Neil Horman [Fri, 4 Mar 2016 18:40:48 +0000 (13:40 -0500)]
vmxnet3: avoid calling pskb_may_pull with interrupts disabled
vmxnet3 has a function vmxnet3_parse_and_copy_hdr which, among other operations,
uses pskb_may_pull to linearize the header portion of an skb. That operation
eventually uses local_bh_disable/enable to ensure that it doesn't race with the
drivers bottom half handler. Unfortunately, vmxnet3 preforms this
parse_and_copy operation with a spinlock held and interrupts disabled. This
causes us to run afoul of the WARN_ON_ONCE(irqs_disabled()) warning in
local_bh_enable, resulting in this:
Fix it by splitting vmxnet3_parse_and_copy_hdr into two functions:
vmxnet3_parse_hdr, which sets up the internal/on stack ctx datastructure, and
pulls the skb (both of which can be done without holding the spinlock with irqs
disabled
and
vmxnet3_copy_header, which just copies the skb to the tx ring under the lock
safely.
tested and shown to correct the described problem. Applies cleanly to the head
of the net tree
Signed-off-by: Neil Horman <nhorman@tuxdriver.com> CC: Shrikrishna Khare <skhare@vmware.com> CC: "VMware, Inc." <pv-drivers@vmware.com> CC: "David S. Miller" <davem@davemloft.net> Acked-by: Shrikrishna Khare <skhare@vmware.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Bjørn Mork [Thu, 3 Mar 2016 21:20:53 +0000 (22:20 +0100)]
cdc_ncm: toggle altsetting to force reset before setup
Some devices will silently fail setup unless they are reset first.
This is necessary even if the data interface is already in
altsetting 0, which it will be when the device is probed for the
first time. Briefly toggling the altsetting forces a function
reset regardless of the initial state.
This fixes a setup problem observed on a number of Huawei devices,
appearing to operate in NTB-32 mode even if we explicitly set them
to NTB-16 mode.
Signed-off-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net>
Takashi Iwai [Wed, 24 Feb 2016 14:35:22 +0000 (15:35 +0100)]
drm/i915: Fix bogus dig_port_map[] assignment for pre-HSW
The recent commit [0bdf5a05647a: drm/i915: Add reverse mapping between
port and intel_encoder] introduced a reverse mapping to retrieve
intel_dig_port object from the port number. The code assumed that the
port vs intel_dig_port are 1:1 mapping. But in reality, this was a
too naive assumption.
As Martin reported about the missing HDMI audio on his SNB machine,
pre-HSW chips may have multiple intel_dig_port objects corresponding
to the same port. Since we assign the mapping statically at the init
time and the multiple objects override the map, it may not match with
the actually enabled output.
This patch tries to address the regression above. The reverse mapping
is provided basically only for the audio callbacks, so now we set /
clear the mapping dynamically at enabling and disabling HDMI/DP audio,
so that we can always track the latest and correct object
corresponding to the given port.
Fixes: 0bdf5a05647a ('drm/i915: Add reverse mapping between port and intel_encoder') Reported-and-tested-by: Martin Kepplinger <martink@posteo.de> Cc: drm-intel-fixes@lists.freedesktop.org Signed-off-by: Takashi Iwai <tiwai@suse.de> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Martin Kepplinger <martink@posteo.de> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1456324522-21591-1-git-send-email-tiwai@suse.de
(cherry picked from commit 9dfbffcf4ac0707097af9e6c1372192b9d03a357) Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Oliver Neukum [Mon, 7 Mar 2016 10:31:10 +0000 (11:31 +0100)]
usbnet: cleanup after bind() in probe()
In case bind() works, but a later error forces bailing
in probe() in error cases work and a timer may be scheduled.
They must be killed. This fixes an error case related to
the double free reported in
http://www.spinics.net/lists/netdev/msg367669.html
and needs to go on top of Linus' fix to cdc-ncm.
Signed-off-by: Oliver Neukum <ONeukum@suse.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ido Schimmel [Mon, 7 Mar 2016 14:15:30 +0000 (15:15 +0100)]
mlxsw: pci: Correctly determine if descriptor queue is full
The descriptor queues for sending (SDQs) and receiving (RDQs) packets
are managed by two counters - producer and consumer - which are both
16-bit in size. A queue is considered full when the difference between
the two equals the queue's maximum number of descriptors.
However, if the producer counter overflows, then it's possible for the
full queue check to fail, as it doesn't take the overflow into account.
In such a case, descriptors already passed to the device - but for which
a completion has yet to be posted - will be overwritten, thereby causing
undefined behavior. The above can be achieved under heavy load (~30
netperf instances).
Fix that by casting the subtraction result to u16, preventing it from
being treated as a signed integer.
Fixes: eda6500a987a ("mlxsw: Add PCI bus implementation") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Since we only support one VLAN filtering bridge we need to associate a
reference count with it, so that when the last port netdev leaves it, we
would know that a different bridge can be offloaded to hardware.
