| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
It's currently possible for cifs_open to fail after it has already
called cifs_new_fileinfo. In that situation, the new fileinfo will be
leaked as the caller doesn't call fput. That in turn leads to a busy
inodes after umount problem since the fileinfo holds an extra inode
reference now. Shuffle cifs_open around a bit so that it only calls
cifs_new_fileinfo if it's going to succeed.
Signed-off-by: Jeff Layton <[email protected]>
Reviewed-and-Tested-by: Suresh Jayaraman <[email protected]>
|
|
...and ensure that we propagate the error back to avoid any surprises.
Signed-off-by: Suresh Jayaraman <[email protected]>
Reviewed-and-Tested-by: Jeff Layton <[email protected]>
|
|
...which takes a ton of unneeded arguments and does a lot more pointer
dereferencing than is really needed.
Signed-off-by: Jeff Layton <[email protected]>
Reviewed-and-Tested-by: Suresh Jayaraman <[email protected]>
|
|
The current scheme of sticking open files on a list and assuming that
cifs_open will scoop them off of it is broken and leads to "Busy
inodes after umount..." errors at unmount time.
The problem is that there is no guarantee that cifs_open will always
be called after a ->lookup or ->create operation. If there are
permissions or other problems, then it's quite likely that it *won't*
be called.
Fix this by fully instantiating the filp whenever the file is created
and pass that filp back to the VFS. If there is a problem, the VFS
can clean up the references.
Signed-off-by: Jeff Layton <[email protected]>
Reviewed-and-Tested-by: Suresh Jayaraman <[email protected]>
|
|
Having cifs_posix_open call cifs_new_fileinfo is problematic and
inconsistent with how "regular" opens work. It's also buggy as
cifs_reopen_file calls this function on a reconnect, which creates a new
struct cifsFileInfo that just gets leaked.
Push it out into the callers. This also allows us to get rid of the
"mnt" arg to cifs_posix_open.
Finally, in the event that a cifsFileInfo isn't or can't be created, we
always want to close the filehandle out on the server as the client
won't have a record of the filehandle and can't actually use it. Make
sure that CIFSSMBClose is called in those cases.
Signed-off-by: Jeff Layton <[email protected]>
Reviewed-and-Tested-by: Suresh Jayaraman <[email protected]>
|
|
|
|
The bridge multicast patches introduced an OOM crash in the forward
path, when deliver_clone fails to clone the skb.
Reported-by: Mark Wagner <[email protected]>
Signed-off-by: Herbert Xu <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
CRB window register is not per pci-func for NX3031,
so caching can result in incorrect values.
Signed-off-by: Amit Kumar Salecha <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Rcv producer should be read in spin-lock.
Signed-off-by: Amit Kumar Salecha <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Fixes memory leak in error path when memory allocation
for adapter data structures fails.
Signed-off-by: Amit Kumar Salecha <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Patch c7c2fa07 removed one line too much from smc91c92_cs.c.
Reported-by: Komuro <[email protected]>
CC: [email protected]
CC: [email protected]
Signed-off-by: Dominik Brodowski <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
According to my schematics, on Lite5200 board ethernet phy uses address
0 (all ADDR lines are pulled down). With this change I can talk to
onboard phy (LXT971) and correctly use autonegotiation.
Signed-off-by: Dmitry Eremin-Solenikov <[email protected]>
Signed-off-by: Grant Likely <[email protected]>
|
|
Compiling in the MPC5200 sound drivers results in the following build error:
sound/soc/fsl/mpc5200_psc_ac97.o: In function `to_psc_dma_stream':
mpc5200_psc_ac97.c:(.text+0x0): multiple definition of `to_psc_dma_stream'
sound/soc/fsl/mpc5200_dma.o:mpc5200_dma.c:(.text+0x0): first defined here
sound/soc/fsl/efika-audio-fabric.o: In function `to_psc_dma_stream':
efika-audio-fabric.c:(.text+0x0): multiple definition of `to_psc_dma_stream'
sound/soc/fsl/mpc5200_dma.o:mpc5200_dma.c:(.text+0x0): first defined here
make[3]: *** [sound/soc/fsl/built-in.o] Error 1
make[2]: *** [sound/soc/fsl] Error 2
make[1]: *** [sound/soc] Error 2
make: *** [sound] Error 2
This patch fixes it by declaring the inline function in the header file to
also be a static.
