user/sven/linux.git/include, branch v3.15.6

kernfs: introduce kernfs_pin_sb()

2014-07-17T23:23:19Z

commit 4e26445faad366d67d7723622bf6a60a6f0f5993 upstream. kernfs_pin_sb() tries to get a refcnt of the superblock. This will be used by cgroupfs. v2: - make kernfs_pin_sb() return the superblock. - drop kernfs_drop_sb(). tj: Updated the comment a bit. [ This is a prerequisite for a bugfix. ] Acked-by: Greg Kroah-Hartman Signed-off-by: Li Zefan Signed-off-by: Tejun Heo Signed-off-by: Greg Kroah-Hartman

kernfs: implement kernfs_root->supers list

2014-07-17T23:23:19Z

commit 7d568a8383bbb9c1f5167781075906acb2bb1550 upstream. Currently, there's no way to find out which super_blocks are associated with a given kernfs_root. Let's implement it - the planned inotify extension to kernfs_notify() needs it. Make kernfs_super_info point back to the super_block and chain it at kernfs_root->supers. Signed-off-by: Tejun Heo [lizf: Backported to 3.15: Adjust context] Signed-off-by: Li Zefan Signed-off-by: Greg Kroah-Hartman

ring-buffer: Check if buffer exists before polling

2014-07-17T23:23:18Z

commit 8b8b36834d0fff67fc8668093f4312dd04dcf21d upstream. The per_cpu buffers are created one per possible CPU. But these do not mean that those CPUs are online, nor do they even exist. With the addition of the ring buffer polling, it assumes that the caller polls on an existing buffer. But this is not the case if the user reads trace_pipe from a CPU that does not exist, and this causes the kernel to crash. Simple fix is to check the cpu against buffer bitmask against to see if the buffer was allocated or not and return -ENODEV if it is not. More updates were done to pass the -ENODEV back up to userspace. Link: http://lkml.kernel.org/r/5393DB61.6060707@oracle.com Reported-by: Sasha Levin Signed-off-by: Steven Rostedt Signed-off-by: Greg Kroah-Hartman

drm/i915: provide interface for audio driver to query cdclk

2014-07-09T18:21:26Z

commit c149dcb5c60bfea8871f16dfcc0690255eeb825f upstream. For Haswell and Broadwell, if the display power well has been disabled, the display audio controller divider values EM4 M VALUE and EM5 N VALUE will have been lost. The CDCLK frequency is required for reprogramming them to generate 24MHz HD-A link BCLK. So provide a private interface for the audio driver to query CDCLK. This is a stopgap solution until a more generic interface between audio and display drivers has been implemented. Signed-off-by: Jani Nikula Reviewed-by: Damien Lespiau Signed-off-by: Mengdong Lin Signed-off-by: Takashi Iwai Signed-off-by: Greg Kroah-Hartman

drm/i915, HD-audio: Don't continue probing when nomodeset is given

2014-07-09T18:21:26Z

commit 74b0c2d75fb4cc89173944e6d8f9eb47aca0c343 upstream. When a machine is booted with nomodeset option, i915 driver skips the whole initialization. Meanwhile, HD-audio tries to bind wth i915 just by request_symbol() without knowing that the initialization was skipped, and eventually it hits WARN_ON() in i915_request_power_well() and i915_release_power_well() wrongly but still continues probing, even though it doesn't work at all. In this patch, both functions are changed to return an error in case of uninitialized state instead of WARN_ON(), so that HD-audio driver can give up HDMI controller initialization at the right time. Acked-by: Daniel Vetter Signed-off-by: Takashi Iwai Signed-off-by: Greg Kroah-Hartman

