| Age | Commit message (Collapse) | Author |
|---|
|
Move drm_pcie_get_speed_cap_mask() under #ifdef CONFIG_PCI because it
it used only for PCI devices (evergreen, r600, r770), and it uses
PCI interfaces that only exist when CONFIG_PCI=y.
Previously, we tried to compile drm_pcie_get_speed_cap_mask() even when
CONFIG_PCI=n, which fails.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
Fix kconfig warning for LGUEST_GUEST config by selecting TTY:
warning: (KVMTOOL_TEST_ENABLE && LGUEST_GUEST) selects VIRTIO_CONSOLE which has unmet direct dependencies (VIRTIO && TTY)
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Joe Millenbach <jmillenbach@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Work requests are passed between the host and the firmware with a
"cookie". This cookie is swapped to big-endian when passed to the
firmware and back to host endianness on return. This swapping seems
to be implemented incorrectly. Moreover, the byte swapping triggers
GCC warnings on 32 bit:
drivers/infiniband/hw/cxgb4/cm.c: In function ‘passive_ofld_conn_reply’:
drivers/infiniband/hw/cxgb4/cm.c:2803:12: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/infiniband/hw/cxgb4/cm.c: In function ‘send_fw_pass_open_req’:
drivers/infiniband/hw/cxgb4/cm.c:2941:16: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
[...]
But byte swapping isn't needed as the firmware doesn't actually touch
the cookie. Dropping byte swapping makes the warnings go away too.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Fixe the following types of sparse warnings
- cast to pointer from integer of different size
- cast from pointer to integer of different size
- incorrect type in assignment (different base types)
- incorrect type in argument 1 (different base types)
- cast from restricted __be64
- cast from restricted __be32
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
CPL_ABORT_REQ_RSS can come before TCP connection is established. In
such case peer_abort was trying to remove the hwtid, which was not
inserted. To avoid this we insert the hwtid when we are sure that we
are surely going to send passive accept request.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Don't wakeup threads blocked in rdma_init/rdma_fini if we are on
MPAv2, and want to retry connection with MPAv1.
Stop ep-timer on getting MPA version mismatch, before doing the
abort_connection - in process_mpa_request.
Take care to stop ep-timer in error paths for process_mpa_request.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Only reconnect if the endpoint wasn't freed.
peer_abort() should only attempt to reconnect if the endpoint wasn't
freed. Also remove hwtid from the debugfs idr.
Add missing check for peer2peer in MPAv2 code
Use correct mpa version on reject.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
The endpoint timeout logic had a race that could cause an endpoint
object to be freed while it was still on the timedout list. This
can happen if the timer is stopped after it had fired, but before
the timedout thread processed the endpoint timeout.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
With newer firmware, we can get streaming data due to connection
errors before the driver moves the QP out of RTS.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Log AEs even if the QP isn't in RTS. It is useful information.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
The driver is currently releasing the last ref on the QP too early.
This can cause bus errors due to HW still fetching WRs from the HW
queue. The fix is to keep a qp ref until we release the HW TID.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
With later firmware, the chances of getting streaming mode data after
we exit RTS is likely, so we don't need to warn for it. The only real
case where we don't expect it is when the QP is in RTS.
Move QP to ERROR when streaming mode data received.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
If a FINI operation fails, then we need to ABORT instead of CLOSE.
Also, if we ABORT due to unexpected STREAMING data, then wake up
anybody blocked in FINI...
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
This error means the RDMA connection was knocked out of RDMA mode,
probably due to an error on the connection.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
When we are converting local data to an extent format as a result of
adding an attribute, the type of data contained in the local fork
determines the behaviour that needs to occur.
xfs_bmap_add_attrfork_local() already handles the directory data
case specially by using S_ISDIR() and calling out to
xfs_dir2_sf_to_block(), but with verifiers we now need to handle
each different type of metadata specially and different metadata
formats require different verifiers (and eventually block header
initialisation).
There is only a single place that we add and attribute fork to
the inode, but that is in the attribute code and it knows nothing
about the specific contents of the data fork. It is only the case of
local data that is the issue here, so adding code to hadnle this
case in the attribute specific code is wrong. Hence we are really
stuck trying to detect the data fork contents in
xfs_bmap_add_attrfork_local() and performing the correct callout
there.
Luckily the current cases can be determined by S_IS* macros, and we
can push the work off to data specific callouts, but each of those
callouts does a lot of work in common with
xfs_bmap_local_to_extents(). The only reason that this fails for
symlinks right now is is that xfs_bmap_local_to_extents() assumes
the data fork contains extent data, and so attaches a a bmap extent
data verifier to the buffer and simply copies the data fork
information straight into it.
To fix this, allow us to pass a "formatting" callback into
xfs_bmap_local_to_extents() which is responsible for setting the
buffer type, initialising it and copying the data fork contents over
to the new buffer. This allows callers to specify how they want to
format the new buffer (which is necessary for the upcoming CRC
enabled metadata blocks) and hence make xfs_bmap_local_to_extents()
useful for any type of data fork content.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
|
|
The trylock log force invoked via xfs_buf_item_push() can attempt
to acquire xa_lock, thus leading to a recursion bug when called
with xa_lock held.
This log force was originally added to xfs_buf_trylock() to address
xfsaild stalls due to pinned and stale buffers. Since the addition
of this behavior, the log item pushing code had been reworked to
detect and track pinned items to inform xfsaild to issue a log
force itself when necessary. As such, the log force on trylock
failure is redundant and safe to remove.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
|
|
The buffer pinned check and trylock sequence in xfs_buf_item_push()
can race with an active transaction on marking the buffer pinned.
This can result in the buffer becoming pinned and stale after the
initial check and the trylock failure, but before the check in
xfs_buf_trylock() that issues a log force. If the log force is
issued from this context, a spinlock recursion occurs on xa_lock.
Prepare xfs_buf_item_push() to handle the race by detecting a
pinned buffer after the trylock failure so xfsaild issues a log
force from a safe context. This, along with various previous fixes,
renders the log force in xfs_buf_trylock() redundant.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
|
|
Speculative preallocation based on the current file size works well
for contiguous files, but is sub-optimal for sparse files where the
EOF preallocation can fill holes and result in large amounts of
zeros being written when it is not necessary.
The algorithm is modified to prevent EOF speculative preallocation
from triggering larger allocations on IO patterns of
truncate--to-zero-seek-write-seek-write-.... which results in
non-sparse files for large files. This, unfortunately, is the way cp
now behaves when copying sparse files and so needs to be fixed.
What this code does is that it looks at the existing extent adjacent
to the current EOF and if it determines that it is a hole we disable
speculative preallocation altogether. To avoid the next write from
doing a large prealloc, it takes the size of subsequent
preallocations from the current size of the existing EOF extent.
IOWs, if you leave a hole in the file, it resets preallocation
behaviour to the same as if it was a zero size file.
Example new behaviour:
$ xfs_io -f -c "pwrite 0 31m" \
-c "pwrite 33m 1m" \
-c "pwrite 128m 1m" \
-c "fiemap -v" /mnt/scratch/blah
wrote 32505856/32505856 bytes at offset 0
31 MiB, 7936 ops; 0.0000 sec (1.608 GiB/sec and 421432.7439 ops/sec)
wrote 1048576/1048576 bytes at offset 34603008
1 MiB, 256 ops; 0.0000 sec (1.462 GiB/sec and 383233.5329 ops/sec)
wrote 1048576/1048576 bytes at offset 134217728
1 MiB, 256 ops; 0.0000 sec (1.719 GiB/sec and 450704.2254 ops/sec)
/mnt/scratch/blah:
EXT: FILE-OFFSET BLOCK-RANGE TOTAL FLAGS
0: [0..65535]: 96..65631 65536 0x0
1: [65536..67583]: hole 2048
2: [67584..69631]: 67680..69727 2048 0x0
3: [69632..262143]: hole 192512
4: [262144..264191]: 262240..264287 2048 0x1
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
|
|
Move the reservation of low memory, except for the 4K which actually
does belong to the BIOS, later in the initialization; in particular,
after we have already reserved the trampoline.
The current code locates the trampoline as high as possible, so by
deferring the allocation we will still be able to reserve as much
memory as is possible. This allows us to run with reservelow=640k
without getting a crash on system startup.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Link: http://lkml.kernel.org/n/tip-0y9dqmmsousf69wutxwl3kkf@git.kernel.org
|
|
Patch 16559ae "kgdb: remove #include <linux/serial_8250.h> from kgdb.h"
removed an implicit inclusion of linux/platform_device.h
In a number of places. This adds back explicit inclusions in a few
more places I found.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Patch 16559ae "kgdb: remove #include <linux/serial_8250.h> from kgdb.h"
removed an implicit inclusion of linux/platform_device.h
from the exynos framebuffer driver. This adds back the required
explicit header file inclusions.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Ajay Kumar <ajaykumar.rs@samsung.com>
Cc: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Patch "16559ae kgdb: remove #include <linux/serial_8250.h> from kgdb.h
caused assabet_defconfig to fail, since assabet.c did not
itself include linux/platform_device.h, although it needs it:
In file included from include/linux/mfd/ucb1x00.h:13:0,
from arch/arm/mach-sa1100/assabet.c:19:
include/linux/mfd/mcp.h:22:16: error: field 'attached_device' has incomplete type
include/linux/mfd/mcp.h:48:23: error: field 'drv' has incomplete type
In file included from arch/arm/mach-sa1100/assabet.c:19:0:
include/linux/mfd/ucb1x00.h:137:16: error: field 'dev' has incomplete type
arch/arm/mach-sa1100/assabet.c: In function 'assabet_init':
arch/arm/mach-sa1100/assabet.c:343:3: error: implicit declaration of function 'platform_device_register_simple' [-Wimplicit-function-declaration]
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
pcieport does nice things like manage AER and we know it doesn't do
DMA or expose any user accessible devices on the host. It also keeps
the Memory, I/O, and Busmaster bits enabled, which is pretty handy
when trying to use anyting below it. Devices owned by pcieport cannot
be given to users via vfio, but we can tolerate them not being owned
by vfio-pci.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
|
|
vfio_dev_present is meant to give us a wait_event callback so that we
can block removing a device from vfio until it becomes unused. The
root of this check depends on being able to get the iommu group from
the device. Unfortunately if the BUS_NOTIFY_DEL_DEVICE notifier has
fired then the device-group reference is no longer searchable and we
fail the lookup.
We don't need to go to such extents for this though. We have a
reference to the device, from which we can acquire a reference to the
group. We can then use the group reference to search for the device
and properly block removal.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
|
|
We can actually handle MMIO and I/O port from the same access function
since PCI already does abstraction of this. The ROM BAR only requires
a minor difference, so it gets included too. vfio_pci_config_readwrite
gets renamed for consistency.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
|
|
The read and write functions are nearly identical, combine them
and convert to a switch statement. This also makes it easy to
narrow the scope of when we use the io/mem accessors in case new
regions are added.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
|
|
Even for non-pfmalloc SKBs, __netif_receive_skb() will do a
tsk_restore_flags() on current unconditionally.
Make __netif_receive_skb() a shim around the existing code, renamed to
__netif_receive_skb_core(). Let __netif_receive_skb() wrap the
__netif_receive_skb_core() call with the task flag modifications, if
necessary.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Commit cb57a2b4cff7edf2a4e32c0163200e9434807e0a ("x86-32: Export
kernel_stack_pointer() for modules") added an include of the
module.h header in conjunction with adding an EXPORT_SYMBOL_GPL
of kernel_stack_pointer.
But module.h should be avoided for simple exports, since it in turn
includes the world. Swap the module.h for export.h instead.
Cc: Jiri Kosina <trivial@kernel.org>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Link: http://lkml.kernel.org/r/1360872842-28417-1-git-send-email-paul.gortmaker@windriver.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
Print some additional debugging context to hopefully help to debug a
warning which is getting triggered by xfstests #74.
Also remove extraneous newlines from when printk's were converted to
ext4_warning() and ext4_msg().
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
|
|
commit 9ec1882df2 (tty: serial: imx: console write routing is unsafe
on SMP) introduced a recursive locking bug in imx_console_write().
The callchain is:
imx_rxint()
spin_lock_irqsave(&sport->port.lock,flags);
...
uart_handle_sysrq_char();
sysrq_function();
printk();
imx_console_write();
spin_lock_irqsave(&sport->port.lock,flags); <--- DEAD
The bad news is that the kernel debugging facilities can dectect the
problem, but the printks never surface on the serial console for
obvious reasons.
There is a similar issue with oops_in_progress. If the kernel crashes
we really don't want to be stuck on the lock and unable to tell what
happened.
In general most UP originated drivers miss these checks and nobody
ever notices because CONFIG_PROVE_LOCKING seems to be still ignored by
a large number of developers.
The solution is to avoid locking in the sysrq case and trylock in the
oops_in_progress case.
This scheme is used in other drivers as well and it would be nice if
we could move this to a common place, so the usual copy/paste/modify
bugs can be avoided.
Now there is another issue with this scheme:
CPU0 CPU1
printk()
rxint()
sysrq_detection() -> sets port->sysrq
return from interrupt
console_write()
if (port->sysrq)
avoid locking
port->sysrq is reset with the next receive character. So as long as
the port->sysrq is not reset and this can take an endless amount of
time if after the break no futher receive character follows, all
console writes happen unlocked.
While the current writer is protected against other console writers by
the console sem, it's unprotected against open/close or other
operations which fiddle with the port. That's what the above mentioned
commit tried to solve.
That's an issue in all drivers which use that scheme and unfortunately
there is no easy workaround. The only solution is to have a separate
indicator port->sysrq_cpu. uart_handle_sysrq_char() then sets it to
smp_processor_id() before calling into handle_sysrq() and resets it to
-1 after that. Then change the locking check to:
if (port->sysrq_cpu == smp_processor_id())
locked = 0;
else if (oops_in_progress)
locked = spin_trylock_irqsave(port->lock, flags);
else
spin_lock_irqsave(port->lock, flags);
That would force all other cpus into the spin_lock path. Problem
solved, but that's way beyond the scope of this fix and really wants
to be implemented in a common function which calls the uart specific
write function to avoid another gazillion of hard to debug
copy/paste/modify bugs.
Reported-and-tested-by: Tim Sander <tim@krieglstein.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable <stable@vger.kernel.org> # 3.6+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Some messages printed related to a WARN_ON(1) were printed using
KERN_NOTICE. Use KERN_WARNING or ext4_warning() instead so that
context related to the WARN_ON() is printed at the same printk warning
level (and log files, etc.)
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
|
|
Obviously this is a typo and could result in memory leaks if kzalloc
fails on a given cpu.
Signed-off-by: Daniel Baluta <dbaluta@ixiacom.com>
Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1360186160-7566-1-git-send-email-dbaluta@ixiacom.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
The 'operations' bitmap corresponds one-for-one with the operation
codes, no adjustment is necessary.
Reported-by: Mark Kettenis <mark.kettenis@xs4all.nl>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Remove the casts of the first argument of memset.
Neaten the style by using the sizeof the actual variable
being memset not the sizeof the type of variable.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
These messages don't seem to be errors but notifications
that some attribute isn't quite right.
Don't mark them as errors.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
This patch check output of pci_map_single() calls. I missed them on
my previous patch "iwlegacy: check for dma mapping errors", which
fixed only pci_map_page() calls.
To handle remaining possible dma mappings errors, we need to rearrange
ilXXXX_tx_skb() and il_enqueue_hcmd() functions.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
Check output of dma_map_single functions which nowadays can fail (when
IOMMU is used). On write_beacon callbacks just print error, similar
like padding error is handled by rt2800_write_beacon.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
This patch adds support for handling of PCIe sleep cookie depending
upon device properties. Some PCIe devices need sleep cookie probing
before accessing HW while some others don't. A new sleep_cookie
variable is defined as part of mwifiex_pcie_card_reg strcture and
set/reset as per device capability.
Sleep cookie is allocated/accessed/freed only when flag sleep_cookie
for this particular device is enabled.
Signed-off-by: Avinash Patil <patila@marvell.com>
Signed-off-by: Bing Zhao <bzhao@marvell.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
As the number of drivers in the rtlwifi family has grown, the Kconfig
section for them has grown unwieldy. This change has two effects: (1)
Variable RTLWIFI_DEBUG is documented, and (2) the entries for the
drivers that depend on RTLWIFI are indented.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
ccode is not NUL terminated. Presumably insisting on a terminator makes
brcms_c_country_valid() return false when it's not intended. ccode[2]
is sprom->leddc_on_time.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless
Conflicts:
drivers/net/wireless/iwlwifi/dvm/tx.c
|
|
The %name-prefix "prefix" syntax is not available on bison 2.3 and
older. Substitute with the -p "prefix" command-line option for
compatibility with older versions of bison.
This patch fixes this build error with older versions of bison.
CC util/sysfs.o
BISON util/pmu-bison.c
util/pmu.y:2.14-24: syntax error, unexpected string, expecting =
make: *** [util/pmu-bison.c] Error 1
Signed-off-by: Vinson Lee <vlee@twitter.com>
Tested-by: Li Zefan <lizefan@huawei.com>
Cc: stable@vger.kernel.org # 3.4+
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Pekka Enberg <penberg@kernel.org>
Link: http://lkml.kernel.org/r/1360792138-29186-1-git-send-email-vlee@twitter.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
There's DWARF unwind support only for x86 archs, so limit the unwind.o
object to them only.
Without this building for other archs (e.g. cross compiling for ARM) is
broken.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Link: http://lkml.kernel.org/n/tip-viqtvd6hppqgt68zz4wlqm20@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
into timers/core
|
|
This fixes a less obvious error on one hand, and prevents futher
similar errors by disambiguating and optimizing RxFCB indication,
on the other hand.
The error consists in NETIF_F_HW_VLAN_TX flag being used as an
indication of Rx FCB insertion. This happened as soon gfar_uses_fcb(),
which despite its name indicates Rx FCB insertion, started
incorporating is_vlan_on().
is_vlan_on(), on the other hand, is also a misleading construct because
we need to differentiate b/w hw VLAN extraction/VLEX (marked by VLAN_RX
flag) and hw VLAN insertion/VLINS (VLAN_TX flag), which are different
mechanisms using different types of FCBs.
The hw spec for the RxFCB feature is as follows:
In the case of RxBD rings, FCBs (Frame Control Block) are inserted by
the eTSEC whenever RCTRL[PRSDEP] is set to a non-zero value. Only one
FCB is inserted per frame (in the buffer pointed to by the RxBD with
bit F set). TOE acceleration for receive is enabled for all rx frames
in this case.
This patch introduces priv->uses_rxfcb field to quickly signal RxFCB
insertion in accordance with the specification above.
The dependency on FSL_GIANFAR_DEV_HAS_TIMER was also eliminated as
another source of confusion. The actual dependency is to priv->hwts_rx_en.
Upon changing priv->hwts_rx_en via IOCTL, the gfar device is being
restarted and on init_mac() the priv->hwts_rx_en flag determines RxFCB
insertion, and rctrl is programmed accordingly. The patch takes care
of this case too.
Though maybe not as self documenting as the inlining version uses_fcb(),
priv->uses_rxfcb has the main purpose to quickly signal, on the hot path,
that the incoming frame has a *Rx* FCB block inserted which needs to be
pulled out before passing the skb to the stack. This is a performance
critical operation, it needs to happen fast, that's why uses_rxfcb is
placed in the first cacheline of gfar_private.