Signed-off-by: Grant Likely <[email protected]>
Cc: Jon Smirl <[email protected]>
Tested-by: John Hilmar Linkhorst <[email protected]>
Acked-by: Mark Brown <[email protected]>
|
|
When going to standby mode mpc code maps the whole soc5200 node
to access warious MBAR registers. However as of_iomap uses 'reg'
property of device node, only small part of MBAR is getting mapped.
Thus pm code gets oops when trying to access high parts of MBAR.
As a way to overcome this, make mpc52xx_pm_prepare() explicitly
map whole MBAR (0xc0000).
Signed-off-by: Dmitry Eremin-Solenikov <[email protected]>
Signed-off-by: Grant Likely <[email protected]>
|
|
Use an irq spinlock to hold off the IRQ handler until
enough early card init is complete such that the handler
can run without faulting.
Signed-off-by: Tim Gardner <[email protected]>
Cc: [email protected]
Signed-off-by: John W. Linville <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/kaber/nf-2.6
|
|
This patch fixes a missing read memory barrier that is needed for the
driver to readout the MAC address correctly from the on-board ROM.
Also it replaces the use of the deprecated functions readl()/writel().
Signed-off-by: Morten H. Larsen <[email protected]>
Signed-off-by: Matt Turner <[email protected]>
|
|
This patch probes for the Super IO chip and reserves the IO range when
found. It avoids enabling the IDE interface on the Avanti family, since
none has IDE. It enables the Enhanced Parallel Port v1.9 feature.
Signed-off-by: Morten H. Larsen <[email protected]>
Signed-off-by: Matt Turner <[email protected]>
|
|
Caused by 2c3c8bea608866d8bd9dcf92657d57fdcac011c5 which was clearly not
even compile tested.
Signed-off-by: Matt Turner <[email protected]>
|
|
Typo in 1527bc8b928dd1399c3d3467dd47d9ede210978a renamed hweight32 to
__arch_weight32.
Signed-off-by: Matt Turner <[email protected]>
|
|
Avoids this:
WARNING: at net/mac80211/scan.c:312 ieee80211_scan_completed+0x5f/0x1f1
[mac80211]()
Hardware name: Latitude E5400
Modules linked in: aes_x86_64 aes_generic fuse ipt_MASQUERADE iptable_nat
nf_nat rfcomm sco bridge stp llc bnep l2cap sunrpc cpufreq_ondemand
acpi_cpufreq freq_table xt_physdev ip6t_REJECT nf_conntrack_ipv6
ip6table_filter ip6_tables ipv6 kvm_intel kvm uinput arc4 ecb
snd_hda_codec_intelhdmi snd_hda_codec_idt snd_hda_intel iwlagn snd_hda_codec
snd_hwdep snd_seq snd_seq_device iwlcore snd_pcm dell_wmi sdhci_pci sdhci
iTCO_wdt tg3 dell_laptop mmc_core i2c_i801 wmi mac80211 snd_timer
iTCO_vendor_support btusb joydev dcdbas cfg80211 bluetooth snd soundcore
microcode rfkill snd_page_alloc firewire_ohci firewire_core crc_itu_t
yenta_socket rsrc_nonstatic i915 drm_kms_helper drm i2c_algo_bit i2c_core video
output [last unloaded: scsi_wait_scan]
Pid: 979, comm: iwlagn Tainted: G W 2.6.33.3-85.fc13.x86_64 #1
Call Trace:
[<ffffffff8104b558>] warn_slowpath_common+0x77/0x8f
[<ffffffff8104b57f>] warn_slowpath_null+0xf/0x11
[<ffffffffa01bb7d9>] ieee80211_scan_completed+0x5f/0x1f1 [mac80211]
[<ffffffffa02a23f0>] iwl_bg_scan_completed+0xbb/0x17a [iwlcore]
[<ffffffff81060d3d>] worker_thread+0x1a4/0x232
[<ffffffffa02a2335>] ? iwl_bg_scan_completed+0x0/0x17a [iwlcore]
[<ffffffff81064817>] ? autoremove_wake_function+0x0/0x34
[<ffffffff81060b99>] ? worker_thread+0x0/0x232
[<ffffffff810643c7>] kthread+0x7a/0x82
[<ffffffff8100a924>] kernel_thread_helper+0x4/0x10
[<ffffffff8106434d>] ? kthread+0x0/0x82
[<ffffffff8100a920>] ? kernel_thread_helper+0x0/0x10
Reported here:
https://bugzilla.redhat.com/show_bug.cgi?id=590436
Signed-off-by: John W. Linville <[email protected]>
Reported-by: Mihai Harpau <[email protected]>
Cc: [email protected]
Acked-by: Reinette Chatre <[email protected]>
|
|
On DA830/OMAP-L137 and DA850/OMAP-L138 SoCs, the McASP peripheral