usb: gadget: f_fs: resurect usb_functionfs_descs_head structure

2014-07-09T18:21:24Z

commit 09122141785348bf9539762a5f5dbbae3761c783 upstream. Even though usb_functionfs_descs_head structure is now deprecated, it has been used by some user space tools. Its removel in commit [ac8dde1: “Add flags to descriptors block”] was an oversight leading to build breakage for such tools. Bring it back so that old user space tools can still be build without problems on newer kernel versions. Reported-by: Lad, Prabhakar Reported-by: Krzysztof Opasiak Signed-off-by: Michal Nazarewicz Signed-off-by: Felipe Balbi Signed-off-by: Greg Kroah-Hartman

usb-storage/SCSI: Add broken_fua blacklist flag

2014-07-09T18:21:23Z

commit b14bf2d0c0358140041d1c1805a674376964d0e0 upstream. Some buggy JMicron USB-ATA bridges don't know how to translate the FUA bit in READs or WRITEs. This patch adds an entry in unusual_devs.h and a blacklist flag to tell the sd driver not to use FUA. Signed-off-by: Alan Stern Reported-by: Michael Büsch Tested-by: Michael Büsch Acked-by: James Bottomley CC: Matthew Dharm Signed-off-by: Greg Kroah-Hartman

SCSI: use the scsi data buffer length to extract transfer size

2014-07-09T18:21:23Z

commit 5616b0a46ed82eb9a093f752fc4d7bd3cc688583 upstream. Commit 8846bab180fa introduced a helper that can be used to query the wire transfer size for a SCSI command taking protection information into account. However, some commands do not have a 1:1 mapping between the block range they work on and the payload size (discard, write same). After the scatterlist has been set up these requests use __data_len to store the number of bytes to report completion on. This means that callers of scsi_transfer_length() would get the wrong byte count for these types of requests. To overcome this we make scsi_transfer_length() use the scatterlist length in the scsi_data_buffer as basis for the wire transfer calculation instead of __data_len. Reported-by: Christoph Hellwig Debugged-by: Mike Christie Signed-off-by: Martin K. Petersen Signed-off-by: Christoph Hellwig Reviewed-by: Sagi Grimberg Fixes: d77e65350f2d82dfa0557707d505711f5a43c8fd Signed-off-by: James Bottomley Signed-off-by: Greg Kroah-Hartman

tracing: Fix syscall_*regfunc() vs copy_process() race

2014-07-07T01:59:12Z

commit 4af4206be2bd1933cae20c2b6fb2058dbc887f7c upstream. syscall_regfunc() and syscall_unregfunc() should set/clear TIF_SYSCALL_TRACEPOINT system-wide, but do_each_thread() can race with copy_process() and miss the new child which was not added to the process/thread lists yet. Change copy_process() to update the child's TIF_SYSCALL_TRACEPOINT under tasklist. Link: http://lkml.kernel.org/p/20140413185854.GB20668@redhat.com Fixes: a871bd33a6c0 "tracing: Add syscall tracepoints" Acked-by: Frederic Weisbecker Acked-by: Paul E. McKenney Signed-off-by: Oleg Nesterov Signed-off-by: Steven Rostedt Signed-off-by: Greg Kroah-Hartman

ptrace,x86: force IRET path after a ptrace_stop()

2014-07-07T01:59:12Z

commit b9cd18de4db3c9ffa7e17b0dc0ca99ed5aa4d43a upstream. The 'sysret' fastpath does not correctly restore even all regular registers, much less any segment registers or reflags values. That is very much part of why it's faster than 'iret'. Normally that isn't a problem, because the normal ptrace() interface catches the process using the signal handler infrastructure, which always returns with an iret. However, some paths can get caught using ptrace_event() instead of the signal path, and for those we need to make sure that we aren't going to return to user space using 'sysret'. Otherwise the modifications that may have been done to the register set by the tracer wouldn't necessarily take effect. Fix it by forcing IRET path by setting TIF_NOTIFY_RESUME from arch_ptrace_stop_needed() which is invoked from ptrace_stop(). Signed-off-by: Tejun Heo Reported-by: Andy Lutomirski Acked-by: Oleg Nesterov Suggested-by: Linus Torvalds Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman