
+
+ /* Cleanup cmds re-use the same TID as the original I/O */
+ xid = io_req->xid;
+ io_req->cmd_type = QEDF_CLEANUP;
+ io_req->return_scsi_cmd_on_abts = return_scsi_cmd_on_abts;
+
+ /* Set the return CPU to be the same as the request one */
+ io_req->cpu = smp_processor_id();
+ qedf_inc_percpu_requests(io_req->cpu);
+
+ set_bit(QEDF_CMD_IN_CLEANUP, &io_req->flags);
+
+ task = qedf_get_task_mem(&qedf->tasks, xid);
+
+ init_completion(&io_req->tm_done);
+
+ /* Obtain free SQ entry */
+ spin_lock_irqsave(&fcport->rport_lock, flags);
+ qedf_add_to_sq(fcport, xid, 0, FCOE_TASK_TYPE_EXCHANGE_CLEANUP, 0);
+
+ /* Ring doorbell */
+ qedf_ring_doorbell(fcport);
+ spin_unlock_irqrestore(&fcport->rport_lock, flags);
+
+ tmo = wait_for_completion_timeout(&io_req->tm_done,
+ QEDF_CLEANUP_TIMEOUT * HZ);
+
+ if (!tmo) {
+ rc = FAILED;
+ /* Timeout case */
+ QEDF_ERR(&(qedf->dbg_ctx), "Cleanup command timeout, "
+ "xid=%x.\n", io_req->xid);
+ clear_bit(QEDF_CMD_IN_CLEANUP, &io_req->flags);
+ /* Issue a drain request if cleanup task times out */
+ QEDF_ERR(&(qedf->dbg_ctx), "Issuing MCP drain request.\n");
+ qedf_drain_request(qedf);
+ }
+
+ if (io_req->sc_cmd) {
+ if (io_req->return_scsi_cmd_on_abts)
+ qedf_scsi_done(qedf, io_req, DID_ERROR);
+ }
+
+ if (rc == SUCCESS)
+ io_req->event = QEDF_IOREQ_EV_CLEANUP_SUCCESS;
+ else
+ io_req->event = QEDF_IOREQ_EV_CLEANUP_FAILED;
+
+ return rc;
+}
+
+void qedf_process_cleanup_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req)
+{
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO, "Entered xid = 0x%x\n",
+ io_req->xid);
+
+ clear_bit(QEDF_CMD_IN_CLEANUP, &io_req->flags);
+
+ /* Complete so we can finish cleaning up the I/O */
+ complete(&io_req->tm_done);
+}
+
+static int qedf_execute_tmf(struct qedf_rport *fcport, struct scsi_cmnd *sc_cmd,
+ uint8_t tm_flags)
+{
+ struct qedf_ioreq *io_req;
+ struct qedf_mp_req *tm_req;
+ struct fcoe_task_context *task;
+ struct fc_frame_header *fc_hdr;
+ struct fcp_cmnd *fcp_cmnd;
+ struct qedf_ctx *qedf = fcport->qedf;
+ int rc = 0;
+ uint16_t xid;
+ uint32_t sid, did;
+ int tmo = 0;
+ unsigned long flags;
+
+ if (!sc_cmd) {
+ QEDF_ERR(&(qedf->dbg_ctx), "invalid arg\n");
+ return FAILED;
+ }
+
+ if (!(test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags))) {
+ QEDF_ERR(&(qedf->dbg_ctx), "fcport not offloaded\n");
+ rc = FAILED;
+ return FAILED;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_SCSI_TM, "portid = 0x%x "
+ "tm_flags = %d\n", fcport->rdata->ids.port_id, tm_flags);
+
+ io_req = qedf_alloc_cmd(fcport, QEDF_TASK_MGMT_CMD);
+ if (!io_req) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed TMF");
+ rc = -EAGAIN;
+ goto reset_tmf_err;
+ }
+
+ /* Initialize rest of io_req fields */
+ io_req->sc_cmd = sc_cmd;
+ io_req->fcport = fcport;
+ io_req->cmd_type = QEDF_TASK_MGMT_CMD;
+
+ /* Set the return CPU to be the same as the request one */
+ io_req->cpu = smp_processor_id();
+ qedf_inc_percpu_requests(io_req->cpu);
+
+ tm_req = (struct qedf_mp_req *)&(io_req->mp_req);
+
+ rc = qedf_init_mp_req(io_req);
+ if (rc == FAILED) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Task mgmt MP request init "
+ "failed\n");
+ kref_put(&io_req->refcount, qedf_release_cmd);
+ goto reset_tmf_err;
+ }
+
+ /* Set TM flags */
+ io_req->io_req_flags = 0;
+ tm_req->tm_flags = tm_flags;
+
+ /* Default is to return a SCSI command when an error occurs */
+ io_req->return_scsi_cmd_on_abts = true;
+
+ /* Fill FCP_CMND */
+ qedf_build_fcp_cmnd(io_req, (struct fcp_cmnd *)tm_req->req_buf);
+ fcp_cmnd = (struct fcp_cmnd *)tm_req->req_buf;
+ memset(fcp_cmnd->fc_cdb, 0, FCP_CMND_LEN);
+ fcp_cmnd->fc_dl = 0;
+
+ /* Fill FC header */
+ fc_hdr = &(tm_req->req_fc_hdr);
+ sid = fcport->sid;
+ did = fcport->rdata->ids.port_id;
+ __fc_fill_fc_hdr(fc_hdr, FC_RCTL_DD_UNSOL_CMD, sid, did,
+ FC_TYPE_FCP, FC_FC_FIRST_SEQ | FC_FC_END_SEQ |
+ FC_FC_SEQ_INIT, 0);
+ /* Obtain exchange id */
+ xid = io_req->xid;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_SCSI_TM, "TMF io_req xid = "
+ "0x%x\n", xid);
+
+ /* Initialize task context for this IO request */
+ task = qedf_get_task_mem(&qedf->tasks, xid);
+ qedf_init_mp_task(io_req, task);
+
+ init_completion(&io_req->tm_done);
+
+ /* Obtain free SQ entry */
+ spin_lock_irqsave(&fcport->rport_lock, flags);
+ qedf_add_to_sq(fcport, xid, 0, FCOE_TASK_TYPE_MIDPATH, 0);
+
+ /* Ring doorbell */
+ qedf_ring_doorbell(fcport);
+ spin_unlock_irqrestore(&fcport->rport_lock, flags);
+
+ tmo = wait_for_completion_timeout(&io_req->tm_done,
+ QEDF_TM_TIMEOUT * HZ);
+
+ if (!tmo) {
+ rc = FAILED;
+ QEDF_ERR(&(qedf->dbg_ctx), "wait for tm_cmpl timeout!\n");
+ } else {
+ /* Check TMF response code */
+ if (io_req->fcp_rsp_code == 0)
+ rc = SUCCESS;
+ else
+ rc = FAILED;
+ }
+
+ if (tm_flags == FCP_TMF_LUN_RESET)
+ qedf_flush_active_ios(fcport, (int)sc_cmd->device->lun);
+ else
+ qedf_flush_active_ios(fcport, -1);
+
+ kref_put(&io_req->refcount, qedf_release_cmd);
+
+ if (rc != SUCCESS) {
+ QEDF_ERR(&(qedf->dbg_ctx), "task mgmt command failed...\n");
+ rc = FAILED;
+ } else {
+ QEDF_ERR(&(qedf->dbg_ctx), "task mgmt command success...\n");
+ rc = SUCCESS;
+ }
+reset_tmf_err:
+ return rc;
+}
+
+int qedf_initiate_tmf(struct scsi_cmnd *sc_cmd, u8 tm_flags)
+{
+ struct fc_rport *rport = starget_to_rport(scsi_target(sc_cmd->device));
+ struct fc_rport_libfc_priv *rp = rport->dd_data;
+ struct qedf_rport *fcport = (struct qedf_rport *)&rp[1];
+ struct qedf_ctx *qedf;
+ struct fc_lport *lport;
+ int rc = SUCCESS;
+ int rval;
+
+ rval = fc_remote_port_chkready(rport);
+
+ if (rval) {
+ QEDF_ERR(NULL, "device_reset rport not ready\n");
+ rc = FAILED;
+ goto tmf_err;
+ }
+
+ if (fcport == NULL) {
+ QEDF_ERR(NULL, "device_reset: rport is NULL\n");
+ rc = FAILED;
+ goto tmf_err;
+ }
+
+ qedf = fcport->qedf;
+ lport = qedf->lport;
+
+ if (test_bit(QEDF_UNLOADING, &qedf->flags)) {
+ rc = SUCCESS;
+ goto tmf_err;
+ }
+
+ if (lport->state != LPORT_ST_READY || !(lport->link_up)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "link is not ready\n");
+ rc = FAILED;
+ goto tmf_err;
+ }
+
+ rc = qedf_execute_tmf(fcport, sc_cmd, tm_flags);
+
+tmf_err:
+ return rc;
+}
+
+void qedf_process_tmf_compl(struct qedf_ctx *qedf, struct fcoe_cqe *cqe,
+ struct qedf_ioreq *io_req)
+{
+ struct fcoe_cqe_rsp_info *fcp_rsp;
+ struct fcoe_cqe_midpath_info *mp_info;
+
+
+ /* Get TMF response length from CQE */
+ mp_info = &cqe->cqe_info.midpath_info;
+ io_req->mp_req.resp_len = mp_info->data_placement_size;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_SCSI_TM,
+ "Response len is %d.\n", io_req->mp_req.resp_len);
+
+ fcp_rsp = &cqe->cqe_info.rsp_info;
+ qedf_parse_fcp_rsp(io_req, fcp_rsp);
+
+ io_req->sc_cmd = NULL;
+ complete(&io_req->tm_done);
+}
+
+void qedf_process_unsol_compl(struct qedf_ctx *qedf, uint16_t que_idx,
+ struct fcoe_cqe *cqe)
+{
+ unsigned long flags;
+ uint16_t tmp;
+ uint16_t pktlen = cqe->cqe_info.unsolic_info.pkt_len;
+ u32 payload_len, crc;
+ struct fc_frame_header *fh;
+ struct fc_frame *fp;
+ struct qedf_io_work *work;
+ struct qedf_percpu_iothread_s *iothread;
+ u32 bdq_idx;
+ void *bdq_addr;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_UNSOL,
+ "address.hi=%x address.lo=%x opaque_data.hi=%x "
+ "opaque_data.lo=%x bdq_prod_idx=%u len=%u.\n",
+ le32_to_cpu(cqe->cqe_info.unsolic_info.bd_info.address.hi),
+ le32_to_cpu(cqe->cqe_info.unsolic_info.bd_info.address.lo),
+ le32_to_cpu(cqe->cqe_info.unsolic_info.bd_info.opaque.hi),
+ le32_to_cpu(cqe->cqe_info.unsolic_info.bd_info.opaque.lo),
+ qedf->bdq_prod_idx, pktlen);
+
+ bdq_idx = le32_to_cpu(cqe->cqe_info.unsolic_info.bd_info.opaque.lo);
+ if (bdq_idx >= QEDF_BDQ_SIZE) {
+ QEDF_ERR(&(qedf->dbg_ctx), "bdq_idx is out of range %d.\n",
+ bdq_idx);
+ goto increment_prod;
+ }
+
+ bdq_addr = qedf->bdq[bdq_idx].buf_addr;
+ if (!bdq_addr) {
+ QEDF_ERR(&(qedf->dbg_ctx), "bdq_addr is NULL, dropping "
+ "unsolicited packet.\n");
+ goto increment_prod;
+ }
+
+ if (qedf_dump_frames) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_UNSOL,
+ "BDQ frame is at addr=%p.\n", bdq_addr);
+ print_hex_dump(KERN_WARNING, "bdq ", DUMP_PREFIX_OFFSET, 16, 1,
+ (void *)bdq_addr, pktlen, false);
+ }
+
+ /* Allocate frame */
+ payload_len = pktlen - sizeof(struct fc_frame_header);
+ fp = fc_frame_alloc(qedf->lport, payload_len);
+ if (!fp) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not allocate fp.\n");
+ goto increment_prod;
+ }
+
+ /* Copy data from BDQ buffer into fc_frame struct */
+ fh = (struct fc_frame_header *)fc_frame_header_get(fp);
+ memcpy(fh, (void *)bdq_addr, pktlen);
+
+ /* Initialize the frame so libfc sees it as a valid frame */
+ crc = fcoe_fc_crc(fp);
+ fc_frame_init(fp);
+ fr_dev(fp) = qedf->lport;
+ fr_sof(fp) = FC_SOF_I3;
+ fr_eof(fp) = FC_EOF_T;
+ fr_crc(fp) = cpu_to_le32(~crc);
+
+ /*
+ * We need to return the frame back up to libfc in a non-atomic
+ * context
+ */
+ work = mempool_alloc(qedf->io_mempool, GFP_ATOMIC);
+ if (!work) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not allocate "
+ "work for I/O completion.\n");
+ fc_frame_free(fp);
+ goto increment_prod;
+ }
+ memset(work, 0, sizeof(struct qedf_io_work));
+
+ INIT_LIST_HEAD(&work->list);
+
+ /* Copy contents of CQE for deferred processing */
+ memcpy(&work->cqe, cqe, sizeof(struct fcoe_cqe));
+
+ work->qedf = qedf;
+ work->fp = fp;
+
+ iothread = &per_cpu(qedf_percpu_iothreads,
+ smp_processor_id());
+ spin_lock_irqsave(&iothread->work_lock, flags);
+ list_add_tail(&work->list, &iothread->work_list);
+ spin_unlock_irqrestore(&iothread->work_lock, flags);
+ wake_up_process(iothread->iothread);
+
+increment_prod:
+ spin_lock_irqsave(&qedf->hba_lock, flags);
+
+ /* Increment producer to let f/w know we've handled the frame */
+ qedf->bdq_prod_idx++;
+
+ /* Producer index wraps at uint16_t boundary */
+ if (qedf->bdq_prod_idx == 0xffff)
+ qedf->bdq_prod_idx = 0;
+
+ writew(qedf->bdq_prod_idx, qedf->bdq_primary_prod);
+ tmp = readw(qedf->bdq_primary_prod);
+ writew(qedf->bdq_prod_idx, qedf->bdq_secondary_prod);
+ tmp = readw(qedf->bdq_secondary_prod);
+
+ spin_unlock_irqrestore(&qedf->hba_lock, flags);
+}
diff --git a/drivers/scsi/qedf/qedf_main.c b/drivers/scsi/qedf/qedf_main.c
new file mode 100644
index 0000000..8526879
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_main.c
@@ -0,0 +1,3483 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/device.h>
+#include <linux/highmem.h>
+#include <linux/crc32.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/kthread.h>
+#include <scsi/libfc.h>
+#include <scsi/scsi_host.h>
+#include <linux/if_ether.h>
+#include <linux/if_vlan.h>
+#include <linux/cpu.h>
+#include "qedf.h"
+
+const struct qed_fcoe_ops *qed_ops;
+
+static int qedf_probe(struct pci_dev *pdev, const struct pci_device_id *id);
+static void qedf_remove(struct pci_dev *pdev);
+
+extern struct qedf_debugfs_ops qedf_debugfs_ops;
+extern struct file_operations qedf_dbg_fops;
+
+/*
+ * Driver module parameters.
+ */
+static unsigned int qedf_dev_loss_tmo = 60;
+module_param_named(dev_loss_tmo, qedf_dev_loss_tmo, int, S_IRUGO);
+MODULE_PARM_DESC(dev_loss_tmo, " dev_loss_tmo setting for attached "
+ "remote ports (default 60)");
+
+uint qedf_debug = QEDF_LOG_INFO;
+module_param_named(debug, qedf_debug, uint, S_IRUGO);
+MODULE_PARM_DESC(qedf_debug, " Debug mask. Pass '1' to enable default debugging"
+ " mask");
+
+static uint qedf_fipvlan_retries = 30;
+module_param_named(fipvlan_retries, qedf_fipvlan_retries, int, S_IRUGO);
+MODULE_PARM_DESC(fipvlan_retries, " Number of FIP VLAN requests to attempt "
+ "before giving up (default 30)");
+
+static uint qedf_fallback_vlan = QEDF_FALLBACK_VLAN;
+module_param_named(fallback_vlan, qedf_fallback_vlan, int, S_IRUGO);
+MODULE_PARM_DESC(fallback_vlan, " VLAN ID to try if fip vlan request fails "
+ "(default 1002).");
+
+static uint qedf_default_prio = QEDF_DEFAULT_PRIO;
+module_param_named(default_prio, qedf_default_prio, int, S_IRUGO);
+MODULE_PARM_DESC(default_prio, " Default 802.1q priority for FIP and FCoE"
+ " traffic (default 3).");
+
+uint qedf_dump_frames;
+module_param_named(dump_frames, qedf_dump_frames, int, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(dump_frames, " Print the skb data of FIP and FCoE frames "
+ "(default off)");
+
+static uint qedf_queue_depth;
+module_param_named(queue_depth, qedf_queue_depth, int, S_IRUGO);
+MODULE_PARM_DESC(queue_depth, " Sets the queue depth for all LUNs discovered "
+ "by the qedf driver. Default is 0 (use OS default).");
+
+uint qedf_io_tracing;
+module_param_named(io_tracing, qedf_io_tracing, int, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(io_tracing, " Enable logging of SCSI requests/completions "
+ "into trace buffer. (default off).");
+
+static uint qedf_max_lun = MAX_FIBRE_LUNS;
+module_param_named(max_lun, qedf_max_lun, int, S_IRUGO);
+MODULE_PARM_DESC(max_lun, " Sets the maximum luns per target that the driver "
+ "supports. (default 0xffffffff)");
+
+uint qedf_link_down_tmo;
+module_param_named(link_down_tmo, qedf_link_down_tmo, int, S_IRUGO);
+MODULE_PARM_DESC(link_down_tmo, " Delays informing the fcoe transport that the "
+ "link is down by N seconds.");
+
+bool qedf_retry_delay;
+module_param_named(retry_delay, qedf_retry_delay, bool, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(retry_delay, " Enable/disable handling of FCP_RSP IU retry "
+ "delay handling (default off).");
+
+static uint qedf_dp_module;
+module_param_named(dp_module, qedf_dp_module, uint, S_IRUGO);
+MODULE_PARM_DESC(dp_module, " bit flags control for verbose printk passed "
+ "qed module during probe.");
+
+static uint qedf_dp_level;
+module_param_named(dp_level, qedf_dp_level, uint, S_IRUGO);
+MODULE_PARM_DESC(dp_level, " printk verbosity control passed to qed module "
+ "during probe (0-3: 0 more verbose).");
+
+
+DEFINE_PER_CPU(struct qedf_percpu_iothread_s, qedf_percpu_iothreads);
+
+static struct fcoe_percpu_s qedf_global;
+static DEFINE_SPINLOCK(qedf_global_lock);
+
+static struct kmem_cache *qedf_io_work_cache;
+
+void qedf_set_vlan_id(struct qedf_ctx *qedf, int vlan_id)
+{
+ qedf->vlan_id = vlan_id;
+ qedf->vlan_id |= qedf_default_prio << VLAN_PRIO_SHIFT;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Setting vlan_id=%04x "
+ "prio=%d.\n", vlan_id, qedf_default_prio);
+}
+
+/* Returns true if we have a valid vlan, false otherwise */
+static bool qedf_initiate_fipvlan_req(struct qedf_ctx *qedf)
+{
+ int rc;
+
+ if (atomic_read(&qedf->link_state) != QEDF_LINK_UP) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Link not up.\n");
+ return false;
+ }
+
+ while (qedf->fipvlan_retries--) {
+ if (qedf->vlan_id > 0)
+ return true;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Retry %d.\n", qedf->fipvlan_retries);
+ init_completion(&qedf->fipvlan_compl);
+ qedf_fcoe_send_vlan_req(qedf);
+ rc = wait_for_completion_timeout(&qedf->fipvlan_compl,
+ 1 * HZ);
+ if (rc > 0) {
+ fcoe_ctlr_link_up(&qedf->ctlr);
+ return true;
+ }
+ }
+
+ return false;
+}
+
+static void qedf_handle_link_update(struct work_struct *work)
+{
+ struct qedf_ctx *qedf =
+ container_of(work, struct qedf_ctx, link_update.work);
+ int rc;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Entered.\n");
+
+ if (atomic_read(&qedf->link_state) == QEDF_LINK_UP) {
+ rc = qedf_initiate_fipvlan_req(qedf);
+ if (rc)
+ return;
+ /*
+ * If we get here then we never received a repsonse to our
+ * fip vlan request so set the vlan_id to the default and
+ * tell FCoE that the link is up
+ */
+ QEDF_WARN(&(qedf->dbg_ctx), "Did not receive FIP VLAN "
+ "response, falling back to default VLAN %d.\n",
+ qedf_fallback_vlan);
+ qedf_set_vlan_id(qedf, QEDF_FALLBACK_VLAN);
+
+ /*
+ * Zero out data_src_addr so we'll update it with the new
+ * lport port_id
+ */
+ eth_zero_addr(qedf->data_src_addr);
+ fcoe_ctlr_link_up(&qedf->ctlr);
+ } else if (atomic_read(&qedf->link_state) == QEDF_LINK_DOWN) {
+ /*
+ * If we hit here and link_down_tmo_valid is still 1 it means
+ * that link_down_tmo timed out so set it to 0 to make sure any
+ * other readers have accurate state.
+ */
+ atomic_set(&qedf->link_down_tmo_valid, 0);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Calling fcoe_ctlr_link_down().\n");
+ fcoe_ctlr_link_down(&qedf->ctlr);
+ qedf_wait_for_upload(qedf);
+ /* Reset the number of FIP VLAN retries */
+ qedf->fipvlan_retries = qedf_fipvlan_retries;
+ }
+}
+
+static void qedf_flogi_resp(struct fc_seq *seq, struct fc_frame *fp,
+ void *arg)
+{
+ struct fc_exch *exch = fc_seq_exch(seq);
+ struct fc_lport *lport = exch->lp;
+ struct qedf_ctx *qedf = lport_priv(lport);
+
+ if (!qedf) {
+ QEDF_ERR(NULL, "qedf is NULL.\n");
+ return;
+ }
+
+ /*
+ * If ERR_PTR is set then don't try to stat anything as it will cause
+ * a crash when we access fp.
+ */
+ if (IS_ERR(fp)) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "fp has IS_ERR() set.\n");
+ goto skip_stat;
+ }
+
+ /* Log stats for FLOGI reject */
+ if (fc_frame_payload_op(fp) == ELS_LS_RJT)
+ qedf->flogi_failed++;
+
+ /* Complete flogi_compl so we can proceed to sending ADISCs */
+ complete(&qedf->flogi_compl);
+
+skip_stat:
+ /* Report response to libfc */
+ fc_lport_flogi_resp(seq, fp, lport);
+}
+
+static struct fc_seq *qedf_elsct_send(struct fc_lport *lport, u32 did,
+ struct fc_frame *fp, unsigned int op,
+ void (*resp)(struct fc_seq *,
+ struct fc_frame *,
+ void *),
+ void *arg, u32 timeout)
+{
+ struct qedf_ctx *qedf = lport_priv(lport);
+
+ /*
+ * Intercept FLOGI for statistic purposes. Note we use the resp
+ * callback to tell if this is really a flogi.
+ */
+ if (resp == fc_lport_flogi_resp) {
+ qedf->flogi_cnt++;
+ return fc_elsct_send(lport, did, fp, op, qedf_flogi_resp,
+ arg, timeout);
+ }
+
+ return fc_elsct_send(lport, did, fp, op, resp, arg, timeout);
+}
+
+int qedf_send_flogi(struct qedf_ctx *qedf)
+{
+ struct fc_lport *lport;
+ struct fc_frame *fp;
+
+ lport = qedf->lport;
+
+ if (!lport->tt.elsct_send)
+ return -EINVAL;
+
+ fp = fc_frame_alloc(lport, sizeof(struct fc_els_flogi));
+ if (!fp) {
+ QEDF_ERR(&(qedf->dbg_ctx), "fc_frame_alloc failed.\n");
+ return -ENOMEM;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_ELS,
+ "Sending FLOGI to reestablish session with switch.\n");
+ lport->tt.elsct_send(lport, FC_FID_FLOGI, fp,
+ ELS_FLOGI, qedf_flogi_resp, lport, lport->r_a_tov);
+
+ init_completion(&qedf->flogi_compl);
+
+ return 0;
+}
+
+struct qedf_tmp_rdata_item {
+ struct fc_rport_priv *rdata;
+ struct list_head list;
+};
+
+/*
+ * This function is called if link_down_tmo is in use. If we get a link up and
+ * link_down_tmo has not expired then use just FLOGI/ADISC to recover our
+ * sessions with targets. Otherwise, just call fcoe_ctlr_link_up().
+ */
+static void qedf_link_recovery(struct work_struct *work)
+{
+ struct qedf_ctx *qedf =
+ container_of(work, struct qedf_ctx, link_recovery.work);
+ struct qedf_rport *fcport;
+ struct fc_rport_priv *rdata;
+ struct qedf_tmp_rdata_item *rdata_item, *tmp_rdata_item;
+ bool rc;
+ int retries = 30;
+ int rval, i;
+ struct list_head rdata_login_list;
+
+ INIT_LIST_HEAD(&rdata_login_list);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Link down tmo did not expire.\n");
+
+ /*
+ * Essentially reset the fcoe_ctlr here without affecting the state
+ * of the libfc structs.
+ */
+ qedf->ctlr.state = FIP_ST_LINK_WAIT;
+ fcoe_ctlr_link_down(&qedf->ctlr);
+
+ /*
+ * Bring the link up before we send the fipvlan request so libfcoe
+ * can select a new fcf in parallel
+ */
+ fcoe_ctlr_link_up(&qedf->ctlr);
+
+ /* Since the link when down and up to verify which vlan we're on */
+ qedf->fipvlan_retries = qedf_fipvlan_retries;
+ rc = qedf_initiate_fipvlan_req(qedf);
+ if (!rc)
+ return;
+
+ /*
+ * We need to wait for an FCF to be selected due to the
+ * fcoe_ctlr_link_up other the FLOGI will be rejected.
+ */
+ while (retries > 0) {
+ if (qedf->ctlr.sel_fcf) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "FCF reselected, proceeding with FLOGI.\n");
+ break;
+ }
+ msleep(500);
+ retries--;
+ }
+
+ if (retries < 1) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Exhausted retries waiting for "
+ "FCF selection.\n");
+ return;
+ }
+
+ rval = qedf_send_flogi(qedf);
+ if (rval)
+ return;
+
+ /* Wait for FLOGI completion before proceeding with sending ADISCs */
+ i = wait_for_completion_timeout(&qedf->flogi_compl,
+ qedf->lport->r_a_tov);
+ if (i == 0) {
+ QEDF_ERR(&(qedf->dbg_ctx), "FLOGI timed out.\n");
+ return;
+ }
+
+ /*
+ * Call lport->tt.rport_login which will cause libfc to send an
+ * ADISC since the rport is in state ready.
+ */
+ rcu_read_lock();
+ list_for_each_entry_rcu(fcport, &qedf->fcports, peers) {
+ rdata = fcport->rdata;
+ if (rdata == NULL)
+ continue;
+ rdata_item = kzalloc(sizeof(struct qedf_tmp_rdata_item),
+ GFP_ATOMIC);
+ if (!rdata_item)
+ continue;
+ if (kref_get_unless_zero(&rdata->kref)) {
+ rdata_item->rdata = rdata;
+ list_add(&rdata_item->list, &rdata_login_list);
+ } else
+ kfree(rdata_item);
+ }
+ rcu_read_unlock();
+ /*
+ * Do the fc_rport_login outside of the rcu lock so we don't take a
+ * mutex in an atomic context.
+ */
+ list_for_each_entry_safe(rdata_item, tmp_rdata_item, &rdata_login_list,
+ list) {
+ list_del(&rdata_item->list);
+ fc_rport_login(rdata_item->rdata);
+ kref_put(&rdata->kref, fc_rport_destroy);
+ kfree(rdata_item);
+ }
+}
+
+static void qedf_update_link_speed(struct qedf_ctx *qedf,
+ struct qed_link_output *link)
+{
+ struct fc_lport *lport = qedf->lport;
+
+ lport->link_speed = FC_PORTSPEED_UNKNOWN;
+ lport->link_supported_speeds = FC_PORTSPEED_UNKNOWN;
+
+ /* Set fc_host link speed */
+ switch (link->speed) {
+ case 10000:
+ lport->link_speed = FC_PORTSPEED_10GBIT;
+ break;
+ case 25000:
+ lport->link_speed = FC_PORTSPEED_25GBIT;
+ break;
+ case 40000:
+ lport->link_speed = FC_PORTSPEED_40GBIT;
+ break;
+ case 50000:
+ lport->link_speed = FC_PORTSPEED_50GBIT;
+ break;
+ case 100000:
+ lport->link_speed = FC_PORTSPEED_100GBIT;
+ break;
+ default:
+ lport->link_speed = FC_PORTSPEED_UNKNOWN;
+ break;
+ }
+
+ /*
+ * Set supported link speed by querying the supported
+ * capabilities of the link.
+ */
+ if (link->supported_caps & SUPPORTED_10000baseKR_Full)
+ lport->link_supported_speeds |= FC_PORTSPEED_10GBIT;
+ if (link->supported_caps & SUPPORTED_25000baseKR_Full)
+ lport->link_supported_speeds |= FC_PORTSPEED_25GBIT;
+ if (link->supported_caps & SUPPORTED_40000baseLR4_Full)
+ lport->link_supported_speeds |= FC_PORTSPEED_40GBIT;
+ if (link->supported_caps & SUPPORTED_50000baseKR2_Full)
+ lport->link_supported_speeds |= FC_PORTSPEED_50GBIT;
+ if (link->supported_caps & SUPPORTED_100000baseKR4_Full)
+ lport->link_supported_speeds |= FC_PORTSPEED_100GBIT;
+ fc_host_supported_speeds(lport->host) = lport->link_supported_speeds;
+}
+
+static void qedf_link_update(void *dev, struct qed_link_output *link)
+{
+ struct qedf_ctx *qedf = (struct qedf_ctx *)dev;
+
+ if (link->link_up) {
+ QEDF_ERR(&(qedf->dbg_ctx), "LINK UP (%d GB/s).\n",
+ link->speed / 1000);
+
+ /* Cancel any pending link down work */
+ cancel_delayed_work(&qedf->link_update);
+
+ atomic_set(&qedf->link_state, QEDF_LINK_UP);
+ qedf_update_link_speed(qedf, link);
+
+ if (atomic_read(&qedf->dcbx) == QEDF_DCBX_DONE) {
+ QEDF_ERR(&(qedf->dbg_ctx), "DCBx done.\n");
+ if (atomic_read(&qedf->link_down_tmo_valid) > 0)
+ queue_delayed_work(qedf->link_update_wq,
+ &qedf->link_recovery, 0);
+ else
+ queue_delayed_work(qedf->link_update_wq,
+ &qedf->link_update, 0);
+ atomic_set(&qedf->link_down_tmo_valid, 0);
+ }
+
+ } else {
+ QEDF_ERR(&(qedf->dbg_ctx), "LINK DOWN.\n");
+
+ atomic_set(&qedf->link_state, QEDF_LINK_DOWN);
+ atomic_set(&qedf->dcbx, QEDF_DCBX_PENDING);
+ /*
+ * Flag that we're waiting for the link to come back up before
+ * informing the fcoe layer of the event.
+ */
+ if (qedf_link_down_tmo > 0) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Starting link down tmo.\n");
+ atomic_set(&qedf->link_down_tmo_valid, 1);
+ }
+ qedf->vlan_id = 0;
+ qedf_update_link_speed(qedf, link);
+ queue_delayed_work(qedf->link_update_wq, &qedf->link_update,
+ qedf_link_down_tmo * HZ);
+ }
+}
+
+
+static void qedf_dcbx_handler(void *dev, struct qed_dcbx_get *get, u32 mib_type)
+{
+ struct qedf_ctx *qedf = (struct qedf_ctx *)dev;
+
+ QEDF_ERR(&(qedf->dbg_ctx), "DCBx event valid=%d enabled=%d fcoe "
+ "prio=%d.\n", get->operational.valid, get->operational.enabled,
+ get->operational.app_prio.fcoe);
+
+ if (get->operational.enabled && get->operational.valid) {
+ /* If DCBX was already negotiated on link up then just exit */
+ if (atomic_read(&qedf->dcbx) == QEDF_DCBX_DONE) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "DCBX already set on link up.\n");
+ return;
+ }
+
+ atomic_set(&qedf->dcbx, QEDF_DCBX_DONE);
+
+ if (atomic_read(&qedf->link_state) == QEDF_LINK_UP) {
+ if (atomic_read(&qedf->link_down_tmo_valid) > 0)
+ queue_delayed_work(qedf->link_update_wq,
+ &qedf->link_recovery, 0);
+ else
+ queue_delayed_work(qedf->link_update_wq,
+ &qedf->link_update, 0);
+ atomic_set(&qedf->link_down_tmo_valid, 0);
+ }
+ }
+
+}
+
+static u32 qedf_get_login_failures(void *cookie)
+{
+ struct qedf_ctx *qedf;
+
+ qedf = (struct qedf_ctx *)cookie;
+ return qedf->flogi_failed;
+}
+
+static struct qed_fcoe_cb_ops qedf_cb_ops = {
+ {
+ .link_update = qedf_link_update,
+ .dcbx_aen = qedf_dcbx_handler,
+ }
+};
+
+/*
+ * Various transport templates.