has FIFO support. This FIFO provides additional data buffering. It
also provides tolerance to variation in host/DMA controller response
times. More details of the FIFO operation can be found at
http://focus.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=sprufm1&fileType=pdf
Existing sequence of steps for audio playback/capture are:
a. DMA configuration
b. McASP configuration (configures and enables FIFO)
c. Start DMA
d. Start McASP (enables FIFO)
During McASP configuration, while FIFO was being configured, FIFO
was being enabled in davinci_hw_common_param() function of
sound/soc/davinci/davinci-mcasp.c file. This generated a transmit
DMA event, which gets serviced when DMA is started.
https://patchwork.kernel.org/patch/84611/ patch clears the DMA
events before starting DMA, which is the right thing to do. But
this resulted in a state where DMA was waiting for an event from
McASP (after step c above), but the event which was already there,
has got cleared (because of step b above).
The fix is not to enable the FIFO during McASP configuration as
FIFO was being enabled as part of McASP start.
Signed-off-by: Sudhakar Rajashekhara <[email protected]>
Acked-by: Liam Girdwood <[email protected]>
Signed-off-by: Mark Brown <[email protected]>
|
|
In the commit below the version string handling was modified, adding
a '+' where no other version information was supplied:
commit 85a256d8e0116c8f5ad276730830f5d4d473344d
Author: David Rientjes <[email protected]>
Date: Wed Jan 13 13:01:05 2010 -0800
From the commit the intent was as below:
- when CONFIG_LOCALVERSION_AUTO is disabled, a `+' is appended if the
repository has been revised beyond a tagged commit and LOCALVERSION=
was not passed to "make".
However if the user supplies an empty LOCALVERSION on the command line
the plus suffix is still added. This form is useful in the case where
the build environment knows that the version as specified is correct and
complete but does not correspond to a specific tag.
This patch changes the implementation to match the documentation
such that specifying LOCALVERSION= on the build line is sufficient
to suppress any suffix.
Signed-off-by: Andy Whitcroft <[email protected]>
Acked-by: David Rientjes <[email protected]>
Signed-off-by: Michal Marek <[email protected]>
|
|
|
|
|
|
Certain revisions of this chipset appear to be broken. There is a shadow
GTT which mirrors the real GTT but contains pre-translated physical
addresses, for performance reasons. When a GTT update happens, the
translations are done once and the resulting physical addresses written
back to the shadow GTT.
Except sometimes, the physical address is actually written back to the
_real_ GTT, not the shadow GTT. Thus we start to see faults when that
physical address is fed through translation again.
Signed-off-by: David Woodhouse <[email protected]>
|
|
stanse found the following double lock.
In get_domain_for_dev:
spin_lock_irqsave(&device_domain_lock, flags);
domain_exit(domain);
domain_remove_dev_info(domain);
spin_lock_irqsave(&device_domain_lock, flags);
spin_unlock_irqrestore(&device_domain_lock, flags);
spin_unlock_irqrestore(&device_domain_lock, flags);
This happens when the domain is created by another CPU at the same time
as this function is creating one, and the other CPU wins the race to
attach it to the device in question, so we have to destroy our own
newly-created one.
Signed-off-by: David Woodhouse <[email protected]>
|
|
Commit a99c47a2 "intel-iommu: errors with smaller iommu widths" replace the
dmar_domain->pgd with the first entry of page table when iommu's supported
width is smaller than dmar_domain's. But it use physical address directly
for new dmar_domain->pgd...
