JC Lafoucriere [Mon, 3 Jun 2013 13:40:44 +0000 (21:40 +0800)]
staging/lustre/ldlm: suppress useless lock RPC for layout
In ldlm_lock_decref_internal() when l_lvb_data is freed to
reduce memory consumption, LDLM_FL_LVB_READY is not
cleared, so later when the lock is reused lvb is not
updated. But clearing LDLM_FL_LVB_READY forces layout refetch
at each file access, so the better is to remove the optimization.
The use case is after a restore in HSM.
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3280
Lustre-change: http://review.whamcloud.com/6268 Signed-off-by: JC Lafoucriere <jacques-charles.lafoucriere@cea.fr> Reviewed-by: Fan Yong <fan.yong@intel.com> Reviewed-by: Andreas Dilger <andreas.dilger@intel.com> Reviewed-by: Oleg Drokin <oleg.drokin@intel.com> Signed-off-by: Peng Tao <tao.peng@emc.com> Signed-off-by: Andreas Dilger <andreas.dilger@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Oleg Drokin [Mon, 3 Jun 2013 13:40:43 +0000 (21:40 +0800)]
staging/lustre: Revert "LU-2139 osc: Track and limit "unstable" pages"
This seems to be causing multiple issues: LU-3274, LU-3277
[The original commit is folded in the large Lusre patch. So we don't
have an exact commit to revert for kernel client -- Peng Tao] Signed-off-by: Peng Tao <tao.peng@emc.com> Signed-off-by: Andreas Dilger <andreas.dilger@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Quiet a number of overly noisy and unhelpful console error
messages. Improve the format of other nearby errors.
In the case of {lod,lov}_fix_desc_stripe_size(), this doesn't
even need a console message unless it is actually changing
some stripe size that is below the minimum. Typically it is
only zero and is being bumped up to the default value.
[picked lov part for upstream kernel submission]
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-1095
Lustre-change: http://review.whamcloud.com/6264 Signed-off-by: Andreas Dilger <andreas.dilger@intel.com> Reviewed-by: Bobi Jam <bobijam@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Oleg Drokin <oleg.drokin@intel.com> Signed-off-by: Peng Tao <tao.peng@emc.com> Signed-off-by: Andreas Dilger <andreas.dilger@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
wang di [Mon, 3 Jun 2013 13:40:41 +0000 (21:40 +0800)]
staging/lustre/utils: missing setting echo seq for getattr/setattr
It should set echo seq before do echo getattr/setattr, otherwise
echo_client will regard it as the object with MDT0 sequence.
[picked echo client part for upstream kernel submission]
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3267
Lustre-change: http://review.whamcloud.com/6263 Signed-off-by: wang di <di.wang@intel.com> Reviewed-by: Andreas Dilger <andreas.dilger@intel.com> Reviewed-by: Lai Siyao <lai.siyao@intel.com> Signed-off-by: Peng Tao <tao.peng@emc.com> Signed-off-by: Andreas Dilger <andreas.dilger@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
wang di [Mon, 3 Jun 2013 13:40:39 +0000 (21:40 +0800)]
staging/lustre/ofd: refill env in ofd_get_info
Because ofd_get_info(KEY_FIEMAP) might be called from
ptlrpc_server_handle_req_in(see the stack below),
where env might not be initialized correctly(see LBUG below),
so it refill refill in ofd_get_info.
This is a patch to the bh.c file that fixes up all indentation errors
found the checkpatch.pl tool. Two lines beyond 80 characters are left
alone to improve readability
Signed-off-by: Davis Abubakr-Sadik Nii Nai <dwa2pac@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Xenia Ragiadakou [Sun, 26 May 2013 19:42:44 +0000 (22:42 +0300)]
rtl8192u: fix printk calls in r8192U_core.c
This patch replaces calls to printk with their
corresponding calls to pr_<loglevel>(fmt, ...)
and netdev_<loglevel>(dev, fmt, ...), when it
is applicable, to make the log messages more
informative.
Also, it fixes some small typos and whitespaces
found in log messages.
Markus Niebel [Mon, 27 May 2013 08:19:09 +0000 (10:19 +0200)]
staging: drm/imx: set correct sync pins for display
partial fix of changes from
"staging: drm/imx: Add support for VGA via TVE on i.MX53"
Have to call imx_drm_crtc_panel_format_pins in imx_drm_crtc_panel_format
with the correct pins instead of (0, 0) This enables configuration of
correct waveforms for vsync / hsync for parallel display, LDB (i.MX53 and
i.MX6) as well as HDMI and MIPI (i.MX6)
TODO: configure pins via device tree
Signed-off-by: Markus Niebel <Markus.Niebel@tqs.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Harsh Kumar [Thu, 30 May 2013 05:48:12 +0000 (11:18 +0530)]
Staging: winbond: Made local functions static
Few functions are used only in one file. They are not included in any
other .h or .c files (I used grep to check). They seem to be local functions.
So, I have made them static.
I have also inlined one function as it is a one line function.
Peng Tao [Wed, 29 May 2013 13:40:56 +0000 (21:40 +0800)]
staging/lustre: fix LBUG and inode refcount bug
ldlm_proc_setup and ldlm_proc_cleanup should reset global
pde pointers otherwise remount may hit
LASSERT(ldlm_ns_proc_dir == NULL);
Also in libcfs_sock_ioctl, fput() includes itself sock_release.
So don't call sock_release twice otherwise kernel may oops
due to incorrect inode ref counting.
Signed-off-by: Peng Tao <tao.peng@emc.com> Signed-off-by: Andreas Dilger <andreas.dilger@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Peng Tao [Wed, 29 May 2013 13:40:55 +0000 (21:40 +0800)]
staging/lustre: adapt proc_dir_entry change
In 3.10 merge window, proc_dir_entry is now private to proc. However,
Lustre lprocfs depends heavily on it and its now-gone read_proc_t and
write_proc_t members.
The patch largely changed the fact, and made lprocfs depend on none of
proc_dir_entry private members. All lprocfs callers are converted to
use the new seq_file scheme.
Also lprocfs_srch is removed so that we can drop lprocfs_lock. All callers
are changed to save created pde in proper place.
See https://jira.hpdd.intel.com/browse/LU-3319 for more details.
Signed-off-by: Peng Tao <tao.peng@emc.com> Signed-off-by: Andreas Dilger <andreas.dilger@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Dan Carpenter [Wed, 29 May 2013 06:57:36 +0000 (09:57 +0300)]
Staging: rtl8192e: printing the wrong array
Smatch complained that there is an array overflow here. It looks
like we are supposed to be printing the ->rates_ex[] array which is
slightly larger than the ->rates[] array. This is a cut and paste
error.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Wei Yongjun [Thu, 23 May 2013 09:28:05 +0000 (17:28 +0800)]
Staging: vt6655: add missing free_netdev() on error in hostap_enable_hostapd()
Add the missing free_netdev() before return from function
hostap_enable_hostapd() in the error handling case.
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Acked-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: comedi: usbduxsigma: tidy up analog input command support
Rename the (*do_cmdtest) and (*do_cmd) functions so they have namespace
associated with the driver. Rename the local variable used for the
private data pointer.
Move all the command argument initialization and testing into the
(*do_cmdtest) function. That function is always called by the comedi
core before the (*do_cmd) function. The only thing the (*do_cmd) function
should have to do is setup the channel list, send the command to start
the analog input sampling, initialize the counter for the conversion,
then either read the acquisition data (submit the urbs) or set the
trigger to read the data later.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: comedi: usbduxsigma: cleanup the private data 'outBuffer'
This buffer is used to cache the analog output values that are written
to the analog output channels. Currently it only caches the single
writes to the channels using the (*insn_write) callback. The async
command writes are not cached. The buffer is also being kzalloc'ed
during the attach of the driver to a size much larger that required.
Rename the CamelCase buffer and change it to an array in the private
data of the correct size to cache the analog output channel values.
Modify the analog output urb callback so it updates the cached values
with those used for the asynchronous command to allow readback after
the command completes.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: comedi: usbduxsigma: tidy up analog output command support
Rename the (*do_cmdtest) and (*do_cmd) functions so they have namespace
associated with the driver. Rename the local variable used for the
private data pointer.
Move all the command argument initialization and testing into the
(*do_cmdtest) function. That function is always called by the comedi
core before the (*do_cmd) function. The only thing the (*do_cmd) function
should have to do is setup the channel list, initialize the counter for
the conversion, then either start the conversion (submit the urbs) or
set the trigger to start the conversion.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: comedi: usbduxsigma: tidy up usbdux_dio_insn_config()
Rename the function so it has namespace associated with the driver.
Rename the local variable used for the private data pointer.
Add a local variable for the channel mask.
Remove the unnecessary comments and clean up the multi-line comment
at the end.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: comedi: usbduxsigma: tidy up usbdux_dio_insn_bits()
Rename the function so it has namespace associated with the driver.
Rename the local variable used for the private data pointer.
Remove the unnecessary casts.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For aesthetics, move this function near its caller. Also, move the
FIRMWARE_MAX_LEN define so its with the other firmware constants
and fix the whitespace.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Rename the function so it has namespace associated with the driver.
For aesthetic reasons, move the function so it's near the one that
allocates the usb buffers.
Tidy up the function so it frees the buffers in the opposite order
they were allocated. Also remove the = NULL for all the buffers,
the driver is being detached so the comedi core will kfree the
private data.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: comedi: usbduxsigma: remove tidy_up() from (*auto_attach)
Comedi drivers do not need to cleanup allocations during the
(*auto_attach). If the (*auto_attach) fails the (*detach) will
automatically be called to handle the cleanup.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Cc: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
projects / linux-beck.git / log