+ */
+
+static struct scsi_transport_template *qedf_fc_transport_template;
+static struct scsi_transport_template *qedf_fc_vport_transport_template;
+
+/*
+ * SCSI EH handlers
+ */
+static int qedf_eh_abort(struct scsi_cmnd *sc_cmd)
+{
+ struct fc_rport *rport = starget_to_rport(scsi_target(sc_cmd->device));
+ struct fc_rport_libfc_priv *rp = rport->dd_data;
+ struct qedf_rport *fcport;
+ struct fc_lport *lport;
+ struct qedf_ctx *qedf;
+ struct qedf_ioreq *io_req;
+ int rc = FAILED;
+ int rval;
+
+ if (fc_remote_port_chkready(rport)) {
+ QEDF_ERR(NULL, "rport not ready\n");
+ goto out;
+ }
+
+ lport = shost_priv(sc_cmd->device->host);
+ qedf = (struct qedf_ctx *)lport_priv(lport);
+
+ if ((lport->state != LPORT_ST_READY) || !(lport->link_up)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "link not ready.\n");
+ goto out;
+ }
+
+ fcport = (struct qedf_rport *)&rp[1];
+
+ io_req = (struct qedf_ioreq *)sc_cmd->SCp.ptr;
+ if (!io_req) {
+ QEDF_ERR(&(qedf->dbg_ctx), "io_req is NULL.\n");
+ rc = SUCCESS;
+ goto out;
+ }
+
+ if (!test_bit(QEDF_CMD_OUTSTANDING, &io_req->flags) ||
+ test_bit(QEDF_CMD_IN_CLEANUP, &io_req->flags) ||
+ test_bit(QEDF_CMD_IN_ABORT, &io_req->flags)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "io_req xid=0x%x already in "
+ "cleanup or abort processing or already "
+ "completed.\n", io_req->xid);
+ rc = SUCCESS;
+ goto out;
+ }
+
+ QEDF_ERR(&(qedf->dbg_ctx), "Aborting io_req sc_cmd=%p xid=0x%x "
+ "fp_idx=%d.\n", sc_cmd, io_req->xid, io_req->fp_idx);
+
+ if (qedf->stop_io_on_error) {
+ qedf_stop_all_io(qedf);
+ rc = SUCCESS;
+ goto out;
+ }
+
+ init_completion(&io_req->abts_done);
+ rval = qedf_initiate_abts(io_req, true);
+ if (rval) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed to queue ABTS.\n");
+ goto out;
+ }
+
+ wait_for_completion(&io_req->abts_done);
+
+ if (io_req->event == QEDF_IOREQ_EV_ABORT_SUCCESS ||
+ io_req->event == QEDF_IOREQ_EV_ABORT_FAILED ||
+ io_req->event == QEDF_IOREQ_EV_CLEANUP_SUCCESS) {
+ /*
+ * If we get a reponse to the abort this is success from
+ * the perspective that all references to the command have
+ * been removed from the driver and firmware
+ */
+ rc = SUCCESS;
+ } else {
+ /* If the abort and cleanup failed then return a failure */
+ rc = FAILED;
+ }
+
+ if (rc == SUCCESS)
+ QEDF_ERR(&(qedf->dbg_ctx), "ABTS succeeded, xid=0x%x.\n",
+ io_req->xid);
+ else
+ QEDF_ERR(&(qedf->dbg_ctx), "ABTS failed, xid=0x%x.\n",
+ io_req->xid);
+
+out:
+ return rc;
+}
+
+static int qedf_eh_target_reset(struct scsi_cmnd *sc_cmd)
+{
+ QEDF_ERR(NULL, "TARGET RESET Issued...");
+ return qedf_initiate_tmf(sc_cmd, FCP_TMF_TGT_RESET);
+}
+
+static int qedf_eh_device_reset(struct scsi_cmnd *sc_cmd)
+{
+ QEDF_ERR(NULL, "LUN RESET Issued...\n");
+ return qedf_initiate_tmf(sc_cmd, FCP_TMF_LUN_RESET);
+}
+
+void qedf_wait_for_upload(struct qedf_ctx *qedf)
+{
+ while (1) {
+ if (atomic_read(&qedf->num_offloads))
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Waiting for all uploads to complete.\n");
+ else
+ break;
+ msleep(500);
+ }
+}
+
+/* Reset the host by gracefully logging out and then logging back in */
+static int qedf_eh_host_reset(struct scsi_cmnd *sc_cmd)
+{
+ struct fc_lport *lport;
+ struct qedf_ctx *qedf;
+
+ lport = shost_priv(sc_cmd->device->host);
+
+ if (lport->vport) {
+ QEDF_ERR(NULL, "Cannot issue host reset on NPIV port.\n");
+ return SUCCESS;
+ }
+
+ qedf = (struct qedf_ctx *)lport_priv(lport);
+
+ if (atomic_read(&qedf->link_state) == QEDF_LINK_DOWN ||
+ test_bit(QEDF_UNLOADING, &qedf->flags))
+ return FAILED;
+
+ QEDF_ERR(&(qedf->dbg_ctx), "HOST RESET Issued...");
+
+ /* For host reset, essentially do a soft link up/down */
+ atomic_set(&qedf->link_state, QEDF_LINK_DOWN);
+ atomic_set(&qedf->dcbx, QEDF_DCBX_PENDING);
+ queue_delayed_work(qedf->link_update_wq, &qedf->link_update,
+ 0);
+ qedf_wait_for_upload(qedf);
+ atomic_set(&qedf->link_state, QEDF_LINK_UP);
+ qedf->vlan_id = 0;
+ queue_delayed_work(qedf->link_update_wq, &qedf->link_update,
+ 0);
+
+ return SUCCESS;
+}
+
+static int qedf_slave_configure(struct scsi_device *sdev)
+{
+ if (qedf_queue_depth) {
+ scsi_change_queue_depth(sdev, qedf_queue_depth);
+ }
+
+ return 0;
+}
+
+static struct scsi_host_template qedf_host_template = {
+ .module = THIS_MODULE,
+ .name = QEDF_MODULE_NAME,
+ .this_id = -1,
+ .cmd_per_lun = 3,
+ .use_clustering = ENABLE_CLUSTERING,
+ .max_sectors = 0xffff,
+ .queuecommand = qedf_queuecommand,
+ .shost_attrs = qedf_host_attrs,
+ .eh_abort_handler = qedf_eh_abort,
+ .eh_device_reset_handler = qedf_eh_device_reset, /* lun reset */
+ .eh_target_reset_handler = qedf_eh_target_reset, /* target reset */
+ .eh_host_reset_handler = qedf_eh_host_reset,
+ .slave_configure = qedf_slave_configure,
+ .dma_boundary = QED_HW_DMA_BOUNDARY,
+ .sg_tablesize = QEDF_MAX_BDS_PER_CMD,
+ .can_queue = FCOE_PARAMS_NUM_TASKS,
+};
+
+static int qedf_get_paged_crc_eof(struct sk_buff *skb, int tlen)
+{
+ int rc;
+
+ spin_lock(&qedf_global_lock);
+ rc = fcoe_get_paged_crc_eof(skb, tlen, &qedf_global);
+ spin_unlock(&qedf_global_lock);
+
+ return rc;
+}
+
+static struct qedf_rport *qedf_fcport_lookup(struct qedf_ctx *qedf, u32 port_id)
+{
+ struct qedf_rport *fcport;
+ struct fc_rport_priv *rdata;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(fcport, &qedf->fcports, peers) {
+ rdata = fcport->rdata;
+ if (rdata == NULL)
+ continue;
+ if (rdata->ids.port_id == port_id) {
+ rcu_read_unlock();
+ return fcport;
+ }
+ }
+ rcu_read_unlock();
+
+ /* Return NULL to caller to let them know fcport was not found */
+ return NULL;
+}
+
+/* Transmits an ELS frame over an offloaded session */
+static int qedf_xmit_l2_frame(struct qedf_rport *fcport, struct fc_frame *fp)
+{
+ struct fc_frame_header *fh;
+ int rc = 0;
+
+ fh = fc_frame_header_get(fp);
+ if ((fh->fh_type == FC_TYPE_ELS) &&
+ (fh->fh_r_ctl == FC_RCTL_ELS_REQ)) {
+ switch (fc_frame_payload_op(fp)) {
+ case ELS_ADISC:
+ qedf_send_adisc(fcport, fp);
+ rc = 1;
+ break;
+ }
+ }
+
+ return rc;
+}
+
+/**
+ * qedf_xmit - qedf FCoE frame transmit function
+ *
+ */
+static int qedf_xmit(struct fc_lport *lport, struct fc_frame *fp)
+{
+ struct fc_lport *base_lport;
+ struct qedf_ctx *qedf;
+ struct ethhdr *eh;
+ struct fcoe_crc_eof *cp;
+ struct sk_buff *skb;
+ struct fc_frame_header *fh;
+ struct fcoe_hdr *hp;
+ u8 sof, eof;
+ u32 crc;
+ unsigned int hlen, tlen, elen;
+ int wlen;
+ struct fc_stats *stats;
+ struct fc_lport *tmp_lport;
+ struct fc_lport *vn_port = NULL;
+ struct qedf_rport *fcport;
+ int rc;
+ u16 vlan_tci = 0;
+
+ qedf = (struct qedf_ctx *)lport_priv(lport);
+
+ fh = fc_frame_header_get(fp);
+ skb = fp_skb(fp);
+
+ /* Filter out traffic to other NPIV ports on the same host */
+ if (lport->vport)
+ base_lport = shost_priv(vport_to_shost(lport->vport));
+ else
+ base_lport = lport;
+
+ /* Flag if the destination is the base port */
+ if (base_lport->port_id == ntoh24(fh->fh_d_id)) {
+ vn_port = base_lport;
+ } else {
+ /* Got through the list of vports attached to the base_lport
+ * and see if we have a match with the destination address.
+ */
+ list_for_each_entry(tmp_lport, &base_lport->vports, list) {
+ if (tmp_lport->port_id == ntoh24(fh->fh_d_id)) {
+ vn_port = tmp_lport;
+ break;
+ }
+ }
+ }
+ if (vn_port && ntoh24(fh->fh_d_id) != FC_FID_FLOGI) {
+ struct fc_rport_priv *rdata = NULL;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2,
+ "Dropping FCoE frame to %06x.\n", ntoh24(fh->fh_d_id));
+ kfree_skb(skb);
+ rdata = fc_rport_lookup(lport, ntoh24(fh->fh_d_id));
+ if (rdata)
+ rdata->retries = lport->max_rport_retry_count;
+ return -EINVAL;
+ }
+ /* End NPIV filtering */
+
+ if (!qedf->ctlr.sel_fcf) {
+ kfree_skb(skb);
+ return 0;
+ }
+
+ if (!test_bit(QEDF_LL2_STARTED, &qedf->flags)) {
+ QEDF_WARN(&(qedf->dbg_ctx), "LL2 not started\n");
+ kfree_skb(skb);
+ return 0;
+ }
+
+ if (atomic_read(&qedf->link_state) != QEDF_LINK_UP) {
+ QEDF_WARN(&(qedf->dbg_ctx), "qedf link down\n");
+ kfree_skb(skb);
+ return 0;
+ }
+
+ if (unlikely(fh->fh_r_ctl == FC_RCTL_ELS_REQ)) {
+ if (fcoe_ctlr_els_send(&qedf->ctlr, lport, skb))
+ return 0;
+ }
+
+ /* Check to see if this needs to be sent on an offloaded session */
+ fcport = qedf_fcport_lookup(qedf, ntoh24(fh->fh_d_id));
+
+ if (fcport && test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ rc = qedf_xmit_l2_frame(fcport, fp);
+ /*
+ * If the frame was successfully sent over the middle path
+ * then do not try to also send it over the LL2 path
+ */
+ if (rc)
+ return 0;
+ }
+
+ sof = fr_sof(fp);
+ eof = fr_eof(fp);
+
+ elen = sizeof(struct ethhdr);
+ hlen = sizeof(struct fcoe_hdr);
+ tlen = sizeof(struct fcoe_crc_eof);
+ wlen = (skb->len - tlen + sizeof(crc)) / FCOE_WORD_TO_BYTE;
+
+ skb->ip_summed = CHECKSUM_NONE;
+ crc = fcoe_fc_crc(fp);
+
+ /* copy port crc and eof to the skb buff */
+ if (skb_is_nonlinear(skb)) {
+ skb_frag_t *frag;
+
+ if (qedf_get_paged_crc_eof(skb, tlen)) {
+ kfree_skb(skb);
+ return -ENOMEM;
+ }
+ frag = &skb_shinfo(skb)->frags[skb_shinfo(skb)->nr_frags - 1];
+ cp = kmap_atomic(skb_frag_page(frag)) + frag->page_offset;
+ } else {
+ cp = (struct fcoe_crc_eof *)skb_put(skb, tlen);
+ }
+
+ memset(cp, 0, sizeof(*cp));
+ cp->fcoe_eof = eof;
+ cp->fcoe_crc32 = cpu_to_le32(~crc);
+ if (skb_is_nonlinear(skb)) {
+ kunmap_atomic(cp);
+ cp = NULL;
+ }
+
+
+ /* adjust skb network/transport offsets to match mac/fcoe/port */
+ skb_push(skb, elen + hlen);
+ skb_reset_mac_header(skb);
+ skb_reset_network_header(skb);
+ skb->mac_len = elen;
+ skb->protocol = htons(ETH_P_FCOE);
+
+ __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), qedf->vlan_id);
+
+ /* fill up mac and fcoe headers */
+ eh = eth_hdr(skb);
+ eh->h_proto = htons(ETH_P_FCOE);
+ if (qedf->ctlr.map_dest)
+ fc_fcoe_set_mac(eh->h_dest, fh->fh_d_id);
+ else
+ /* insert GW address */
+ ether_addr_copy(eh->h_dest, qedf->ctlr.dest_addr);
+
+ /* Set the source MAC address */
+ fc_fcoe_set_mac(eh->h_source, fh->fh_s_id);
+
+ hp = (struct fcoe_hdr *)(eh + 1);
+ memset(hp, 0, sizeof(*hp));
+ if (FC_FCOE_VER)
+ FC_FCOE_ENCAPS_VER(hp, FC_FCOE_VER);
+ hp->fcoe_sof = sof;
+
+ /*update tx stats */
+ stats = per_cpu_ptr(lport->stats, get_cpu());
+ stats->TxFrames++;
+ stats->TxWords += wlen;
+ put_cpu();
+
+ /* Get VLAN ID from skb for printing purposes */
+ __vlan_hwaccel_get_tag(skb, &vlan_tci);
+
+ /* send down to lld */
+ fr_dev(fp) = lport;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2, "FCoE frame send: "
+ "src=%06x dest=%06x r_ctl=%x type=%x vlan=%04x.\n",
+ ntoh24(fh->fh_s_id), ntoh24(fh->fh_d_id), fh->fh_r_ctl, fh->fh_type,
+ vlan_tci);
+ if (qedf_dump_frames)
+ print_hex_dump(KERN_WARNING, "fcoe: ", DUMP_PREFIX_OFFSET, 16,
+ 1, skb->data, skb->len, false);
+ qed_ops->ll2->start_xmit(qedf->cdev, skb);
+
+ return 0;
+}
+
+static int qedf_alloc_sq(struct qedf_ctx *qedf, struct qedf_rport *fcport)
+{
+ int rval = 0;
+ u32 *pbl;
+ dma_addr_t page;
+ int num_pages;
+
+ /* Calculate appropriate queue and PBL sizes */
+ fcport->sq_mem_size = SQ_NUM_ENTRIES * sizeof(struct fcoe_wqe);
+ fcport->sq_mem_size = ALIGN(fcport->sq_mem_size, QEDF_PAGE_SIZE);
+ fcport->sq_pbl_size = (fcport->sq_mem_size / QEDF_PAGE_SIZE) *
+ sizeof(void *);
+ fcport->sq_pbl_size = fcport->sq_pbl_size + QEDF_PAGE_SIZE;
+
+ fcport->sq = dma_alloc_coherent(&qedf->pdev->dev, fcport->sq_mem_size,
+ &fcport->sq_dma, GFP_KERNEL);
+ if (!fcport->sq) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not allocate send "
+ "queue.\n");
+ rval = 1;
+ goto out;
+ }
+ memset(fcport->sq, 0, fcport->sq_mem_size);
+
+ fcport->sq_pbl = dma_alloc_coherent(&qedf->pdev->dev,
+ fcport->sq_pbl_size, &fcport->sq_pbl_dma, GFP_KERNEL);
+ if (!fcport->sq_pbl) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not allocate send "
+ "queue PBL.\n");
+ rval = 1;
+ goto out_free_sq;
+ }
+ memset(fcport->sq_pbl, 0, fcport->sq_pbl_size);
+
+ /* Create PBL */
+ num_pages = fcport->sq_mem_size / QEDF_PAGE_SIZE;
+ page = fcport->sq_dma;
+ pbl = (u32 *)fcport->sq_pbl;
+
+ while (num_pages--) {
+ *pbl = U64_LO(page);
+ pbl++;
+ *pbl = U64_HI(page);
+ pbl++;
+ page += QEDF_PAGE_SIZE;
+ }
+
+ return rval;
+
+out_free_sq:
+ dma_free_coherent(&qedf->pdev->dev, fcport->sq_mem_size, fcport->sq,
+ fcport->sq_dma);
+out:
+ return rval;
+}
+
+static void qedf_free_sq(struct qedf_ctx *qedf, struct qedf_rport *fcport)
+{
+ if (fcport->sq_pbl)
+ dma_free_coherent(&qedf->pdev->dev, fcport->sq_pbl_size,
+ fcport->sq_pbl, fcport->sq_pbl_dma);
+ if (fcport->sq)
+ dma_free_coherent(&qedf->pdev->dev, fcport->sq_mem_size,
+ fcport->sq, fcport->sq_dma);
+}
+
+static int qedf_offload_connection(struct qedf_ctx *qedf,
+ struct qedf_rport *fcport)
+{
+ struct qed_fcoe_params_offload conn_info;
+ u32 port_id;
+ u8 lport_src_id[3];
+ int rval;
+ uint16_t total_sqe = (fcport->sq_mem_size / sizeof(struct fcoe_wqe));
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_CONN, "Offloading connection "
+ "portid=%06x.\n", fcport->rdata->ids.port_id);
+ rval = qed_ops->acquire_conn(qedf->cdev, &fcport->handle,
+ &fcport->fw_cid, &fcport->p_doorbell);
+ if (rval) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not acquire connection "
+ "for portid=%06x.\n", fcport->rdata->ids.port_id);
+ rval = 1; /* For some reason qed returns 0 on failure here */
+ goto out;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_CONN, "portid=%06x "
+ "fw_cid=%08x handle=%d.\n", fcport->rdata->ids.port_id,
+ fcport->fw_cid, fcport->handle);
+
+ memset(&conn_info, 0, sizeof(struct qed_fcoe_params_offload));
+
+ /* Fill in the offload connection info */
+ conn_info.sq_pbl_addr = fcport->sq_pbl_dma;
+
+ conn_info.sq_curr_page_addr = (dma_addr_t)(*(u64 *)fcport->sq_pbl);
+ conn_info.sq_next_page_addr =
+ (dma_addr_t)(*(u64 *)(fcport->sq_pbl + 8));
+
+ /* Need to use our FCoE MAC for the offload session */
+ port_id = fc_host_port_id(qedf->lport->host);
+ lport_src_id[2] = (port_id & 0x000000FF);
+ lport_src_id[1] = (port_id & 0x0000FF00) >> 8;
+ lport_src_id[0] = (port_id & 0x00FF0000) >> 16;
+ fc_fcoe_set_mac(conn_info.src_mac, lport_src_id);
+
+ ether_addr_copy(conn_info.dst_mac, qedf->ctlr.dest_addr);
+
+ conn_info.tx_max_fc_pay_len = fcport->rdata->maxframe_size;
+ conn_info.e_d_tov_timer_val = qedf->lport->e_d_tov / 20;
+ conn_info.rec_tov_timer_val = 3; /* I think this is what E3 was */
+ conn_info.rx_max_fc_pay_len = fcport->rdata->maxframe_size;
+
+ /* Set VLAN data */
+ conn_info.vlan_tag = qedf->vlan_id <<
+ FCOE_CONN_OFFLOAD_RAMROD_DATA_VLAN_ID_SHIFT;
+ conn_info.vlan_tag |=
+ qedf_default_prio << FCOE_CONN_OFFLOAD_RAMROD_DATA_PRIORITY_SHIFT;
+ conn_info.flags |= (FCOE_CONN_OFFLOAD_RAMROD_DATA_B_VLAN_FLAG_MASK <<
+ FCOE_CONN_OFFLOAD_RAMROD_DATA_B_VLAN_FLAG_SHIFT);
+
+ /* Set host port source id */
+ port_id = fc_host_port_id(qedf->lport->host);
+ fcport->sid = port_id;
+ conn_info.s_id.addr_hi = (port_id & 0x000000FF);
+ conn_info.s_id.addr_mid = (port_id & 0x0000FF00) >> 8;
+ conn_info.s_id.addr_lo = (port_id & 0x00FF0000) >> 16;
+
+ conn_info.max_conc_seqs_c3 = fcport->rdata->max_seq;
+
+ /* Set remote port destination id */
+ port_id = fcport->rdata->rport->port_id;
+ conn_info.d_id.addr_hi = (port_id & 0x000000FF);
+ conn_info.d_id.addr_mid = (port_id & 0x0000FF00) >> 8;
+ conn_info.d_id.addr_lo = (port_id & 0x00FF0000) >> 16;
+
+ conn_info.def_q_idx = 0; /* Default index for send queue? */
+
+ /* Set FC-TAPE specific flags if needed */
+ if (fcport->dev_type == QEDF_RPORT_TYPE_TAPE) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_CONN,
+ "Enable CONF, REC for portid=%06x.\n",
+ fcport->rdata->ids.port_id);
+ conn_info.flags |= 1 <<
+ FCOE_CONN_OFFLOAD_RAMROD_DATA_B_CONF_REQ_SHIFT;
+ conn_info.flags |=
+ ((fcport->rdata->sp_features & FC_SP_FT_SEQC) ? 1 : 0) <<
+ FCOE_CONN_OFFLOAD_RAMROD_DATA_B_REC_VALID_SHIFT;
+ }
+
+ rval = qed_ops->offload_conn(qedf->cdev, fcport->handle, &conn_info);
+ if (rval) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not offload connection "
+ "for portid=%06x.\n", fcport->rdata->ids.port_id);
+ goto out_free_conn;
+ } else
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_CONN, "Offload "
+ "succeeded portid=%06x total_sqe=%d.\n",
+ fcport->rdata->ids.port_id, total_sqe);
+
+ spin_lock_init(&fcport->rport_lock);
+ atomic_set(&fcport->free_sqes, total_sqe);
+ return 0;
+out_free_conn:
+ qed_ops->release_conn(qedf->cdev, fcport->handle);
+out:
+ return rval;
+}
+
+#define QEDF_TERM_BUFF_SIZE 10
+static void qedf_upload_connection(struct qedf_ctx *qedf,
+ struct qedf_rport *fcport)
+{
+ void *term_params;
+ dma_addr_t term_params_dma;
+
+ /* Term params needs to be a DMA coherent buffer as qed shared the
+ * physical DMA address with the firmware. The buffer may be used in
+ * the receive path so we may eventually have to move this.
+ */
+ term_params = dma_alloc_coherent(&qedf->pdev->dev, QEDF_TERM_BUFF_SIZE,
+ &term_params_dma, GFP_KERNEL);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_CONN, "Uploading connection "
+ "port_id=%06x.\n", fcport->rdata->ids.port_id);
+
+ qed_ops->destroy_conn(qedf->cdev, fcport->handle, term_params_dma);
+ qed_ops->release_conn(qedf->cdev, fcport->handle);
+
+ dma_free_coherent(&qedf->pdev->dev, QEDF_TERM_BUFF_SIZE, term_params,
+ term_params_dma);
+}
+
+static void qedf_cleanup_fcport(struct qedf_ctx *qedf,
+ struct qedf_rport *fcport)
+{
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_CONN, "Cleaning up portid=%06x.\n",
+ fcport->rdata->ids.port_id);
+
+ /* Flush any remaining i/o's before we upload the connection */
+ qedf_flush_active_ios(fcport, -1);
+
+ if (test_and_clear_bit(QEDF_RPORT_SESSION_READY, &fcport->flags))
+ qedf_upload_connection(qedf, fcport);
+ qedf_free_sq(qedf, fcport);
+ fcport->rdata = NULL;
+ fcport->qedf = NULL;
+}
+
+/**
+ * This event_callback is called after successful completion of libfc
+ * initiated target login. qedf can proceed with initiating the session
+ * establishment.
+ */
+static void qedf_rport_event_handler(struct fc_lport *lport,
+ struct fc_rport_priv *rdata,
+ enum fc_rport_event event)
+{
+ struct qedf_ctx *qedf = lport_priv(lport);
+ struct fc_rport *rport = rdata->rport;
+ struct fc_rport_libfc_priv *rp;
+ struct qedf_rport *fcport;
+ u32 port_id;
+ int rval;
+ unsigned long flags;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "event = %d, "
+ "port_id = 0x%x\n", event, rdata->ids.port_id);
+
+ switch (event) {
+ case RPORT_EV_READY:
+ if (!rport) {
+ QEDF_WARN(&(qedf->dbg_ctx), "rport is NULL.\n");
+ break;
+ }
+
+ rp = rport->dd_data;
+ fcport = (struct qedf_rport *)&rp[1];
+ fcport->qedf = qedf;
+
+ if (atomic_read(&qedf->num_offloads) >= QEDF_MAX_SESSIONS) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Not offloading "
+ "portid=0x%x as max number of offloaded sessions "
+ "reached.\n", rdata->ids.port_id);
+ return;
+ }
+
+ /*
+ * Don't try to offload the session again. Can happen when we
+ * get an ADISC
+ */
+ if (test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Session already "
+ "offloaded, portid=0x%x.\n",
+ rdata->ids.port_id);
+ return;
+ }
+
+ if (rport->port_id == FC_FID_DIR_SERV) {
+ /*
+ * qedf_rport structure doesn't exist for
+ * directory server.
+ * We should not come here, as lport will
+ * take care of fabric login
+ */
+ QEDF_WARN(&(qedf->dbg_ctx), "rport struct does not "
+ "exist for dir server port_id=%x\n",
+ rdata->ids.port_id);
+ break;
+ }
+
+ if (rdata->spp_type != FC_TYPE_FCP) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Not offlading since since spp type isn't FCP\n");
+ break;
+ }
+ if (!(rdata->ids.roles & FC_RPORT_ROLE_FCP_TARGET)) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Not FCP target so not offloading\n");
+ break;
+ }
+
+ fcport->rdata = rdata;
+ fcport->rport = rport;
+
+ rval = qedf_alloc_sq(qedf, fcport);
+ if (rval) {
+ qedf_cleanup_fcport(qedf, fcport);
+ break;
+ }
+
+ /* Set device type */
+ if (rdata->flags & FC_RP_FLAGS_RETRY &&
+ rdata->ids.roles & FC_RPORT_ROLE_FCP_TARGET &&
+ !(rdata->ids.roles & FC_RPORT_ROLE_FCP_INITIATOR)) {
+ fcport->dev_type = QEDF_RPORT_TYPE_TAPE;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "portid=%06x is a TAPE device.\n",
+ rdata->ids.port_id);
+ } else {
+ fcport->dev_type = QEDF_RPORT_TYPE_DISK;
+ }
+
+ rval = qedf_offload_connection(qedf, fcport);
+ if (rval) {
+ qedf_cleanup_fcport(qedf, fcport);
+ break;
+ }
+
+ /* Add fcport to list of qedf_ctx list of offloaded ports */
+ spin_lock_irqsave(&qedf->hba_lock, flags);
+ list_add_rcu(&fcport->peers, &qedf->fcports);
+ spin_unlock_irqrestore(&qedf->hba_lock, flags);
+
+ /*
+ * Set the session ready bit to let everyone know that this
+ * connection is ready for I/O
+ */
+ set_bit(QEDF_RPORT_SESSION_READY, &fcport->flags);
+ atomic_inc(&qedf->num_offloads);
+
+ break;
+ case RPORT_EV_LOGO:
+ case RPORT_EV_FAILED:
+ case RPORT_EV_STOP:
+ port_id = rdata->ids.port_id;
+ if (port_id == FC_FID_DIR_SERV)
+ break;
+
+ if (!rport) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "port_id=%x - rport notcreated Yet!!\n", port_id);
+ break;
+ }
+ rp = rport->dd_data;
+ /*
+ * Perform session upload. Note that rdata->peers is already
+ * removed from disc->rports list before we get this event.
+ */
+ fcport = (struct qedf_rport *)&rp[1];
+
+ /* Only free this fcport if it is offloaded already */
+ if (test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ set_bit(QEDF_RPORT_UPLOADING_CONNECTION, &fcport->flags);
+ qedf_cleanup_fcport(qedf, fcport);
+
+ /*
+ * Remove fcport to list of qedf_ctx list of offloaded
+ * ports
+ */
+ spin_lock_irqsave(&qedf->hba_lock, flags);
+ list_del_rcu(&fcport->peers);
+ spin_unlock_irqrestore(&qedf->hba_lock, flags);
+
+ clear_bit(QEDF_RPORT_UPLOADING_CONNECTION,
+ &fcport->flags);
+ atomic_dec(&qedf->num_offloads);
+ }
+
+ break;
+
+ case RPORT_EV_NONE:
+ break;
+ }
+}
+
+static void qedf_abort_io(struct fc_lport *lport)
+{
+ /* NO-OP but need to fill in the template */
+}
+
+static void qedf_fcp_cleanup(struct fc_lport *lport)
+{
+ /*
+ * NO-OP but need to fill in template to prevent a NULL
+ * function pointer dereference during link down. I/Os
+ * will be flushed when port is uploaded.
+ */
+}
+
+static struct libfc_function_template qedf_lport_template = {
+ .frame_send = qedf_xmit,
+ .fcp_abort_io = qedf_abort_io,
+ .fcp_cleanup = qedf_fcp_cleanup,
+ .rport_event_callback = qedf_rport_event_handler,
+ .elsct_send = qedf_elsct_send,
+};
+
+static void qedf_fcoe_ctlr_setup(struct qedf_ctx *qedf)
+{
+ fcoe_ctlr_init(&qedf->ctlr, FIP_ST_AUTO);
+
+ qedf->ctlr.send = qedf_fip_send;
+ qedf->ctlr.update_mac = qedf_update_src_mac;
+ qedf->ctlr.get_src_addr = qedf_get_src_mac;
+ ether_addr_copy(qedf->ctlr.ctl_src_addr, qedf->mac);
+}
+
+static int qedf_lport_setup(struct qedf_ctx *qedf)
+{
+ struct fc_lport *lport = qedf->lport;
+
+ lport->link_up = 0;
+ lport->max_retry_count = QEDF_FLOGI_RETRY_CNT;
+ lport->max_rport_retry_count = QEDF_RPORT_RETRY_CNT;
+ lport->service_params = (FCP_SPPF_INIT_FCN | FCP_SPPF_RD_XRDY_DIS |
+ FCP_SPPF_RETRY | FCP_SPPF_CONF_COMPL);
+ lport->boot_time = jiffies;
+ lport->e_d_tov = 2 * 1000;
+ lport->r_a_tov = 10 * 1000;
+
+ /* Set NPIV support */
+ lport->does_npiv = 1;
+ fc_host_max_npiv_vports(lport->host) = QEDF_MAX_NPIV;
+
+ fc_set_wwnn(lport, qedf->wwnn);
+ fc_set_wwpn(lport, qedf->wwpn);
+
+ fcoe_libfc_config(lport, &qedf->ctlr, &qedf_lport_template, 0);
+
+ /* Allocate the exchange manager */
+ fc_exch_mgr_alloc(lport, FC_CLASS_3, qedf->max_scsi_xid + 1,
+ qedf->max_els_xid, NULL);
+
+ if (fc_lport_init_stats(lport))
+ return -ENOMEM;
+
+ /* Finish lport config */
+ fc_lport_config(lport);
+
+ /* Set max frame size */
+ fc_set_mfs(lport, QEDF_MFS);
+ fc_host_maxframe_size(lport->host) = lport->mfs;
+
+ /* Set default dev_loss_tmo based on module parameter */
+ fc_host_dev_loss_tmo(lport->host) = qedf_dev_loss_tmo;
+
+ /* Set symbolic node name */
+ snprintf(fc_host_symbolic_name(lport->host), 256,
+ "QLogic %s v%s", QEDF_MODULE_NAME, QEDF_VERSION);
+
+ return 0;
+}
+
+/*
+ * NPIV functions
+ */
+
+static int qedf_vport_libfc_config(struct fc_vport *vport,
+ struct fc_lport *lport)
+{
+ lport->link_up = 0;
+ lport->qfull = 0;
+ lport->max_retry_count = QEDF_FLOGI_RETRY_CNT;
+ lport->max_rport_retry_count = QEDF_RPORT_RETRY_CNT;
+ lport->service_params = (FCP_SPPF_INIT_FCN | FCP_SPPF_RD_XRDY_DIS |
+ FCP_SPPF_RETRY | FCP_SPPF_CONF_COMPL);
+ lport->boot_time = jiffies;
+ lport->e_d_tov = 2 * 1000;
+ lport->r_a_tov = 10 * 1000;
+ lport->does_npiv = 1; /* Temporary until we add NPIV support */
+
+ /* Allocate stats for vport */
+ if (fc_lport_init_stats(lport))
+ return -ENOMEM;
+
+ /* Finish lport config */
+ fc_lport_config(lport);
+
+ /* offload related configuration */
+ lport->crc_offload = 0;
+ lport->seq_offload = 0;
+ lport->lro_enabled = 0;
+ lport->lro_xid = 0;
+ lport->lso_max = 0;
+
+ return 0;
+}
+
+static int qedf_vport_create(struct fc_vport *vport, bool disabled)
+{
+ struct Scsi_Host *shost = vport_to_shost(vport);
+ struct fc_lport *n_port = shost_priv(shost);
+ struct fc_lport *vn_port;
+ struct qedf_ctx *base_qedf = lport_priv(n_port);
+ struct qedf_ctx *vport_qedf;
+
+ char buf[32];
+ int rc = 0;
+
+ rc = fcoe_validate_vport_create(vport);
+ if (rc) {
+ fcoe_wwn_to_str(vport->port_name, buf, sizeof(buf));
+ QEDF_WARN(&(base_qedf->dbg_ctx), "Failed to create vport, "
+ "WWPN (0x%s) already exists.\n", buf);
+ goto err1;
+ }
+
+ if (atomic_read(&base_qedf->link_state) != QEDF_LINK_UP) {
+ QEDF_WARN(&(base_qedf->dbg_ctx), "Cannot create vport "
+ "because link is not up.\n");
+ rc = -EIO;
+ goto err1;
+ }
+
+ vn_port = libfc_vport_create(vport, sizeof(struct qedf_ctx));
+ if (!vn_port) {
+ QEDF_WARN(&(base_qedf->dbg_ctx), "Could not create lport "
+ "for vport.\n");
+ rc = -ENOMEM;
+ goto err1;
+ }
+
+ fcoe_wwn_to_str(vport->port_name, buf, sizeof(buf));
+ QEDF_ERR(&(base_qedf->dbg_ctx), "Creating NPIV port, WWPN=%s.\n",
+ buf);
+
+ /* Copy some fields from base_qedf */
+ vport_qedf = lport_priv(vn_port);
+ memcpy(vport_qedf, base_qedf, sizeof(struct qedf_ctx));
+
+ /* Set qedf data specific to this vport */
+ vport_qedf->lport = vn_port;
+ /* Use same hba_lock as base_qedf */
+ vport_qedf->hba_lock = base_qedf->hba_lock;
+ vport_qedf->pdev = base_qedf->pdev;
+ vport_qedf->cmd_mgr = base_qedf->cmd_mgr;
+ init_completion(&vport_qedf->flogi_compl);
+ INIT_LIST_HEAD(&vport_qedf->fcports);
+
+ rc = qedf_vport_libfc_config(vport, vn_port);
+ if (rc) {
+ QEDF_ERR(&(base_qedf->dbg_ctx), "Could not allocate memory "
+ "for lport stats.\n");
+ goto err2;
+ }
+
+ fc_set_wwnn(vn_port, vport->node_name);
+ fc_set_wwpn(vn_port, vport->port_name);
+ vport_qedf->wwnn = vn_port->wwnn;
+ vport_qedf->wwpn = vn_port->wwpn;
+
+ vn_port->host->transportt = qedf_fc_vport_transport_template;
+ vn_port->host->can_queue = QEDF_MAX_ELS_XID;
+ vn_port->host->max_lun = qedf_max_lun;
+ vn_port->host->sg_tablesize = QEDF_MAX_BDS_PER_CMD;
+ vn_port->host->max_cmd_len = QEDF_MAX_CDB_LEN;
+
+ rc = scsi_add_host(vn_port->host, &vport->dev);
+ if (rc) {
+ QEDF_WARN(&(base_qedf->dbg_ctx), "Error adding Scsi_Host.\n");
+ goto err2;
+ }
+
+ /* Set default dev_loss_tmo based on module parameter */
+ fc_host_dev_loss_tmo(vn_port->host) = qedf_dev_loss_tmo;
+
+ /* Init libfc stuffs */
+ memcpy(&vn_port->tt, &qedf_lport_template,
+ sizeof(qedf_lport_template));
+ fc_exch_init(vn_port);
+ fc_elsct_init(vn_port);
+ fc_lport_init(vn_port);
+ fc_disc_init(vn_port);
+ fc_disc_config(vn_port, vn_port);
+
+
+ /* Allocate the exchange manager */
+ shost = vport_to_shost(vport);
+ n_port = shost_priv(shost);
+ fc_exch_mgr_list_clone(n_port, vn_port);
+
+ /* Set max frame size */
+ fc_set_mfs(vn_port, QEDF_MFS);
+
+ fc_host_port_type(vn_port->host) = FC_PORTTYPE_UNKNOWN;
+
+ if (disabled) {
+ fc_vport_set_state(vport, FC_VPORT_DISABLED);
+ } else {
+ vn_port->boot_time = jiffies;
+ fc_fabric_login(vn_port);
+ fc_vport_setlink(vn_port);
+ }
+
+ QEDF_INFO(&(base_qedf->dbg_ctx), QEDF_LOG_NPIV, "vn_port=%p.\n",
+ vn_port);
+
+ /* Set up debug context for vport */
+ vport_qedf->dbg_ctx.host_no = vn_port->host->host_no;
+ vport_qedf->dbg_ctx.pdev = base_qedf->pdev;
+
+err2:
+ scsi_host_put(vn_port->host);
+err1:
+ return rc;
+}
+
+static int qedf_vport_destroy(struct fc_vport *vport)
+{
+ struct Scsi_Host *shost = vport_to_shost(vport);
+ struct fc_lport *n_port = shost_priv(shost);
+ struct fc_lport *vn_port = vport->dd_data;
+
+ mutex_lock(&n_port->lp_mutex);
+ list_del(&vn_port->list);
+ mutex_unlock(&n_port->lp_mutex);
+
+ fc_fabric_logoff(vn_port);
+ fc_lport_destroy(vn_port);
+
+ /* Detach from scsi-ml */
+ fc_remove_host(vn_port->host);
+ scsi_remove_host(vn_port->host);
+
+ /*
+ * Only try to release the exchange manager if the vn_port
+ * configuration is complete.