This result in KVM oops with VT-d on some machines.
Reported-by: Allen Kay <[email protected]>
Cc: Tom Lyon <[email protected]>
Signed-off-by: Sheng Yang <[email protected]>
Signed-off-by: David Woodhouse <[email protected]>
|
|
The code in rfbi.c tried to get the omapdss platform_device via a static
member defined in dispc.c, leading to a compile error. The same
platform_device is available through rfbi-struct.
Signed-off-by: Tomi Valkeinen <[email protected]>
|
|
I was observing the following error messages on my OMAP1 based Amstrad Delta
board when first changing from text to graphics mode or vice versa after the
LCD display had been blanked:
omapfb omapfb: timeout waiting for FRAME DONE
with a followup error message while unblanking it back:
omapfb omapfb: resetting (status 0xffffffb2,reset count 1)
As a visible result, image pixels happened to be shifted by a few bits,
giving wrong colors.
Examining the code, I found that this problem occures when an OMAP1 internal
LCD controller is disabled from omap_lcdc_suspend() and then a subsequent
omap_lcdc_setup_plane() calls disable_controller() again. This potentially
error provoking behaviour is triggered by the lcdc.update_mode flag being kept
at OMAP_AUTO_UPDATE, regardless of the controller and panel being suspended.
This patch tries to correct the problem by replacing both omap_lcdc_suspend()
and omap_lcdc_resume() function bodies with single calls to
omap_lcdc_set_update_mode() with a respective OMAP_UPDATE_DISABLE or
OMAP_AUTO_UPDATE argument. As a result, exactly the same lower level
operations are performed, with addition of changing the lcdc.update_mode flag
to a value better suited for the controller state. This prevents any further
calls to disable_controller() from omap_lcdc_setup_plane() while the display
is suspended.
Created against linux-2.6.34-rc7.
Tested on Amstrad Delta.
Signed-off-by: Janusz Krzysztofik <[email protected]>
Signed-off-by: Tomi Valkeinen <[email protected]>
|
|
Signed-off-by: Stephen M. Cameron <[email protected]>
Cc: Mike Miller <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
|
|
Add dts descriptions for onboard 256 byte I2C eeprom (pcf8582C-2)
and 16MB NOR flash (am29lv652d).
Signed-off-by: Dmitry Eremin-Solenikov <[email protected]>
[[email protected]: added lite5200b eeprom and declared lite5200 gpios]
Signed-off-by: Grant Likely <[email protected]>
|
|
Signed-off-by: Grant Likely <[email protected]>
|
|
mpic_resume() on G5 macs blindly dereferences mpic->fixups, but
it may legitimately be NULL (as on PowerMac7,2). Add an explicit
check.
This fixes suspend-to-disk with one processor (maxcpus=1) for me.
Signed-off-by: Alastair Bridgewater <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
When trying to flash a machine via the update_flash command, Anton received the
following error:
Restarting system.
FLASH: kernel bug...flash list header addr above 4GB
The code in question has a comment that the flash list should be in
the kernel data and therefore under 4GB:
/* NOTE: the "first" block list is a global var with no data
* blocks in the kernel data segment. We do this because
* we want to ensure this block_list addr is under 4GB.
*/
Unfortunately the Kconfig option is marked tristate which means the variable
may not be in the kernel data and could be above 4GB.
Instead of relying on the data segment being below 4GB, use the static
data buffer allocated by the kernel for use by rtas. Since we don't
use the header struct directly anymore, convert it to a simple pointer.
Reported-By: Anton Blanchard <[email protected]>
Signed-Off-By: Milton Miller <[email protected]
Tested-By: Anton Blanchard <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
Irq stacks provide an essential protection from stack overflows through
external interrupts, at the cost of two additionals stacks per CPU.
Enable them unconditionally to simplify the kernel build and prevent
people from accidentally disabling them.
Signed-off-by: Christoph Hellwig <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
kexec_perpare_cpus_wait() iterates i through NR_CPUS to check
paca[i].kexec_state of each to make sure they have quiesced.