When a LAG device is memeber in a bridge and port netdevs are leaving
the LAG, we should always decrement the bridge's reference count, as it's
incremented for any port in the LAG.
Fixes: 4dc236c31733 ("mlxsw: spectrum: Handle port leaving LAG while bridged") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Provide a user space interface to issue call logical-processor instructions.
Only selected CLP commands are allowed, enough to get the full overview of
the installed PCI functions.
Reviewed-by: Sebastian Ott <sebott@linux.vnet.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vineet Gupta [Fri, 1 Jan 2016 09:42:54 +0000 (15:12 +0530)]
ARC: [intc-*] Do a domain lookup in primary handler for hwirq -> linux virq
The primary interrupt handler arch_do_IRQ() was passing hwirq as linux
virq to core code. This was fragile and worked so far as we only had legacy/linear
domains.
This came out of a rant by Marc Zyngier.
http://lists.infradead.org/pipermail/linux-snps-arc/2015-December/000298.html
Cc: Marc Zyngier <marc.zyngier@arm.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Noam Camus <noamc@ezchip.com> Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Alexey Brodkin [Mon, 1 Feb 2016 14:30:17 +0000 (17:30 +0300)]
ARC: [plat-axs] Don't use arc_{get|set}_core_freq() for manipulating core clk
For AXS103, certain bitfile configurations may not work with stock
"clock-frequency" specified in DT. Instead of duplicating the DT files, we
fixup the DT in-place.
This used to be done differently - as in top level "clock-frequency" was
read very early from FDT and exported using arc_{get|set}_core_freq()
also used in setting up clockevent/clocksource timers
This homebrew clk API served well for legacy timer probe (non DT)
However TIMERS are now probed from DT and use "core_clk" defined in DT,
and thus can no longer use the top level "clock-frequency".
This change reduces the number of users of ARC clk hack and paves way
for removal.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
[vgupta: broken out of from bigger patch] Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Vineet Gupta [Fri, 1 Jan 2016 12:35:48 +0000 (18:05 +0530)]
ARC: clocksource: DT based probe
- Remove explicit clocksource setup and let it be done by OF framework
by defining CLOCKSOURCE_OF_DECLARE() for various timers
- This allows multiple clocksources to be potentially registered
simultaneouly: previously we could only do one - as all of them had
same arc_counter_setup() routine for registration
- Setup routines also ensure that the underlying timer actually exists.
- Remove some of the panic() calls if underlying timer is NOT detected as
fallback clocksource might still be available
1. If GRFC doesn't exist, jiffies clocksource gets registered anyways
2. if RTC doesn't exist, TIMER1 can take over (as it is always
present)
Cc: Daniel Lezcano <daniel.lezcano@linaro.org> Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Vineet Gupta [Fri, 1 Jan 2016 13:18:40 +0000 (18:48 +0530)]
ARC: [dts] Introduce Timer bindings
ARC Timers have historically been probed directly.
As precursor to start probing Timers thru DT introduce these bindings
Note that to keep series bisectable, these bindings are not yet used in
code.
Cc: Daniel Lezcano <daniel.lezcano@linaro.org> Cc: Rob Herring <robh@kernel.org> Cc: devicetree@vger.kernel.org Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Sebastian Ott [Fri, 4 Mar 2016 09:34:05 +0000 (10:34 +0100)]
s390/dasd: remove casts to dasd_*_private
Convert dasd_device.private to be a void pointer to get
rid of a lot of explicit casts.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com> Reviewed-by: Stefan Haberland <sth@linux.vnet.ibm.com> Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Jan Höppner [Wed, 19 Aug 2015 09:16:35 +0000 (11:16 +0200)]
s390/dasd: Refactor dasd format functions
Prepare for new format checking function by renaming functions and
moving reusable code to separate functions:
- Move sanity checks into a new function and make it reusable.
- Move common format code to a new function called
dasd_eckd_format_process_data.
- Create the generic function dasd_eckd_format_build_ccw_req, which
itself will then decide what ccw request is being built according to
the input data. (with upcoming functionality).
Signed-off-by: Jan Höppner <hoeppner@linux.vnet.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Jan Höppner [Mon, 21 Sep 2015 15:32:11 +0000 (17:32 +0200)]
s390/dasd: Simplify code in format logic
Currently, dasd_format is calling the format logic of a DASD discipline
with PAV enabled. If that fails with an error code of -EAGAIN the value
of retries is decremented and the discipline function is called with PAV
turned off.
The loop is supposed to try this up to 255 times until success. However,
-EAGAIN can only occur once here and therefore the loop will never reach
the 255 retries.
So, replace the unnecessarily complicated loop logic and simply try again
without PAV enabled in case of an -EAGAIN error.