+ */
+ if (vn_port->state == LPORT_ST_READY)
+ fc_exch_mgr_free(vn_port);
+
+ /* Free memory used by statistical counters */
+ fc_lport_free_stats(vn_port);
+
+ /* Release Scsi_Host */
+ if (vn_port->host)
+ scsi_host_put(vn_port->host);
+
+ return 0;
+}
+
+static int qedf_vport_disable(struct fc_vport *vport, bool disable)
+{
+ struct fc_lport *lport = vport->dd_data;
+
+ if (disable) {
+ fc_vport_set_state(vport, FC_VPORT_DISABLED);
+ fc_fabric_logoff(lport);
+ } else {
+ lport->boot_time = jiffies;
+ fc_fabric_login(lport);
+ fc_vport_setlink(lport);
+ }
+ return 0;
+}
+
+/*
+ * During removal we need to wait for all the vports associated with a port
+ * to be destroyed so we avoid a race condition where libfc is still trying
+ * to reap vports while the driver remove function has already reaped the
+ * driver contexts associated with the physical port.
+ */
+static void qedf_wait_for_vport_destroy(struct qedf_ctx *qedf)
+{
+ struct fc_host_attrs *fc_host = shost_to_fc_host(qedf->lport->host);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_NPIV,
+ "Entered.\n");
+ while (fc_host->npiv_vports_inuse > 0) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_NPIV,
+ "Waiting for all vports to be reaped.\n");
+ msleep(1000);
+ }
+}
+
+/**
+ * qedf_fcoe_reset - Resets the fcoe
+ *
+ * @shost: shost the reset is from
+ *
+ * Returns: always 0
+ */
+static int qedf_fcoe_reset(struct Scsi_Host *shost)
+{
+ struct fc_lport *lport = shost_priv(shost);
+
+ fc_fabric_logoff(lport);
+ fc_fabric_login(lport);
+ return 0;
+}
+
+static struct fc_host_statistics *qedf_fc_get_host_stats(struct Scsi_Host
+ *shost)
+{
+ struct fc_host_statistics *qedf_stats;
+ struct fc_lport *lport = shost_priv(shost);
+ struct qedf_ctx *qedf = lport_priv(lport);
+ struct qed_fcoe_stats *fw_fcoe_stats;
+
+ qedf_stats = fc_get_host_stats(shost);
+
+ /* We don't collect offload stats for specific NPIV ports */
+ if (lport->vport)
+ goto out;
+
+ fw_fcoe_stats = kmalloc(sizeof(struct qed_fcoe_stats), GFP_KERNEL);
+ if (!fw_fcoe_stats) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not allocate memory for "
+ "fw_fcoe_stats.\n");
+ goto out;
+ }
+
+ /* Query firmware for offload stats */
+ qed_ops->get_stats(qedf->cdev, fw_fcoe_stats);
+
+ /*
+ * The expectation is that we add our offload stats to the stats
+ * being maintained by libfc each time the fc_get_host_status callback
+ * is invoked. The additions are not carried over for each call to
+ * the fc_get_host_stats callback.
+ */
+ qedf_stats->tx_frames += fw_fcoe_stats->fcoe_tx_data_pkt_cnt +
+ fw_fcoe_stats->fcoe_tx_xfer_pkt_cnt +
+ fw_fcoe_stats->fcoe_tx_other_pkt_cnt;
+ qedf_stats->rx_frames += fw_fcoe_stats->fcoe_rx_data_pkt_cnt +
+ fw_fcoe_stats->fcoe_rx_xfer_pkt_cnt +
+ fw_fcoe_stats->fcoe_rx_other_pkt_cnt;
+ qedf_stats->fcp_input_megabytes += fw_fcoe_stats->fcoe_rx_byte_cnt /
+ 1000000;
+ qedf_stats->fcp_output_megabytes += fw_fcoe_stats->fcoe_tx_byte_cnt /
+ 1000000;
+ qedf_stats->rx_words += fw_fcoe_stats->fcoe_rx_byte_cnt / 4;
+ qedf_stats->tx_words += fw_fcoe_stats->fcoe_tx_byte_cnt / 4;
+ qedf_stats->invalid_crc_count +=
+ fw_fcoe_stats->fcoe_silent_drop_pkt_crc_error_cnt;
+ qedf_stats->dumped_frames =
+ fw_fcoe_stats->fcoe_silent_drop_total_pkt_cnt;
+ qedf_stats->error_frames +=
+ fw_fcoe_stats->fcoe_silent_drop_total_pkt_cnt;
+ qedf_stats->fcp_input_requests += qedf->input_requests;
+ qedf_stats->fcp_output_requests += qedf->output_requests;
+ qedf_stats->fcp_control_requests += qedf->control_requests;
+ qedf_stats->fcp_packet_aborts += qedf->packet_aborts;
+ qedf_stats->fcp_frame_alloc_failures += qedf->alloc_failures;
+
+ kfree(fw_fcoe_stats);
+out:
+ return qedf_stats;
+}
+
+static struct fc_function_template qedf_fc_transport_fn = {
+ .show_host_node_name = 1,
+ .show_host_port_name = 1,
+ .show_host_supported_classes = 1,
+ .show_host_supported_fc4s = 1,
+ .show_host_active_fc4s = 1,
+ .show_host_maxframe_size = 1,
+
+ .show_host_port_id = 1,
+ .show_host_supported_speeds = 1,
+ .get_host_speed = fc_get_host_speed,
+ .show_host_speed = 1,
+ .show_host_port_type = 1,
+ .get_host_port_state = fc_get_host_port_state,
+ .show_host_port_state = 1,
+ .show_host_symbolic_name = 1,
+
+ /*
+ * Tell FC transport to allocate enough space to store the backpointer
+ * for the associate qedf_rport struct.
+ */
+ .dd_fcrport_size = (sizeof(struct fc_rport_libfc_priv) +
+ sizeof(struct qedf_rport)),
+ .show_rport_maxframe_size = 1,
+ .show_rport_supported_classes = 1,
+ .show_host_fabric_name = 1,
+ .show_starget_node_name = 1,
+ .show_starget_port_name = 1,
+ .show_starget_port_id = 1,
+ .set_rport_dev_loss_tmo = fc_set_rport_loss_tmo,
+ .show_rport_dev_loss_tmo = 1,
+ .get_fc_host_stats = qedf_fc_get_host_stats,
+ .issue_fc_host_lip = qedf_fcoe_reset,
+ .vport_create = qedf_vport_create,
+ .vport_delete = qedf_vport_destroy,
+ .vport_disable = qedf_vport_disable,
+ .bsg_request = fc_lport_bsg_request,
+};
+
+static struct fc_function_template qedf_fc_vport_transport_fn = {
+ .show_host_node_name = 1,
+ .show_host_port_name = 1,
+ .show_host_supported_classes = 1,
+ .show_host_supported_fc4s = 1,
+ .show_host_active_fc4s = 1,
+ .show_host_maxframe_size = 1,
+ .show_host_port_id = 1,
+ .show_host_supported_speeds = 1,
+ .get_host_speed = fc_get_host_speed,
+ .show_host_speed = 1,
+ .show_host_port_type = 1,
+ .get_host_port_state = fc_get_host_port_state,
+ .show_host_port_state = 1,
+ .show_host_symbolic_name = 1,
+ .dd_fcrport_size = (sizeof(struct fc_rport_libfc_priv) +
+ sizeof(struct qedf_rport)),
+ .show_rport_maxframe_size = 1,
+ .show_rport_supported_classes = 1,
+ .show_host_fabric_name = 1,
+ .show_starget_node_name = 1,
+ .show_starget_port_name = 1,
+ .show_starget_port_id = 1,
+ .set_rport_dev_loss_tmo = fc_set_rport_loss_tmo,
+ .show_rport_dev_loss_tmo = 1,
+ .get_fc_host_stats = fc_get_host_stats,
+ .issue_fc_host_lip = qedf_fcoe_reset,
+ .bsg_request = fc_lport_bsg_request,
+};
+
+static bool qedf_fp_has_work(struct qedf_fastpath *fp)
+{
+ struct qedf_ctx *qedf = fp->qedf;
+ struct global_queue *que;
+ struct qed_sb_info *sb_info = fp->sb_info;
+ struct status_block *sb = sb_info->sb_virt;
+ u16 prod_idx;
+
+ /* Get the pointer to the global CQ this completion is on */
+ que = qedf->global_queues[fp->sb_id];
+
+ /* Be sure all responses have been written to PI */
+ rmb();
+
+ /* Get the current firmware producer index */
+ prod_idx = sb->pi_array[QEDF_FCOE_PARAMS_GL_RQ_PI];
+
+ return (que->cq_prod_idx != prod_idx);
+}
+
+/*
+ * Interrupt handler code.
+ */
+
+/* Process completion queue and copy CQE contents for deferred processesing
+ *
+ * Return true if we should wake the I/O thread, false if not.
+ */
+static bool qedf_process_completions(struct qedf_fastpath *fp)
+{
+ struct qedf_ctx *qedf = fp->qedf;
+ struct qed_sb_info *sb_info = fp->sb_info;
+ struct status_block *sb = sb_info->sb_virt;
+ struct global_queue *que;
+ u16 prod_idx;
+ struct fcoe_cqe *cqe;
+ struct qedf_io_work *work;
+ unsigned long flags;
+ int num_handled = 0;
+ unsigned int cpu;
+ struct qedf_ioreq *io_req = NULL;
+ struct qedf_percpu_iothread_s *iothread;
+ u16 xid;
+ u16 new_cqes;
+ u32 comp_type;
+
+ /* Get the current firmware producer index */
+ prod_idx = sb->pi_array[QEDF_FCOE_PARAMS_GL_RQ_PI];
+
+ /* Get the pointer to the global CQ this completion is on */
+ que = qedf->global_queues[fp->sb_id];
+
+ /* Calculate the amount of new elements since last processing */
+ new_cqes = (prod_idx >= que->cq_prod_idx) ?
+ (prod_idx - que->cq_prod_idx) :
+ 0x10000 - que->cq_prod_idx + prod_idx;
+
+ /* Save producer index */
+ que->cq_prod_idx = prod_idx;
+
+ while (new_cqes) {
+ fp->completions++;
+ num_handled++;
+ cqe = &que->cq[que->cq_cons_idx];
+
+ comp_type = (cqe->cqe_data >> FCOE_CQE_CQE_TYPE_SHIFT) &
+ FCOE_CQE_CQE_TYPE_MASK;
+
+ /*
+ * Process unsolicited CQEs directly in the interrupt handler
+ * sine we need the fastpath ID
+ */
+ if (comp_type == FCOE_UNSOLIC_CQE_TYPE) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_UNSOL,
+ "Unsolicated CQE.\n");
+ qedf_process_unsol_compl(qedf, fp->sb_id, cqe);
+ /*
+ * Don't add a work list item. Increment consumer
+ * consumer index and move on.
+ */
+ goto inc_idx;
+ }
+
+ xid = cqe->cqe_data & FCOE_CQE_TASK_ID_MASK;
+ io_req = &qedf->cmd_mgr->cmds[xid];
+
+ /*
+ * Figure out which percpu thread we should queue this I/O
+ * on.
+ */
+ if (!io_req)
+ /* If there is not io_req assocated with this CQE
+ * just queue it on CPU 0
+ */
+ cpu = 0;
+ else {
+ cpu = io_req->cpu;
+ io_req->int_cpu = smp_processor_id();
+ }
+
+ work = mempool_alloc(qedf->io_mempool, GFP_ATOMIC);
+ if (!work) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not allocate "
+ "work for I/O completion.\n");
+ continue;
+ }
+ memset(work, 0, sizeof(struct qedf_io_work));
+
+ INIT_LIST_HEAD(&work->list);
+
+ /* Copy contents of CQE for deferred processing */
+ memcpy(&work->cqe, cqe, sizeof(struct fcoe_cqe));
+
+ work->qedf = fp->qedf;
+ work->fp = NULL; /* Only used for unsolicited frames */
+
+ iothread = &per_cpu(qedf_percpu_iothreads, cpu);
+ spin_lock_irqsave(&iothread->work_lock, flags);
+ list_add_tail(&work->list, &iothread->work_list);
+ spin_unlock_irqrestore(&iothread->work_lock, flags);
+ wake_up_process(iothread->iothread);
+
+inc_idx:
+ que->cq_cons_idx++;
+ if (que->cq_cons_idx == fp->cq_num_entries)
+ que->cq_cons_idx = 0;
+ new_cqes--;
+ }
+
+ return true;
+}
+
+
+/* MSI-X fastpath handler code */
+static irqreturn_t qedf_msix_handler(int irq, void *dev_id)
+{
+ struct qedf_fastpath *fp = dev_id;
+
+ if (!fp) {
+ QEDF_ERR(NULL, "fp is null.\n");
+ return IRQ_HANDLED;
+ }
+ if (!fp->sb_info) {
+ QEDF_ERR(NULL, "fp->sb_info in null.");
+ return IRQ_HANDLED;
+ }
+
+ /*
+ * Disable interrupts for this status block while we process new
+ * completions
+ */
+ qed_sb_ack(fp->sb_info, IGU_INT_DISABLE, 0 /*do not update*/);
+
+ while (1) {
+ qedf_process_completions(fp);
+
+ if (qedf_fp_has_work(fp) == 0) {
+ /* Update the sb information */
+ qed_sb_update_sb_idx(fp->sb_info);
+
+ /* Check for more work */
+ rmb();
+
+ if (qedf_fp_has_work(fp) == 0) {
+ /* Re-enable interrupts */
+ qed_sb_ack(fp->sb_info, IGU_INT_ENABLE, 1);
+ return IRQ_HANDLED;
+ }
+ }
+ }
+
+ /* Do we ever want to break out of above loop? */
+ return IRQ_HANDLED;
+}
+
+/* simd handler for MSI/INTa */
+static void qedf_simd_int_handler(void *cookie)
+{
+ /* Cookie is qedf_ctx struct */
+ struct qedf_ctx *qedf = (struct qedf_ctx *)cookie;
+
+ QEDF_WARN(&(qedf->dbg_ctx), "qedf=%p.\n", qedf);
+}
+
+#define QEDF_SIMD_HANDLER_NUM 0
+static void qedf_sync_free_irqs(struct qedf_ctx *qedf)
+{
+ int i;
+
+ if (qedf->int_info.msix_cnt) {
+ for (i = 0; i < qedf->int_info.used_cnt; i++) {
+ synchronize_irq(qedf->int_info.msix[i].vector);
+ irq_set_affinity_hint(qedf->int_info.msix[i].vector,
+ NULL);
+ irq_set_affinity_notifier(qedf->int_info.msix[i].vector,
+ NULL);
+ free_irq(qedf->int_info.msix[i].vector,
+ &qedf->fp_array[i]);
+ }
+ } else
+ qed_ops->common->simd_handler_clean(qedf->cdev,
+ QEDF_SIMD_HANDLER_NUM);
+
+ qedf->int_info.used_cnt = 0;
+ qed_ops->common->set_fp_int(qedf->cdev, 0);
+}
+
+static int qedf_request_msix_irq(struct qedf_ctx *qedf)
+{
+ int i, rc, cpu;
+
+ cpu = cpumask_first(cpu_online_mask);
+ for (i = 0; i < qedf->num_queues; i++) {
+ rc = request_irq(qedf->int_info.msix[i].vector,
+ qedf_msix_handler, 0, "qedf", &qedf->fp_array[i]);
+
+ if (rc) {
+ QEDF_WARN(&(qedf->dbg_ctx), "request_irq failed.\n");
+ qedf_sync_free_irqs(qedf);
+ return rc;
+ }
+
+ qedf->int_info.used_cnt++;
+ rc = irq_set_affinity_hint(qedf->int_info.msix[i].vector,
+ get_cpu_mask(cpu));
+ cpu = cpumask_next(cpu, cpu_online_mask);
+ }
+
+ return 0;
+}
+
+static int qedf_setup_int(struct qedf_ctx *qedf)
+{
+ int rc = 0;
+
+ /*
+ * Learn interrupt configuration
+ */
+ rc = qed_ops->common->set_fp_int(qedf->cdev, num_online_cpus());
+
+ rc = qed_ops->common->get_fp_int(qedf->cdev, &qedf->int_info);
+ if (rc)
+ return 0;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Number of msix_cnt = "
+ "0x%x num of cpus = 0x%x\n", qedf->int_info.msix_cnt,
+ num_online_cpus());
+
+ if (qedf->int_info.msix_cnt)
+ return qedf_request_msix_irq(qedf);
+
+ qed_ops->common->simd_handler_config(qedf->cdev, &qedf,
+ QEDF_SIMD_HANDLER_NUM, qedf_simd_int_handler);
+ qedf->int_info.used_cnt = 1;
+
+ return 0;
+}
+
+/* Main function for libfc frame reception */
+static void qedf_recv_frame(struct qedf_ctx *qedf,
+ struct sk_buff *skb)
+{
+ u32 fr_len;
+ struct fc_lport *lport;
+ struct fc_frame_header *fh;
+ struct fcoe_crc_eof crc_eof;
+ struct fc_frame *fp;
+ u8 *mac = NULL;
+ u8 *dest_mac = NULL;
+ struct fcoe_hdr *hp;
+ struct qedf_rport *fcport;
+
+ lport = qedf->lport;
+ if (lport == NULL || lport->state == LPORT_ST_DISABLED) {
+ QEDF_WARN(NULL, "Invalid lport struct or lport disabled.\n");
+ kfree_skb(skb);
+ return;
+ }
+
+ if (skb_is_nonlinear(skb))
+ skb_linearize(skb);
+ mac = eth_hdr(skb)->h_source;
+ dest_mac = eth_hdr(skb)->h_dest;
+
+ /* Pull the header */
+ hp = (struct fcoe_hdr *)skb->data;
+ fh = (struct fc_frame_header *) skb_transport_header(skb);
+ skb_pull(skb, sizeof(struct fcoe_hdr));
+ fr_len = skb->len - sizeof(struct fcoe_crc_eof);
+
+ fp = (struct fc_frame *)skb;
+ fc_frame_init(fp);
+ fr_dev(fp) = lport;
+ fr_sof(fp) = hp->fcoe_sof;
+ if (skb_copy_bits(skb, fr_len, &crc_eof, sizeof(crc_eof))) {
+ kfree_skb(skb);
+ return;
+ }
+ fr_eof(fp) = crc_eof.fcoe_eof;
+ fr_crc(fp) = crc_eof.fcoe_crc32;
+ if (pskb_trim(skb, fr_len)) {
+ kfree_skb(skb);
+ return;
+ }
+
+ fh = fc_frame_header_get(fp);
+
+ if (fh->fh_r_ctl == FC_RCTL_DD_SOL_DATA &&
+ fh->fh_type == FC_TYPE_FCP) {
+ /* Drop FCP data. We dont this in L2 path */
+ kfree_skb(skb);
+ return;
+ }
+ if (fh->fh_r_ctl == FC_RCTL_ELS_REQ &&
+ fh->fh_type == FC_TYPE_ELS) {
+ switch (fc_frame_payload_op(fp)) {
+ case ELS_LOGO:
+ if (ntoh24(fh->fh_s_id) == FC_FID_FLOGI) {
+ /* drop non-FIP LOGO */
+ kfree_skb(skb);
+ return;
+ }
+ break;
+ }
+ }
+
+ if (fh->fh_r_ctl == FC_RCTL_BA_ABTS) {
+ /* Drop incoming ABTS */
+ kfree_skb(skb);
+ return;
+ }
+
+ /*
+ * If a connection is uploading, drop incoming FCoE frames as there
+ * is a small window where we could try to return a frame while libfc
+ * is trying to clean things up.