However now we have dynamic PACA allocation, paca[NR_CPUS] is not necessarily
valid and we overrun the array; spurious "cpu is not possible, ignoring"
errors result. This patch iterates for_each_online_cpu so stays
within the bounds of paca[] -- and every CPU is now 'possible'.
Signed-off-by: Matt Evans <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
Acked-by: Grant Likely <[email protected]>
On 5 May 2010 21:33, "Anton Blanchard" <[email protected]> wrote:
CONFIG_SYSFS_DEPRECATED can cause issues with newer distros and should not
be required for any distro in the last 3 or 4 years, so disable it.
Signed-off-by: Anton Blanchard <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
It was used in the dim distant past for adding initrds to images
for legacy iSeries, but it's not even used for that now that we
have initramfs. So remove it.
Signed-off-by: Paul Mackerras <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
We are seeing boot fails on some System p machines when using the kdump
crashkernel= boot option. The default kdump base address is 32MB, so if we
reserve 256MB for kdump then we reserve all of the RMO except the first 32MB.
We really want kdump to reserve some memory in the RMO and most of it
elsewhere but that will require more significant changes. For now we can shift
the default base address to 64MB when CONFIG_PPC64 and CONFIG_RELOCATABLE are
set. This isn't quite correct since what we really care about is the kdump
kernel is relocatable, but we already make the assumption that base kernel
and kdump kernel have the same CONFIG_RELOCATABLE setting, eg:
#ifndef CONFIG_RELOCATABLE
if (crashk_res.start != KDUMP_KERNELBASE)
printk("Crash kernel location must be 0x%x\n",
KDUMP_KERNELBASE);
...
RTAS is instantiated towards the top of our RMO, so if we were to go any
higher we risk not having enough RMO memory for the kdump kernel on boxes
with a 128MB RMO.
Signed-off-by: Anton Blanchard <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
CONFIG_HIGHPTE doesn't exist in Kconfig, therefore removing all
references for it from the source code.
Signed-off-by: Christoph Egger <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
|
|
|
|
This patch adds a missing element of the ReadPubEK command output,
that prevents future overflow of this buffer when copying the
TPM output result into it.
Prevents a kernel panic in case the user tries to read the
pubek from sysfs.
Signed-off-by: Rajiv Andrade <[email protected]>
Signed-off-by: James Morris <[email protected]>
|
|
Signed-off-by: Zhu Yi <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
|
|
Cc: [email protected]
Signed-off-by: Christian Lamparter <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
|
|
Fixes linux-2.6 warning:
drivers/net/wireless/libertas_tf/main.c: In function 'lbtf_rx':
drivers/net/wireless/libertas_tf/main.c:578: warning: 'stats.antenna' is used uninitialized in this function
drivers/net/wireless/libertas_tf/main.c:578: warning: 'stats.mactime' is used uninitialized in this function
stats struct needs to be set to 0 before use.
Signed-off-by: Prarit Bhargava <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
|
|
The flow id (scd_flow) in a compressed BA packet should match the txq_id
of the queue from which the aggregated packets were sent. However, in
some hardware like the 1000 series, sometimes the flow id is 0 for the
txq_id (10 to 19). This can cause the annoying message:
[ 2213.306191] iwlagn 0000:01:00.0: Received BA when not expected
[ 2213.310178] iwlagn 0000:01:00.0: Read index for DMA queue txq id (0),
index 5, is out of range [0-256] 7 7.
And even worse, if agg->wait_for_ba is true when the bad BA is arriving,
this can cause system hang due to NULL pointer dereference because the
code is operating in a wrong tx queue!
Signed-off-by: Shanyu Zhao <[email protected]>
Signed-off-by: Pradeep Kulkarni <[email protected]>
Signed-off-by: Reinette Chatre <[email protected]>
|
|
We are seeing some race conditions between incoming station management
requests (station add/remove) and the internal unassoc RXON command that
modifies station table. Modify these flows to require the mutex to be held
and thus serializing them.
This fixes http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2207
Signed-off-by: Reinette Chatre <[email protected]>
|
|
cpqarray_register_ctlr() and cpqarray_eisa_detect() also
need to be marked as __devinit.
Signed-off-by: Jens Axboe <[email protected]>
|
|
It needs to be __devinit, not __init.
Signed-off-by: Jens Axboe <[email protected]>
|