This is also why a cached rctrl cannot be used instead: 1) because
we don't have 32 bits available in the first cacheline of gfar_priv
(but only 16); 2) bit operations are expensive on the hot path.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The controller's ref manual states clearly that when the hw Rx vlan
offload feature is enabled, meaning that the VLEX bit from RCTRL is
correctly enabled, then the hw performs automatic VLAN tag extraction
and deletion from the ethernet frames. So there's no point in trying to
increase the rx buff size when rxvlan is on, as the frame is actually
smaller.
And the Tx vlan hw accel feature (VLINS) has nothing to do with rx buff
size computation.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
No return code is expected from gfar_process_frame(), hence
change it to return void.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The change is significant since it affects the rx hot path.
Paul observed and documented the effects at asm level, see
below:
"It turns out that it does make a difference, since gfar_process_frame
gets inlined, and so the increment code gets moved out of line (I have
marked the if statment with * and the increment code within "-----"):
------------------------- as is currently ------------------
4d14: 80 61 00 18 lwz r3,24(r1)
4d18: 7f c4 f3 78 mr r4,r30
4d1c: 48 00 00 01 bl 4d1c <gfar_clean_rx_ring+0x10c>
* 4d20: 2f 83 00 04 cmpwi cr7,r3,4
4d24: 40 9e 00 1c bne- cr7,4d40
<gfar_clean_rx_ring+0x130>
----------------------------
4d28: 81 3c 01 f8 lwz r9,504(r28)
4d2c: 81 5c 01 fc lwz r10,508(r28)
4d30: 31 4a 00 01 addic r10,r10,1
4d34: 7d 29 01 94 addze r9,r9
4d38: 91 3c 01 f8 stw r9,504(r28)
4d3c: 91 5c 01 fc stw r10,508(r28)
----------------------------
4d40: a0 1f 00 24 lhz r0,36(r31)
4d44: 81 3f 00 00 lwz r9,0(r31)
4d48: 7f a4 eb 78 mr r4,r29
4d4c: 7f e3 fb 78 mr r3,r31
-------------------------- unlikely ------------------------
4d14: 80 61 00 18 lwz r3,24(r1)
4d18: 7f c4 f3 78 mr r4,r30
4d1c: 48 00 00 01 bl 4d1c <gfar_clean_rx_ring+0x10c>
* 4d20: 2f 83 00 04 cmpwi cr7,r3,4
4d24: 41 9e 03 94 beq- cr7,50b8
<gfar_clean_rx_ring+0x4a8>
4d28: a0 1f 00 24 lhz r0,36(r31)
4d2c: 81 3f 00 00 lwz r9,0(r31)
4d30: 7f a4 eb 78 mr r4,r29
4d34: 7f e3 fb 78 mr r3,r31
[...]
50b8: 81 3c 01 f8 lwz r9,504(r28)
50bc: 81 5c 01 fc lwz r10,508(r28)
50c0: 31 4a 00 01 addic r10,r10,1
50c4: 7d 29 01 94 addze r9,r9
50c8: 91 3c 01 f8 stw r9,504(r28)
50cc: 91 5c 01 fc stw r10,508(r28)
50d0: 4b ff fc 58 b 4d28 <gfar_clean_rx_ring+0x118>
So, the increment does actually get moved ~1k away."
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Group run-time critical fields within the 1st cacheline (32B)
followed by the tx|rx_queue reference arrays and the interrupt
group instances (gfargrp), all cacheline aligned.
This has several benefits. Firstly comes the performance benefit
by having the members required by the driver's hot path re-grouped
in the structure's first cache lines, whereas the unimportant
members were pushed towards the end of the struct.
Another benefit comes from eliminating a 24 byte memory hole that
was rendering gfar_priv's 2nd cacheline useless. The default gcc
layout of gfar_private leaves an implicit 24 byte hole after the
errata (enum) member. This patch fixes it.
The uchar bitfields were pushed towards the end of the struct
as these are not run-time performance critical (used for init
time operations). Because there is no other 2 byte member
around to couple the uchar bitfields memeber with, we will
have an addititnal 2 byte hole after the bitfields. This is
unsignificant however, and it doesn't influence gfar_priv's
size, because the whole structure is padded to be a 32B multiple.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