+ */
+
+ /* Get fcport associated with d_id if it exists */
+ fcport = qedf_fcport_lookup(qedf, ntoh24(fh->fh_d_id));
+
+ if (fcport && test_bit(QEDF_RPORT_UPLOADING_CONNECTION,
+ &fcport->flags)) {
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2,
+ "Connection uploading, dropping fp=%p.\n", fp);
+ kfree_skb(skb);
+ return;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_LL2, "FCoE frame receive: "
+ "skb=%p fp=%p src=%06x dest=%06x r_ctl=%x fh_type=%x.\n", skb, fp,
+ ntoh24(fh->fh_s_id), ntoh24(fh->fh_d_id), fh->fh_r_ctl,
+ fh->fh_type);
+ if (qedf_dump_frames)
+ print_hex_dump(KERN_WARNING, "fcoe: ", DUMP_PREFIX_OFFSET, 16,
+ 1, skb->data, skb->len, false);
+ fc_exch_recv(lport, fp);
+}
+
+
+static int qedf_ll2_rx(void *cookie, struct sk_buff *skb,
+ u32 arg1, u32 arg2)
+{
+ struct qedf_ctx *qedf = (struct qedf_ctx *)cookie;
+ struct qedf_skb_work *work;
+ unsigned long flags;
+
+ work = kzalloc(sizeof(struct qedf_skb_work), GFP_ATOMIC);
+ if (!work) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not allocate work so "
+ "dropping frame.\n");
+ kfree_skb(skb);
+ return 0;
+ }
+
+ INIT_LIST_HEAD(&work->list);
+ work->skb = skb;
+ spin_lock_irqsave(&qedf->ll2_lock, flags);
+ list_add_tail(&work->list, &qedf->ll2_skb_list);
+ spin_unlock_irqrestore(&qedf->ll2_lock, flags);
+ wake_up_process(qedf->ll2_recv_thread);
+
+ return 0;
+}
+
+static struct qed_ll2_cb_ops qedf_ll2_cb_ops = {
+ .rx_cb = qedf_ll2_rx,
+ .tx_cb = NULL,
+};
+
+static int qedf_ll2_process_skb(struct qedf_ctx *qedf, struct sk_buff *skb)
+{
+ struct ethhdr *eh;
+
+ if (!qedf) {
+ QEDF_ERR(NULL, "qedf is NULL\n");
+ goto err_out;
+ }
+
+ eh = (struct ethhdr *)skb->data;
+ /* Undo VLAN encapsulation */
+ if (eh->h_proto == htons(ETH_P_8021Q)) {
+ memmove((u8 *)eh + VLAN_HLEN, eh, ETH_ALEN * 2);
+ eh = (struct ethhdr *)skb_pull(skb, VLAN_HLEN);
+ skb_reset_mac_header(skb);
+ }
+
+ /*
+ * Process either a FIP frame or FCoE frame based on the
+ * protocol value. If it's not either just drop the
+ * frame.
+ */
+ if (eh->h_proto == htons(ETH_P_FIP)) {
+ qedf_fip_recv(qedf, skb);
+ goto out;
+ } else if (eh->h_proto == htons(ETH_P_FCOE)) {
+ __skb_pull(skb, ETH_HLEN);
+ qedf_recv_frame(qedf, skb);
+ goto out;
+ } else
+ goto err_out;
+
+err_out:
+ kfree_skb(skb);
+out:
+ return 0;
+}
+
+/* Main thread to process skb's from light-L2 interface */
+static int qedf_ll2_recv_thread(void *arg)
+{
+ struct qedf_ctx *qedf = (struct qedf_ctx *)arg;
+ struct qedf_skb_work *work, *work_tmp;
+ unsigned long flags;
+
+ set_user_nice(current, -20);
+ set_current_state(TASK_INTERRUPTIBLE);
+
+ while (!kthread_should_stop()) {
+ schedule();
+ if (!list_empty(&qedf->ll2_skb_list)) {
+ list_for_each_entry_safe(work, work_tmp,
+ &qedf->ll2_skb_list, list) {
+ spin_lock_irqsave(&qedf->ll2_lock, flags);
+ list_del(&work->list);
+ spin_unlock_irqrestore(&qedf->ll2_lock, flags);
+ qedf_ll2_process_skb(qedf, work->skb);
+ kfree(work);
+ }
+ }
+ __set_current_state(TASK_INTERRUPTIBLE);
+ }
+
+ __set_current_state(TASK_RUNNING);
+ return 0;
+}
+
+/* Free any remaining skb's in the LL2 receive list */
+static void qedf_ll2_free_skbs(struct qedf_ctx *qedf)
+{
+ struct qedf_skb_work *work, *work_tmp;
+ unsigned long flags;
+
+ spin_lock_irqsave(&qedf->ll2_lock, flags);
+ list_for_each_entry_safe(work, work_tmp, &qedf->ll2_skb_list, list) {
+ list_del(&work->list);
+ if (work->skb)
+ kfree_skb(work->skb);
+ kfree(work);
+ }
+ spin_unlock_irqrestore(&qedf->ll2_lock, flags);
+}
+
+/* Main thread to process I/O completions */
+static int qedf_fp_io_thread(void *arg)
+{
+ struct qedf_percpu_iothread_s *iothread;
+ struct qedf_io_work *work, *work_tmp;
+ unsigned long flags;
+ u32 comp_type;
+
+ set_current_state(TASK_INTERRUPTIBLE);
+
+ iothread = (struct qedf_percpu_iothread_s *)arg;
+
+ while (!kthread_should_stop()) {
+ schedule();
+ if (!list_empty(&iothread->work_list)) {
+ list_for_each_entry_safe(work, work_tmp,
+ &iothread->work_list, list) {
+ spin_lock_irqsave(&iothread->work_lock, flags);
+ list_del(&work->list);
+ spin_unlock_irqrestore(&iothread->work_lock,
+ flags);
+ /* Record the response */
+ iothread->responses++;
+ /*
+ * Deferred part of unsolicited CQE sends
+ * frame to libfc.
+ */
+ comp_type = (work->cqe.cqe_data >>
+ FCOE_CQE_CQE_TYPE_SHIFT) &
+ FCOE_CQE_CQE_TYPE_MASK;
+ if (comp_type == FCOE_UNSOLIC_CQE_TYPE &&
+ work->fp)
+ fc_exch_recv(work->qedf->lport,
+ work->fp);
+ else
+ qedf_process_cqe(work->qedf,
+ &work->cqe);
+ mempool_free(work, work->qedf->io_mempool);
+ }
+ } else
+ __set_current_state(TASK_INTERRUPTIBLE);
+ }
+
+ __set_current_state(TASK_RUNNING);
+ return 0;
+}
+
+static int qedf_alloc_and_init_sb(struct qedf_ctx *qedf,
+ struct qed_sb_info *sb_info, u16 sb_id)
+{
+ struct status_block *sb_virt;
+ dma_addr_t sb_phys;
+ int ret;
+
+ sb_virt = dma_alloc_coherent(&qedf->pdev->dev,
+ sizeof(struct status_block), &sb_phys, GFP_KERNEL);
+
+ if (!sb_virt) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Status block allocation failed "
+ "for id = %d.\n", sb_id);
+ return -ENOMEM;
+ }
+
+ ret = qed_ops->common->sb_init(qedf->cdev, sb_info, sb_virt, sb_phys,
+ sb_id, QED_SB_TYPE_STORAGE);
+
+ if (ret) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Status block initialization "
+ "failed for id = %d.\n", sb_id);
+ return ret;
+ }
+
+ return 0;
+}
+
+static void qedf_free_sb(struct qedf_ctx *qedf, struct qed_sb_info *sb_info)
+{
+ if (sb_info->sb_virt)
+ dma_free_coherent(&qedf->pdev->dev, sizeof(*sb_info->sb_virt),
+ (void *)sb_info->sb_virt, sb_info->sb_phys);
+}
+
+static void qedf_destroy_sb(struct qedf_ctx *qedf)
+{
+ int id;
+ struct qedf_fastpath *fp = NULL;
+
+ for (id = 0; id < qedf->num_queues; id++) {
+ fp = &(qedf->fp_array[id]);
+ if (fp->sb_id == QEDF_SB_ID_NULL)
+ break;
+ qedf_free_sb(qedf, fp->sb_info);
+ kfree(fp->sb_info);
+ }
+ kfree(qedf->fp_array);
+}
+
+static int qedf_prepare_sb(struct qedf_ctx *qedf)
+{
+ int id;
+ struct qedf_fastpath *fp;
+ int ret;
+
+ qedf->fp_array =
+ kcalloc(qedf->num_queues, sizeof(struct qedf_fastpath),
+ GFP_KERNEL);
+
+ if (!qedf->fp_array) {
+ QEDF_ERR(&(qedf->dbg_ctx), "fastpath array allocation "
+ "failed.\n");
+ return -ENOMEM;
+ }
+
+ for (id = 0; id < qedf->num_queues; id++) {
+ fp = &(qedf->fp_array[id]);
+ fp->sb_id = QEDF_SB_ID_NULL;
+ fp->sb_info = kcalloc(1, sizeof(*fp->sb_info), GFP_KERNEL);
+ if (!fp->sb_info) {
+ QEDF_ERR(&(qedf->dbg_ctx), "SB info struct "
+ "allocation failed.\n");
+ goto err;
+ }
+ ret = qedf_alloc_and_init_sb(qedf, fp->sb_info, id);
+ if (ret) {
+ QEDF_ERR(&(qedf->dbg_ctx), "SB allocation and "
+ "initialization failed.\n");
+ goto err;
+ }
+ fp->sb_id = id;
+ fp->qedf = qedf;
+ fp->cq_num_entries =
+ qedf->global_queues[id]->cq_mem_size /
+ sizeof(struct fcoe_cqe);
+ }
+err:
+ return 0;
+}
+
+void qedf_process_cqe(struct qedf_ctx *qedf, struct fcoe_cqe *cqe)
+{
+ u16 xid;
+ struct qedf_ioreq *io_req;
+ struct qedf_rport *fcport;
+ u32 comp_type;
+
+ comp_type = (cqe->cqe_data >> FCOE_CQE_CQE_TYPE_SHIFT) &
+ FCOE_CQE_CQE_TYPE_MASK;
+
+ xid = cqe->cqe_data & FCOE_CQE_TASK_ID_MASK;
+ io_req = &qedf->cmd_mgr->cmds[xid];
+
+ /* Completion not for a valid I/O anymore so just return */
+ if (!io_req)
+ return;
+
+ fcport = io_req->fcport;
+
+ if (fcport == NULL) {
+ QEDF_ERR(&(qedf->dbg_ctx), "fcport is NULL.\n");
+ return;
+ }
+
+ /*
+ * Check that fcport is offloaded. If it isn't then the spinlock
+ * isn't valid and shouldn't be taken. We should just return.
+ */
+ if (!test_bit(QEDF_RPORT_SESSION_READY, &fcport->flags)) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Session not offloaded yet.\n");
+ return;
+ }
+
+
+ switch (comp_type) {
+ case FCOE_GOOD_COMPLETION_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ switch (io_req->cmd_type) {
+ case QEDF_SCSI_CMD:
+ qedf_scsi_completion(qedf, cqe, io_req);
+ break;
+ case QEDF_ELS:
+ qedf_process_els_compl(qedf, cqe, io_req);
+ break;
+ case QEDF_TASK_MGMT_CMD:
+ qedf_process_tmf_compl(qedf, cqe, io_req);
+ break;
+ case QEDF_SEQ_CLEANUP:
+ qedf_process_seq_cleanup_compl(qedf, cqe, io_req);
+ break;
+ }
+ break;
+ case FCOE_ERROR_DETECTION_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Error detect CQE.\n");
+ qedf_process_error_detect(qedf, cqe, io_req);
+ break;
+ case FCOE_EXCH_CLEANUP_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Cleanup CQE.\n");
+ qedf_process_cleanup_compl(qedf, cqe, io_req);
+ break;
+ case FCOE_ABTS_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Abort CQE.\n");
+ qedf_process_abts_compl(qedf, cqe, io_req);
+ break;
+ case FCOE_DUMMY_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Dummy CQE.\n");
+ break;
+ case FCOE_LOCAL_COMP_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Local completion CQE.\n");
+ break;
+ case FCOE_WARNING_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Warning CQE.\n");
+ qedf_process_warning_compl(qedf, cqe, io_req);
+ break;
+ case MAX_FCOE_CQE_TYPE:
+ atomic_inc(&fcport->free_sqes);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Max FCoE CQE.\n");
+ break;
+ default:
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_IO,
+ "Default CQE.\n");
+ break;
+ }
+}
+
+static void qedf_free_bdq(struct qedf_ctx *qedf)
+{
+ int i;
+
+ if (qedf->bdq_pbl_list)
+ dma_free_coherent(&qedf->pdev->dev, QEDF_PAGE_SIZE,
+ qedf->bdq_pbl_list, qedf->bdq_pbl_list_dma);
+
+ if (qedf->bdq_pbl)
+ dma_free_coherent(&qedf->pdev->dev, qedf->bdq_pbl_mem_size,
+ qedf->bdq_pbl, qedf->bdq_pbl_dma);
+
+ for (i = 0; i < QEDF_BDQ_SIZE; i++) {
+ if (qedf->bdq[i].buf_addr) {
+ dma_free_coherent(&qedf->pdev->dev, QEDF_BDQ_BUF_SIZE,
+ qedf->bdq[i].buf_addr, qedf->bdq[i].buf_dma);
+ }
+ }
+}
+
+static void qedf_free_global_queues(struct qedf_ctx *qedf)
+{
+ int i;
+ struct global_queue **gl = qedf->global_queues;
+
+ for (i = 0; i < qedf->num_queues; i++) {
+ if (!gl[i])
+ continue;
+
+ if (gl[i]->cq)
+ dma_free_coherent(&qedf->pdev->dev,
+ gl[i]->cq_mem_size, gl[i]->cq, gl[i]->cq_dma);
+ if (gl[i]->cq_pbl)
+ dma_free_coherent(&qedf->pdev->dev, gl[i]->cq_pbl_size,
+ gl[i]->cq_pbl, gl[i]->cq_pbl_dma);
+
+ kfree(gl[i]);
+ }
+
+ qedf_free_bdq(qedf);
+}
+
+static int qedf_alloc_bdq(struct qedf_ctx *qedf)
+{
+ int i;
+ struct scsi_bd *pbl;
+ u64 *list;
+ dma_addr_t page;
+
+ /* Alloc dma memory for BDQ buffers */
+ for (i = 0; i < QEDF_BDQ_SIZE; i++) {
+ qedf->bdq[i].buf_addr = dma_alloc_coherent(&qedf->pdev->dev,
+ QEDF_BDQ_BUF_SIZE, &qedf->bdq[i].buf_dma, GFP_KERNEL);
+ if (!qedf->bdq[i].buf_addr) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not allocate BDQ "
+ "buffer %d.\n", i);
+ return -ENOMEM;
+ }
+ }
+
+ /* Alloc dma memory for BDQ page buffer list */
+ qedf->bdq_pbl_mem_size =
+ QEDF_BDQ_SIZE * sizeof(struct scsi_bd);
+ qedf->bdq_pbl_mem_size =
+ ALIGN(qedf->bdq_pbl_mem_size, QEDF_PAGE_SIZE);
+
+ qedf->bdq_pbl = dma_alloc_coherent(&qedf->pdev->dev,
+ qedf->bdq_pbl_mem_size, &qedf->bdq_pbl_dma, GFP_KERNEL);
+ if (!qedf->bdq_pbl) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not allocate BDQ PBL.\n");
+ return -ENOMEM;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "BDQ PBL addr=0x%p dma=0x%llx.\n", qedf->bdq_pbl,
+ qedf->bdq_pbl_dma);
+
+ /*
+ * Populate BDQ PBL with physical and virtual address of individual
+ * BDQ buffers
+ */
+ pbl = (struct scsi_bd *)qedf->bdq_pbl;
+ for (i = 0; i < QEDF_BDQ_SIZE; i++) {
+ pbl->address.hi = cpu_to_le32(U64_HI(qedf->bdq[i].buf_dma));
+ pbl->address.lo = cpu_to_le32(U64_LO(qedf->bdq[i].buf_dma));
+ pbl->opaque.hi = 0;
+ /* Opaque lo data is an index into the BDQ array */
+ pbl->opaque.lo = cpu_to_le32(i);
+ pbl++;
+ }
+
+ /* Allocate list of PBL pages */
+ qedf->bdq_pbl_list = dma_alloc_coherent(&qedf->pdev->dev,
+ QEDF_PAGE_SIZE, &qedf->bdq_pbl_list_dma, GFP_KERNEL);
+ if (!qedf->bdq_pbl_list) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not allocate list of PBL "
+ "pages.\n");
+ return -ENOMEM;
+ }
+ memset(qedf->bdq_pbl_list, 0, QEDF_PAGE_SIZE);
+
+ /*
+ * Now populate PBL list with pages that contain pointers to the
+ * individual buffers.