Signed-off-by: Jan Höppner <hoeppner@linux.vnet.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Jan Höppner [Wed, 14 Oct 2015 15:01:04 +0000 (17:01 +0200)]
s390/dasd: Improve dasd format code
- Make sure a calling function can rely on data in fdata by resetting to
its initial values
- Move special treatment for track 0 and 1 to dasd_eckd_build_format
- Replace dangerous backward goto with a loop logic
- Add define for number that specifies the maximum amount of CCWs per
request and is used for format_step calculation
- Remove unused variable
Signed-off-by: Jan Höppner <hoeppner@linux.vnet.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
MIPS: traps: Correct the SIGTRAP debug ABI in `do_watch' and `do_trap_or_bp'
Follow our own rules set in <asm/siginfo.h> for SIGTRAP signals issued
from `do_watch' and `do_trap_or_bp' by setting the signal code to
TRAP_HWBKPT and TRAP_BRKPT respectively, for Watch exceptions and for
those Breakpoint exceptions whose originating BREAK instruction's code
does not have a special meaning. Keep Trap exceptions unaffected as
these are not debug events.
No existing user software is expected to examine signal codes for these
signals as SI_KERNEL has been always used here. This change makes the
MIPS port more like other Linux ports, which reduces the complexity and
provides for performance improvement in GDB.
Signed-off-by: Maciej W. Rozycki <macro@imgtec.com> Cc: Pedro Alves <palves@redhat.com> Cc: Luis Machado <lgustavo@codesourcery.com> Cc: linux-mips@linux-mips.org Cc: gdb@sourceware.org
Patchwork: https://patchwork.linux-mips.org/patch/12758/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Rafał Miłecki [Mon, 25 Jan 2016 08:50:29 +0000 (09:50 +0100)]
MIPS: BCM47xx: Move SPROM driver to drivers/firmware/
Broadcom ARM home routers store SPROM content in NVRAM just like MIPS
ones. To share SPROM code we need to move it out of arch/mips/ to some
common place. We already have bcm47xx_nvram in firmware path and SPROM
should fit there as well.
This driver is responsible for parsing SoC configuration data into a
struct shared between ssb and bcma buses.
This was tested with BCM4706 & BCM5357C0 (BCM47XX) and BCM4708A0
(ARCH_BCM_5301X).
Aaro Koskinen [Mon, 22 Feb 2016 22:22:57 +0000 (00:22 +0200)]
MIPS: Octeon: Use model string from DTB for unknown board type
Use model string from DTB for board type if the board is unknown.
This is more informative, e.g. with EdgeRouter Pro the /proc/cpuinfo
will display "ubnt,e200 (CN6120p1.1-1000-NSP)" instead of misleading
"Unsupported Board".
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi> Cc: David Daney <ddaney.cavm@gmail.com> Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/12582/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Paul Burton [Mon, 13 Jul 2015 16:12:44 +0000 (17:12 +0100)]
MIPS: tlb-r4k: panic if the MMU doesn't support PAGE_SIZE
After writing the appropriate mask to the cop0 PageMask register, read
the register back & check it matches what we want. If it doesn't then
the MMU does not support the page size the kernel is configured for and
we're better off bailing than continuing to do odd things with TLB
exceptions.
Signed-off-by: Paul Burton <paul.burton@imgtec.com> Cc: Steven J. Hill <Steven.Hill@imgtec.com> Cc: Joshua Kinard <kumba@gentoo.org> Cc: Rafał Miłecki <zajec5@gmail.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Markos Chandras <markos.chandras@imgtec.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/10691/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
David Daney [Tue, 9 Feb 2016 19:00:06 +0000 (11:00 -0800)]
MIPS: OCTEON: Remove some code limiting NR_IRQS to 255
Follow-on patches for OCTEON III will increase the number of irqs to
potentially more than 256.
Increase the width of the octeon_irq_ciu_to_irq to int to be able to
handle this case. Remove the hacky code that verified that u8 would
not be overflowed.
David Daney [Tue, 2 Feb 2016 01:46:54 +0000 (17:46 -0800)]
MIPS: OCTEON: Extend number of supported CPUs past 32
To support more than 48 CPUs, the bootinfo structure grows a new
coremask structure. Add the definition of the structure and add it to
struct cvmx_bootinfo. In prom_init(), copy the new coremask data into
the sysinfo structure, and use it in smp_setup().
Ralf Baechle [Thu, 28 Jan 2016 13:26:00 +0000 (14:26 +0100)]
MIPS: highmem: Turn flush_cache_kmaps into a no-op.
It was calling flush_cache_all() which is a no-op since a long time anyway
and which was overkill in the old days when it was actually doing something
because only the D-cache needs to be flushed, never the I-cache, never
the S-cache. Since however highmem on MIPS is still only supported on
processors that don't suffer from cache aliases, we could turn
flush_cache_kmaps() into a no-op - but for paranoia's sake we rather make
it BUG_ON(cpu_has_dc_aliases()).
projects / karo-tx-linux.git / log