+ */
+ qedf->bdq_pbl_list_num_entries = qedf->bdq_pbl_mem_size /
+ QEDF_PAGE_SIZE;
+ list = (u64 *)qedf->bdq_pbl_list;
+ page = qedf->bdq_pbl_list_dma;
+ for (i = 0; i < qedf->bdq_pbl_list_num_entries; i++) {
+ *list = qedf->bdq_pbl_dma;
+ list++;
+ page += QEDF_PAGE_SIZE;
+ }
+
+ return 0;
+}
+
+static int qedf_alloc_global_queues(struct qedf_ctx *qedf)
+{
+ u32 *list;
+ int i;
+ int status = 0, rc;
+ u32 *pbl;
+ dma_addr_t page;
+ int num_pages;
+
+ /* Allocate and map CQs, RQs */
+ /*
+ * Number of global queues (CQ / RQ). This should
+ * be <= number of available MSIX vectors for the PF
+ */
+ if (!qedf->num_queues) {
+ QEDF_ERR(&(qedf->dbg_ctx), "No MSI-X vectors available!\n");
+ return 1;
+ }
+
+ /*
+ * Make sure we allocated the PBL that will contain the physical
+ * addresses of our queues
+ */
+ if (!qedf->p_cpuq) {
+ status = 1;
+ goto mem_alloc_failure;
+ }
+
+ qedf->global_queues = kzalloc((sizeof(struct global_queue *)
+ * qedf->num_queues), GFP_KERNEL);
+ if (!qedf->global_queues) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Unable to allocate global "
+ "queues array ptr memory\n");
+ return -ENOMEM;
+ }
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "qedf->global_queues=%p.\n", qedf->global_queues);
+
+ /* Allocate DMA coherent buffers for BDQ */
+ rc = qedf_alloc_bdq(qedf);
+ if (rc)
+ goto mem_alloc_failure;
+
+ /* Allocate a CQ and an associated PBL for each MSI-X vector */
+ for (i = 0; i < qedf->num_queues; i++) {
+ qedf->global_queues[i] = kzalloc(sizeof(struct global_queue),
+ GFP_KERNEL);
+ if (!qedf->global_queues[i]) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Unable to allocation "
+ "global queue %d.\n", i);
+ goto mem_alloc_failure;
+ }
+
+ qedf->global_queues[i]->cq_mem_size =
+ FCOE_PARAMS_CQ_NUM_ENTRIES * sizeof(struct fcoe_cqe);
+ qedf->global_queues[i]->cq_mem_size =
+ ALIGN(qedf->global_queues[i]->cq_mem_size, QEDF_PAGE_SIZE);
+
+ qedf->global_queues[i]->cq_pbl_size =
+ (qedf->global_queues[i]->cq_mem_size /
+ PAGE_SIZE) * sizeof(void *);
+ qedf->global_queues[i]->cq_pbl_size =
+ ALIGN(qedf->global_queues[i]->cq_pbl_size, QEDF_PAGE_SIZE);
+
+ qedf->global_queues[i]->cq =
+ dma_alloc_coherent(&qedf->pdev->dev,
+ qedf->global_queues[i]->cq_mem_size,
+ &qedf->global_queues[i]->cq_dma, GFP_KERNEL);
+
+ if (!qedf->global_queues[i]->cq) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not allocate "
+ "cq.\n");
+ status = -ENOMEM;
+ goto mem_alloc_failure;
+ }
+ memset(qedf->global_queues[i]->cq, 0,
+ qedf->global_queues[i]->cq_mem_size);
+
+ qedf->global_queues[i]->cq_pbl =
+ dma_alloc_coherent(&qedf->pdev->dev,
+ qedf->global_queues[i]->cq_pbl_size,
+ &qedf->global_queues[i]->cq_pbl_dma, GFP_KERNEL);
+
+ if (!qedf->global_queues[i]->cq_pbl) {
+ QEDF_WARN(&(qedf->dbg_ctx), "Could not allocate "
+ "cq PBL.\n");
+ status = -ENOMEM;
+ goto mem_alloc_failure;
+ }
+ memset(qedf->global_queues[i]->cq_pbl, 0,
+ qedf->global_queues[i]->cq_pbl_size);
+
+ /* Create PBL */
+ num_pages = qedf->global_queues[i]->cq_mem_size /
+ QEDF_PAGE_SIZE;
+ page = qedf->global_queues[i]->cq_dma;
+ pbl = (u32 *)qedf->global_queues[i]->cq_pbl;
+
+ while (num_pages--) {
+ *pbl = U64_LO(page);
+ pbl++;
+ *pbl = U64_HI(page);
+ pbl++;
+ page += QEDF_PAGE_SIZE;
+ }
+ /* Set the initial consumer index for cq */
+ qedf->global_queues[i]->cq_cons_idx = 0;
+ }
+
+ list = (u32 *)qedf->p_cpuq;
+
+ /*
+ * The list is built as follows: CQ#0 PBL pointer, RQ#0 PBL pointer,
+ * CQ#1 PBL pointer, RQ#1 PBL pointer, etc. Each PBL pointer points
+ * to the physical address which contains an array of pointers to
+ * the physical addresses of the specific queue pages.
+ */
+ for (i = 0; i < qedf->num_queues; i++) {
+ *list = U64_LO(qedf->global_queues[i]->cq_pbl_dma);
+ list++;
+ *list = U64_HI(qedf->global_queues[i]->cq_pbl_dma);
+ list++;
+ *list = U64_LO(0);
+ list++;
+ *list = U64_HI(0);
+ list++;
+ }
+
+ return 0;
+
+mem_alloc_failure:
+ qedf_free_global_queues(qedf);
+ return status;
+}
+
+static int qedf_set_fcoe_pf_param(struct qedf_ctx *qedf)
+{
+ u8 sq_num_pbl_pages;
+ u32 sq_mem_size;
+ u32 cq_mem_size;
+ u32 cq_num_entries;
+ int rval;
+
+ /*
+ * The number of completion queues/fastpath interrupts/status blocks
+ * we allocation is the minimum off:
+ *
+ * Number of CPUs
+ * Number of MSI-X vectors
+ * Max number allocated in hardware (QEDF_MAX_NUM_CQS)
+ */
+ qedf->num_queues = min((unsigned int)QEDF_MAX_NUM_CQS,
+ num_online_cpus());
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Number of CQs is %d.\n",
+ qedf->num_queues);
+
+ qedf->p_cpuq = pci_alloc_consistent(qedf->pdev,
+ qedf->num_queues * sizeof(struct qedf_glbl_q_params),
+ &qedf->hw_p_cpuq);
+
+ if (!qedf->p_cpuq) {
+ QEDF_ERR(&(qedf->dbg_ctx), "pci_alloc_consistent failed.\n");
+ return 1;
+ }
+
+ rval = qedf_alloc_global_queues(qedf);
+ if (rval) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Global queue allocation "
+ "failed.\n");
+ return 1;
+ }
+
+ /* Calculate SQ PBL size in the same manner as in qedf_sq_alloc() */
+ sq_mem_size = SQ_NUM_ENTRIES * sizeof(struct fcoe_wqe);
+ sq_mem_size = ALIGN(sq_mem_size, QEDF_PAGE_SIZE);
+ sq_num_pbl_pages = (sq_mem_size / QEDF_PAGE_SIZE);
+
+ /* Calculate CQ num entries */
+ cq_mem_size = FCOE_PARAMS_CQ_NUM_ENTRIES * sizeof(struct fcoe_cqe);
+ cq_mem_size = ALIGN(cq_mem_size, QEDF_PAGE_SIZE);
+ cq_num_entries = cq_mem_size / sizeof(struct fcoe_cqe);
+
+ memset(&(qedf->pf_params), 0,
+ sizeof(qedf->pf_params));
+
+ /* Setup the value for fcoe PF */
+ qedf->pf_params.fcoe_pf_params.num_cons = QEDF_MAX_SESSIONS;
+ qedf->pf_params.fcoe_pf_params.num_tasks = FCOE_PARAMS_NUM_TASKS;
+ qedf->pf_params.fcoe_pf_params.glbl_q_params_addr =
+ (u64)qedf->hw_p_cpuq;
+ qedf->pf_params.fcoe_pf_params.sq_num_pbl_pages = sq_num_pbl_pages;
+
+ qedf->pf_params.fcoe_pf_params.rq_buffer_log_size = 0;
+
+ qedf->pf_params.fcoe_pf_params.cq_num_entries = cq_num_entries;
+ qedf->pf_params.fcoe_pf_params.num_cqs = qedf->num_queues;
+
+ /* log_page_size: 12 for 4KB pages */
+ qedf->pf_params.fcoe_pf_params.log_page_size = ilog2(QEDF_PAGE_SIZE);
+
+ qedf->pf_params.fcoe_pf_params.mtu = 9000;
+ qedf->pf_params.fcoe_pf_params.gl_rq_pi = QEDF_FCOE_PARAMS_GL_RQ_PI;
+ qedf->pf_params.fcoe_pf_params.gl_cmd_pi = QEDF_FCOE_PARAMS_GL_CMD_PI;
+
+ /* BDQ address and size */
+ qedf->pf_params.fcoe_pf_params.bdq_pbl_base_addr[0] =
+ qedf->bdq_pbl_list_dma;
+ qedf->pf_params.fcoe_pf_params.bdq_pbl_num_entries[0] =
+ qedf->bdq_pbl_list_num_entries;
+ qedf->pf_params.fcoe_pf_params.rq_buffer_size = QEDF_BDQ_BUF_SIZE;
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "bdq_list=%p bdq_pbl_list_dma=%llx bdq_pbl_list_entries=%d.\n",
+ qedf->bdq_pbl_list,
+ qedf->pf_params.fcoe_pf_params.bdq_pbl_base_addr[0],
+ qedf->pf_params.fcoe_pf_params.bdq_pbl_num_entries[0]);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "cq_num_entries=%d.\n",
+ qedf->pf_params.fcoe_pf_params.cq_num_entries);
+
+ return 0;
+}
+
+/* Free DMA coherent memory for array of queue pointers we pass to qed */
+static void qedf_free_fcoe_pf_param(struct qedf_ctx *qedf)
+{
+ size_t size = 0;
+
+ if (qedf->p_cpuq) {
+ size = qedf->num_queues * sizeof(struct qedf_glbl_q_params);
+ pci_free_consistent(qedf->pdev, size, qedf->p_cpuq,
+ qedf->hw_p_cpuq);
+ }
+
+ qedf_free_global_queues(qedf);
+
+ if (qedf->global_queues)
+ kfree(qedf->global_queues);
+}
+
+/*
+ * PCI driver functions
+ */
+
+static const struct pci_device_id qedf_pci_tbl[] = {
+ { PCI_DEVICE(PCI_VENDOR_ID_QLOGIC, 0x165c) },
+ { PCI_DEVICE(PCI_VENDOR_ID_QLOGIC, 0x8080) },
+ {0}
+};
+MODULE_DEVICE_TABLE(pci, qedf_pci_tbl);
+
+static enum cpuhp_state qedf_cpuhp_state;
+
+static struct pci_driver qedf_pci_driver = {
+ .name = QEDF_MODULE_NAME,
+ .id_table = qedf_pci_tbl,
+ .probe = qedf_probe,
+ .remove = qedf_remove,
+};
+
+static int __qedf_probe(struct pci_dev *pdev, int mode)
+{
+ int rc;
+ struct fc_lport *lport;
+ struct qedf_ctx *qedf;
+ struct Scsi_Host *host;
+ bool is_vf = false;
+ struct qed_ll2_params params;
+ char host_buf[20];
+ struct qed_link_params link_params;
+ int status;
+ void *task_start, *task_end;
+ struct qed_slowpath_params slowpath_params;
+ struct qed_probe_params qed_params;
+ u16 tmp;
+
+ /*
+ * When doing error recovery we didn't reap the lport so don't try
+ * to reallocate it.
+ */
+ if (mode != QEDF_MODE_RECOVERY) {
+ lport = libfc_host_alloc(&qedf_host_template,
+ sizeof(struct qedf_ctx));
+
+ if (!lport) {
+ QEDF_ERR(NULL, "Could not allocate lport.\n");
+ rc = -ENOMEM;
+ goto err0;
+ }
+
+ /* Initialize qedf_ctx */
+ qedf = lport_priv(lport);
+ qedf->lport = lport;
+ qedf->ctlr.lp = lport;
+ qedf->pdev = pdev;
+ qedf->dbg_ctx.pdev = pdev;
+ qedf->dbg_ctx.host_no = lport->host->host_no;
+ spin_lock_init(&qedf->hba_lock);
+ spin_lock_init(&qedf->ll2_lock);
+ INIT_LIST_HEAD(&qedf->ll2_skb_list);
+ INIT_LIST_HEAD(&qedf->fcports);
+ qedf->curr_conn_id = QEDF_MAX_SESSIONS - 1;
+ atomic_set(&qedf->num_offloads, 0);
+ qedf->stop_io_on_error = false;
+ pci_set_drvdata(pdev, qedf);
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_INFO,
+ "QLogic FastLinQ FCoE Module qedf %s, "
+ "FW %d.%d.%d.%d\n", QEDF_VERSION,
+ FW_MAJOR_VERSION, FW_MINOR_VERSION, FW_REVISION_VERSION,
+ FW_ENGINEERING_VERSION);
+ } else {
+ /* Init pointers during recovery */
+ qedf = pci_get_drvdata(pdev);
+ lport = qedf->lport;
+ }
+
+ host = lport->host;
+
+ /* Allocate mempool for qedf_io_work structs */
+ qedf->io_mempool = mempool_create_slab_pool(QEDF_IO_WORK_MIN,
+ qedf_io_work_cache);
+ if (qedf->io_mempool == NULL) {
+ QEDF_ERR(&(qedf->dbg_ctx), "qedf->io_mempool is NULL.\n");
+ goto err1;
+ }
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_INFO, "qedf->io_mempool=%p.\n",
+ qedf->io_mempool);
+
+ sprintf(host_buf, "qedf_%u_link",
+ qedf->lport->host->host_no);
+ qedf->link_update_wq = create_singlethread_workqueue(host_buf);
+ INIT_DELAYED_WORK(&qedf->link_update, qedf_handle_link_update);
+ INIT_DELAYED_WORK(&qedf->link_recovery, qedf_link_recovery);
+
+ qedf->fipvlan_retries = qedf_fipvlan_retries;
+
+ /*
+ * Common probe. Takes care of basic hardware init and pci_*
+ * functions.
+ */
+ memset(&qed_params, 0, sizeof(qed_params));
+ qed_params.protocol = QED_PROTOCOL_FCOE;
+ qed_params.dp_module = qedf_dp_module;
+ qed_params.dp_level = qedf_dp_level;
+ qed_params.is_vf = is_vf;
+ qedf->cdev = qed_ops->common->probe(pdev, &qed_params);
+ if (!qedf->cdev) {
+ rc = -ENODEV;
+ goto err1;
+ }
+
+ /* queue allocation code should come here
+ * order should be
+ * slowpath_start
+ * status block allocation
+ * interrupt registration (to get min number of queues)
+ * set_fcoe_pf_param
+ * qed_sp_fcoe_func_start
+ */
+ rc = qedf_set_fcoe_pf_param(qedf);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Cannot set fcoe pf param.\n");
+ goto err2;
+ }
+ qed_ops->common->update_pf_params(qedf->cdev, &qedf->pf_params);
+
+ /* Learn information crucial for qedf to progress */
+ rc = qed_ops->fill_dev_info(qedf->cdev, &qedf->dev_info);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed to dev info.\n");
+ goto err1;
+ }
+
+ /* Record BDQ producer doorbell addresses */
+ qedf->bdq_primary_prod = qedf->dev_info.primary_dbq_rq_addr;
+ qedf->bdq_secondary_prod = qedf->dev_info.secondary_bdq_rq_addr;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "BDQ primary_prod=%p secondary_prod=%p.\n", qedf->bdq_primary_prod,
+ qedf->bdq_secondary_prod);
+
+ qed_ops->register_ops(qedf->cdev, &qedf_cb_ops, qedf);
+
+ rc = qedf_prepare_sb(qedf);
+ if (rc) {
+
+ QEDF_ERR(&(qedf->dbg_ctx), "Cannot start slowpath.\n");
+ goto err2;
+ }
+
+ /* Start the Slowpath-process */
+ slowpath_params.int_mode = QED_INT_MODE_MSIX;
+ slowpath_params.drv_major = QEDF_DRIVER_MAJOR_VER;
+ slowpath_params.drv_minor = QEDF_DRIVER_MINOR_VER;
+ slowpath_params.drv_rev = QEDF_DRIVER_REV_VER;
+ slowpath_params.drv_eng = QEDF_DRIVER_ENG_VER;
+ memcpy(slowpath_params.name, "qedf", QED_DRV_VER_STR_SIZE);
+ rc = qed_ops->common->slowpath_start(qedf->cdev, &slowpath_params);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Cannot start slowpath.\n");
+ goto err2;
+ }
+
+ /*
+ * update_pf_params needs to be called before and after slowpath
+ * start
+ */
+ qed_ops->common->update_pf_params(qedf->cdev, &qedf->pf_params);
+
+ /* Setup interrupts */
+ rc = qedf_setup_int(qedf);
+ if (rc)
+ goto err3;
+
+ rc = qed_ops->start(qedf->cdev, &qedf->tasks);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Cannot start FCoE function.\n");
+ goto err4;
+ }
+ task_start = qedf_get_task_mem(&qedf->tasks, 0);
+ task_end = qedf_get_task_mem(&qedf->tasks, MAX_TID_BLOCKS_FCOE - 1);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "Task context start=%p, "
+ "end=%p block_size=%u.\n", task_start, task_end,
+ qedf->tasks.size);
+
+ /*
+ * We need to write the number of BDs in the BDQ we've preallocated so
+ * the f/w will do a prefetch and we'll get an unsolicited CQE when a
+ * packet arrives.
+ */
+ qedf->bdq_prod_idx = QEDF_BDQ_SIZE;
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "Writing %d to primary and secondary BDQ doorbell registers.\n",
+ qedf->bdq_prod_idx);
+ writew(qedf->bdq_prod_idx, qedf->bdq_primary_prod);
+ tmp = readw(qedf->bdq_primary_prod);
+ writew(qedf->bdq_prod_idx, qedf->bdq_secondary_prod);
+ tmp = readw(qedf->bdq_secondary_prod);
+
+ qed_ops->common->set_power_state(qedf->cdev, PCI_D0);
+
+ /* Now that the dev_info struct has been filled in set the MAC
+ * address
+ */
+ ether_addr_copy(qedf->mac, qedf->dev_info.common.hw_mac);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "MAC address is %pM.\n",
+ qedf->mac);
+
+ /* Set the WWNN and WWPN based on the MAC address */
+ qedf->wwnn = fcoe_wwn_from_mac(qedf->mac, 1, 0);
+ qedf->wwpn = fcoe_wwn_from_mac(qedf->mac, 2, 0);
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC, "WWNN=%016llx "
+ "WWPN=%016llx.\n", qedf->wwnn, qedf->wwpn);
+
+ sprintf(host_buf, "host_%d", host->host_no);
+ qed_ops->common->set_id(qedf->cdev, host_buf, QEDF_VERSION);
+
+
+ /* Set xid max values */
+ qedf->max_scsi_xid = QEDF_MAX_SCSI_XID;
+ qedf->max_els_xid = QEDF_MAX_ELS_XID;
+
+ /* Allocate cmd mgr */
+ qedf->cmd_mgr = qedf_cmd_mgr_alloc(qedf);
+ if (!qedf->cmd_mgr) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed to allocate cmd mgr.\n");
+ goto err5;
+ }
+
+ if (mode != QEDF_MODE_RECOVERY) {
+ host->transportt = qedf_fc_transport_template;
+ host->can_queue = QEDF_MAX_ELS_XID;
+ host->max_lun = qedf_max_lun;
+ host->max_cmd_len = QEDF_MAX_CDB_LEN;
+ rc = scsi_add_host(host, &pdev->dev);
+ if (rc)
+ goto err6;
+ }
+
+ memset(¶ms, 0, sizeof(params));
+ params.mtu = 9000;
+ ether_addr_copy(params.ll2_mac_address, qedf->mac);
+
+ /* Start LL2 processing thread */
+ snprintf(host_buf, 20, "qedf_%d_ll2", host->host_no);
+ qedf->ll2_recv_thread = kthread_run(qedf_ll2_recv_thread,
+ (void *)qedf, host_buf);
+
+#ifdef CONFIG_DEBUG_FS
+ qedf_dbg_host_init(&(qedf->dbg_ctx), &qedf_debugfs_ops,
+ &qedf_dbg_fops);
+#endif
+
+ /* Start LL2 */
+ qed_ops->ll2->register_cb_ops(qedf->cdev, &qedf_ll2_cb_ops, qedf);
+ rc = qed_ops->ll2->start(qedf->cdev, ¶ms);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Could not start Light L2.\n");
+ goto err7;
+ }
+ set_bit(QEDF_LL2_STARTED, &qedf->flags);
+
+ /* hw will be insterting vlan tag*/
+ qedf->vlan_hw_insert = 1;
+ qedf->vlan_id = 0;
+
+ /*
+ * No need to setup fcoe_ctlr or fc_lport objects during recovery since
+ * they were not reaped during the unload process.
+ */
+ if (mode != QEDF_MODE_RECOVERY) {
+ /* Setup imbedded fcoe controller */
+ qedf_fcoe_ctlr_setup(qedf);
+
+ /* Setup lport */
+ rc = qedf_lport_setup(qedf);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "qedf_lport_setup failed.\n");
+ goto err7;
+ }
+ }
+
+ sprintf(host_buf, "qedf_%u_timer", qedf->lport->host->host_no);
+ qedf->timer_work_queue =
+ create_singlethread_workqueue(host_buf);
+ if (!qedf->timer_work_queue) {
+ QEDF_ERR(&(qedf->dbg_ctx), "Failed to start timer "
+ "workqueue.\n");
+ goto err7;
+ }
+
+ /* DPC workqueue is not reaped during recovery unload */
+ if (mode != QEDF_MODE_RECOVERY) {
+ sprintf(host_buf, "qedf_%u_dpc",
+ qedf->lport->host->host_no);
+ qedf->dpc_wq = create_singlethread_workqueue(host_buf);
+ }
+
+ /*
+ * GRC dump and sysfs parameters are not reaped during the recovery
+ * unload process.
+ */
+ if (mode != QEDF_MODE_RECOVERY) {
+ qedf->grcdump_size = qed_ops->common->dbg_grc_size(qedf->cdev);
+ if (qedf->grcdump_size) {
+ rc = qedf_alloc_grc_dump_buf(&qedf->grcdump,
+ qedf->grcdump_size);
+ if (rc) {
+ QEDF_ERR(&(qedf->dbg_ctx),
+ "GRC Dump buffer alloc failed.\n");
+ qedf->grcdump = NULL;
+ }
+
+ QEDF_INFO(&(qedf->dbg_ctx), QEDF_LOG_DISC,
+ "grcdump: addr=%p, size=%u.\n",
+ qedf->grcdump, qedf->grcdump_size);
+ }
+ qedf_create_sysfs_ctx_attr(qedf);
+
+ /* Initialize I/O tracing for this adapter */
+ spin_lock_init(&qedf->io_trace_lock);
+ qedf->io_trace_idx = 0;
+ }
+
+ init_completion(&qedf->flogi_compl);
+
+ memset(&link_params, 0, sizeof(struct qed_link_params));
+ link_params.link_up = true;
+ status = qed_ops->common->set_link(qedf->cdev, &link_params);
+ if (status)
+ QEDF_WARN(&(qedf->dbg_ctx), "set_link failed.\n");
+
+ /* Start/restart discovery */
+ if (mode == QEDF_MODE_RECOVERY)
+ fcoe_ctlr_link_up(&qedf->ctlr);
+ else
+ fc_fabric_login(lport);
+
+ /* All good */
+ return 0;
+
+err7:
+ fc_remove_host(qedf->lport->host);
+ scsi_remove_host(qedf->lport->host);
+#ifdef CONFIG_DEBUG_FS
+ qedf_dbg_host_exit(&(qedf->dbg_ctx));
+#endif
+err6:
+ qedf_cmd_mgr_free(qedf->cmd_mgr);
+err5:
+ qed_ops->stop(qedf->cdev);
+err4:
+ qedf_free_fcoe_pf_param(qedf);
+ qedf_sync_free_irqs(qedf);
+err3:
+ qed_ops->common->slowpath_stop(qedf->cdev);
+err2:
+ qed_ops->common->remove(qedf->cdev);
+err1:
+ scsi_host_put(lport->host);
+err0:
+ return rc;
+}
+
+static int qedf_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+ return __qedf_probe(pdev, QEDF_MODE_NORMAL);
+}
+
+static void __qedf_remove(struct pci_dev *pdev, int mode)
+{
+ struct qedf_ctx *qedf;
+
+ if (!pdev) {
+ QEDF_ERR(NULL, "pdev is NULL.\n");
+ return;
+ }
+
+ qedf = pci_get_drvdata(pdev);
+
+ /*
+ * Prevent race where we're in board disable work and then try to
+ * rmmod the module.
+ */
+ if (test_bit(QEDF_UNLOADING, &qedf->flags)) {
+ QEDF_ERR(&qedf->dbg_ctx, "Already removing PCI function.\n");
+ return;
+ }
+
+ if (mode != QEDF_MODE_RECOVERY)
+ set_bit(QEDF_UNLOADING, &qedf->flags);
+
+ /* Logoff the fabric to upload all connections */
+ if (mode == QEDF_MODE_RECOVERY)
+ fcoe_ctlr_link_down(&qedf->ctlr);
+ else
+ fc_fabric_logoff(qedf->lport);
+ qedf_wait_for_upload(qedf);
+
+#ifdef CONFIG_DEBUG_FS
+ qedf_dbg_host_exit(&(qedf->dbg_ctx));
+#endif
+
+ /* Stop any link update handling */
+ cancel_delayed_work_sync(&qedf->link_update);
+ destroy_workqueue(qedf->link_update_wq);
+ qedf->link_update_wq = NULL;
+
+ if (qedf->timer_work_queue)
+ destroy_workqueue(qedf->timer_work_queue);
+
+ /* Stop Light L2 */
+ clear_bit(QEDF_LL2_STARTED, &qedf->flags);
+ qed_ops->ll2->stop(qedf->cdev);
+ if (qedf->ll2_recv_thread)
+ kthread_stop(qedf->ll2_recv_thread);
+ qedf_ll2_free_skbs(qedf);
+
+ /* Stop fastpath */
+ qedf_sync_free_irqs(qedf);
+ qedf_destroy_sb(qedf);
+
+ /*
+ * During recovery don't destroy OS constructs that represent the
+ * physical port.
+ */
+ if (mode != QEDF_MODE_RECOVERY) {
+ qedf_free_grc_dump_buf(&qedf->grcdump);
+ qedf_remove_sysfs_ctx_attr(qedf);
+
+ /* Remove all SCSI/libfc/libfcoe structures */
+ fcoe_ctlr_destroy(&qedf->ctlr);
+ fc_lport_destroy(qedf->lport);
+ fc_remove_host(qedf->lport->host);
+ scsi_remove_host(qedf->lport->host);
+ }
+
+ qedf_cmd_mgr_free(qedf->cmd_mgr);
+
+ if (mode != QEDF_MODE_RECOVERY) {
+ fc_exch_mgr_free(qedf->lport);
+ fc_lport_free_stats(qedf->lport);
+
+ /* Wait for all vports to be reaped */
+ qedf_wait_for_vport_destroy(qedf);
+ }
+
+ /*
+ * Now that all connections have been uploaded we can stop the
+ * rest of the qed operations
+ */
+ qed_ops->stop(qedf->cdev);
+
+ if (mode != QEDF_MODE_RECOVERY) {
+ if (qedf->dpc_wq) {
+ /* Stop general DPC handling */
+ destroy_workqueue(qedf->dpc_wq);
+ qedf->dpc_wq = NULL;
+ }
+ }
+
+ /* Final shutdown for the board */
+ qedf_free_fcoe_pf_param(qedf);
+ if (mode != QEDF_MODE_RECOVERY) {
+ qed_ops->common->set_power_state(qedf->cdev, PCI_D0);
+ pci_set_drvdata(pdev, NULL);
+ }
+ qed_ops->common->slowpath_stop(qedf->cdev);
+ qed_ops->common->remove(qedf->cdev);
+
+ mempool_destroy(qedf->io_mempool);
+
+ /* Only reap the Scsi_host on a real removal */
+ if (mode != QEDF_MODE_RECOVERY)
+ scsi_host_put(qedf->lport->host);
+}
+
+static void qedf_remove(struct pci_dev *pdev)
+{
+ /* Check to make sure this function wasn't already disabled */
+ if (!atomic_read(&pdev->enable_cnt))
+ return;
+
+ __qedf_remove(pdev, QEDF_MODE_NORMAL);
+}
+
+/*
+ * Per CPU I/O thread initialization code.
+ */
+static int qedf_cpu_online(unsigned int cpu)
+{
+ struct qedf_percpu_iothread_s *iothread =
+ this_cpu_ptr(&qedf_percpu_iothreads);
+ char io_thread_str[20];
+
+ sprintf(io_thread_str, "qedf_io_%d", cpu);
+ iothread->iothread = kthread_create(qedf_fp_io_thread,
+ (void *)iothread, io_thread_str);
+
+ if (IS_ERR(iothread->iothread)) {
+ QEDF_ERR(NULL, "Failed to create I/O thread for CPU %d.\n",
+ cpu);
+ return PTR_ERR(iothread->iothread);
+ }
+
+ kthread_bind(iothread->iothread, cpu);
+ wake_up_process(iothread->iothread);
+
+ QEDF_INFO(NULL, QEDF_LOG_DISC,
+ "Creating I/O thread for CPU %d at %p.\n", cpu,
+ iothread->iothread);
+ return 0;
+}
+
+/* Free any remaining I/O work items */
+static void qedf_percpu_iothread_flush(struct qedf_percpu_iothread_s *iothread)
+{
+ struct qedf_io_work *work, *work_tmp;
+ unsigned long flags;
+
+ spin_lock_irqsave(&iothread->work_lock, flags);
+ list_for_each_entry_safe(work, work_tmp, &iothread->work_list, list) {
+ list_del(&work->list);
+ kfree(work);
+ }
+ spin_unlock_irqrestore(&iothread->work_lock, flags);
+}
+
+static int qedf_cpu_offline(unsigned int cpu)
+{
+ struct qedf_percpu_iothread_s *iothread;
+
+ iothread = this_cpu_ptr(&qedf_percpu_iothreads);
+
+ if (iothread->iothread == NULL)
+ goto out;
+
+ QEDF_INFO(NULL, QEDF_LOG_DISC,
+ "Freeing I/O thread for CPU %d.\n", cpu);
+ qedf_percpu_iothread_flush(iothread);
+ if (iothread->iothread) {
+ kthread_stop(iothread->iothread);
+ iothread->iothread = NULL;
+ }
+
+out:
+ return 0;
+}
+
+/*
+ * Module Init/Remove
+ */
+
+static int __init qedf_init(void)
+{
+ int ret, rc;
+ struct qedf_percpu_iothread_s *iothread;
+ unsigned int cpu;
+
+ /* If debug=1 passed, set the default log mask */
+ if (qedf_debug == QEDF_LOG_DEFAULT)
+ qedf_debug = QEDF_DEFAULT_LOG_MASK;
+
+ /* Print driver banner */
+ QEDF_INFO(NULL, QEDF_LOG_INFO, "%s v%s.\n", QEDF_DESCR,
+ QEDF_VERSION);
+
+ /* Create kmem_cache for qedf_io_work structs */
+ qedf_io_work_cache = kmem_cache_create("qedf_io_work_cache",
+ sizeof(struct qedf_io_work), 0, SLAB_HWCACHE_ALIGN, NULL);
+ if (qedf_io_work_cache == NULL) {
+ QEDF_ERR(NULL, "qedf_io_work_cache is NULL.\n");
+ goto err1;
+ }
+ QEDF_INFO(NULL, QEDF_LOG_DISC, "qedf_io_work_cache=%p.\n",
+ qedf_io_work_cache);
+
+ qed_ops = qed_get_fcoe_ops();
+ if (!qed_ops) {
+ QEDF_ERR(NULL, "Failed to get qed fcoe operations\n");
+ goto err1;
+ }
+
+#ifdef CONFIG_DEBUG_FS
+ qedf_dbg_init("qedf");
+#endif
+
+ qedf_fc_transport_template =
+ fc_attach_transport(&qedf_fc_transport_fn);
+ if (!qedf_fc_transport_template) {
+ QEDF_ERR(NULL, "Could not register with FC transport\n");
+ goto err2;
+ }
+
+ qedf_fc_vport_transport_template =
+ fc_attach_transport(&qedf_fc_vport_transport_fn);
+ if (!qedf_fc_vport_transport_template) {
+ QEDF_ERR(NULL, "Could not register vport template with FC "
+ "transport\n");
+ goto err3;
+ }
+
+ /* Initialize the per CPU I/O threads */
+ for_each_possible_cpu(cpu) {
+ iothread = &per_cpu(qedf_percpu_iothreads, cpu);
+ INIT_LIST_HEAD(&iothread->work_list);
+ spin_lock_init(&iothread->work_lock);
+ }
+
+ rc = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "scsi/qedf:online",
+ qedf_cpu_online, qedf_cpu_offline);
+
+ if (rc < 0)
+ goto err4;
+ qedf_cpuhp_state = rc;
+
+ qedf_cb_ops.get_login_failures = qedf_get_login_failures;
+
+ ret = pci_register_driver(&qedf_pci_driver);
+ if (ret) {
+ QEDF_ERR(NULL, "Failed to register driver\n");
+ goto err4;
+ }
+
+ return 0;
+
+err4:
+ /* Release any I/O threads */
+ cpuhp_remove_state(qedf_cpuhp_state);
+
+ fc_release_transport(qedf_fc_vport_transport_template);
+err3:
+ fc_release_transport(qedf_fc_transport_template);
+err2:
+#ifdef CONFIG_DEBUG_FS
+ qedf_dbg_exit();
+#endif
+ qed_put_fcoe_ops();
+err1:
+ return -EINVAL;
+}
+
+static void __exit qedf_cleanup(void)
+{
+ pci_unregister_driver(&qedf_pci_driver);
+
+ /* Destroy percpu I/O threads */
+ cpuhp_remove_state(qedf_cpuhp_state);
+
+ fc_release_transport(qedf_fc_vport_transport_template);
+ fc_release_transport(qedf_fc_transport_template);
+#ifdef CONFIG_DEBUG_FS
+ qedf_dbg_exit();
+#endif
+ qed_put_fcoe_ops();
+
+ kmem_cache_destroy(qedf_io_work_cache);
+}
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("QLogic QEDF 25/40/50/100Gb FCoE Driver");
+MODULE_AUTHOR("QLogic Corporation");
+MODULE_VERSION(QEDF_VERSION);
+module_init(qedf_init);
+module_exit(qedf_cleanup);
diff --git a/drivers/scsi/qedf/qedf_version.h b/drivers/scsi/qedf/qedf_version.h
new file mode 100644
index 0000000..4ae5f53
--- /dev/null
+++ b/drivers/scsi/qedf/qedf_version.h
@@ -0,0 +1,15 @@
+/*
+ * QLogic FCoE Offload Driver
+ * Copyright (c) 2016 Cavium Inc.
+ *
+ * This software is available under the terms of the GNU General Public License
+ * (GPL) Version 2, available from the file COPYING in the main directory of
+ * this source tree.
+ */
+
+#define QEDF_VERSION "8.10.7.0"
+#define QEDF_DRIVER_MAJOR_VER 8
+#define QEDF_DRIVER_MINOR_VER 10
+#define QEDF_DRIVER_REV_VER 7
+#define QEDF_DRIVER_ENG_VER 0
+
--
1.8.5.6
^ 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