In commit:
21cbc2822aa1 ("x86/mm/cpa: Unbreak populate_pgd(): stop trying to deallocate failed PUDs")
I intended to add this comment, but I failed at using git.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/242baf8612394f4e31216f96d13c4d2e9b90d1b7.1469293159.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Valdis Kletnieks bisected a boot failure back to this recent commit:
360cb4d155 ("x86/mm/cpa: In populate_pgd(), don't set the PGD entry until it's populated")
I broke the case where a PUD table got allocated -- populate_pud()
would wander off a pgd_none entry and get lost. I'm not sure how
this survived my testing.
Fix the original issue in a much simpler way. The problem
was that, if we allocated a PUD table, failed to populate it, and
freed it, another CPU could potentially keep using the PGD entry we
installed (either by copying it via vmalloc_fault or by speculatively
caching it). There's a straightforward fix: simply leave the
top-level entry in place if this happens. This can't waste any
significant amount of memory -- there are at most 256 entries like
this systemwide and, as a practical matter, if we hit this failure
path repeatedly, we're likely to reuse the same page anyway.
For context, this is a reversion with this hunk added in:
if (ret < 0) {
+ /*
+ * Leave the PUD page in place in case some other CPU or thread
+ * already found it, but remove any useless entries we just
+ * added to it.
+ */
- unmap_pgd_range(cpa->pgd, addr,
+ unmap_pud_range(pgd_entry, addr,
addr + (cpa->numpages << PAGE_SHIFT));
return ret;
}
This effectively open-codes what the now-deleted unmap_pgd_range()
function used to do except that unmap_pgd_range() used to try to
free the page as well.
Reported-by: Valdis Kletnieks <Valdis.Kletnieks@vt.edu>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luis R. Rodriguez <mcgrof@suse.com>
Cc: Mike Krinkin <krinkin.m.u@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Toshi Kani <toshi.kani@hp.com>
Link: http://lkml.kernel.org/r/21cbc2822aa18aa812c0215f4231dbf5f65afa7f.1469249789.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
A recent effort to add a new nvdimm bus provider attribute highlighted a
race between interrogating nvdimm_bus->nd_desc and nvdimm_bus tear down.
The typical way to handle these races is to take the device_lock() in
the attribute method and validate that the device is still active. In
order for a device to be 'active' it needs to be associated with a
driver. So, we create the small boilerplate for a driver and register
nvdimm_bus devices on the 'nvdimm_bus_type' bus.
A result of this change is that ndbusX devices now appear under
/sys/bus/nd/devices. In fact this makes /sys/class/nd somewhat
redundant, but removing that will need to take a long deprecation period
given its use by ndctl binaries in the field.
This change naturally pulls code from drivers/nvdimm/core.c to
drivers/nvdimm/bus.c, so it is a nice code organization clean-up as
well.
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Prefix the sector number being cleared with a '0x' to make it clear that
this is a hex value.
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The pcommit instruction is being deprecated in favor of either ADR
(asynchronous DRAM refresh: flush-on-power-fail) at the platform level, or
posted-write-queue flush addresses as defined by the ACPI 6.x NFIT (NVDIMM
Firmware Interface Table).
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: x86@kernel.org
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Acked-by: Ingo Molnar <mingo@redhat.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This reverts commit 8b3e34e46a.
Given the deprecation of the pcommit instruction, the relevant VMX
features and CPUID bits are not going to be rolled into the SDM. Remove
their usage from KVM.
Cc: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The kprobe enablement work has uncovered that changes made by
a guest to MDSCR_EL1 were propagated to the host when VHE was
enabled, leading to unexpected exception being delivered.
Moving this register to the list of registers that are always
context-switched fixes the issue.
Fixes: 9c6c356832 ("arm64: KVM: VHE: Split save/restore of registers shared between guest and host")
Cc: stable@vger.kernel.org #4.6
Reported-by: Tirumalesh Chalamarla <Tirumalesh.Chalamarla@cavium.com>
Tested-by: Tirumalesh Chalamarla <Tirumalesh.Chalamarla@cavium.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
References at the rst files for C functions generated via
kernel-doc should use :c:func:.
Fix it.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
changeset b7e67f6c1b ("doc-rst: linux_tv: supress lots of warnings")
were meant to touch only on media files, but it also touched
at this script by mistake. Revert such change.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Better document a note on this header.
While here, better format dvb_ringbuffer_pkt_read_user()
to adjust it to CodingStyle.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Sphinx can't handle well typedefs. Change two typedef
occurrences, in order to cleanup some of such warnings.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Reorganize the order of the document, putting the chapters
on a more logical order and renaming some sections.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Now that the only remaining chapters at v4l2-framework are
the introduction ones, let' s rename the file.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Move the v4l2 clocks stuff from v4l2-framework to a separate
file and adds an attention that came from the v4l2-clk.h.
Note: as this is meant to be a temporary kAPI, and it is
used only by two drivers (soc_camera and em28xx), where
the first one is in deprecation process, it probably not
a worth effort to document its header.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Several functions weren't documented. Document them all.
While here, makes checkpatch.pl happy.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Move the v4l2 event-specific text from v4l2-framework.rst
to v4l2-event.rst. That helps to keep the text together with
the functions it describes, and makes easier to identify
documentation gaps.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
As we merged the videobuf chapter at the kABI section, and it
is a way more complete, just remove the small videobuf chapter
that came from framework.txt.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Add cross-references for the functions/structs and add
the markup tags to improve its display.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Add documentation for v4l2-dev.h, and put it at v4l2-framework.rst,
where struct video_device is currently documented.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Not all functions at v4l2-common.h are documented. Add
documentation for some other ones.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
There are some subdev-specific functions at v4l2-common.h
that are mentioned at v4l2-subdev.rst.
Document them.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The two new sections were missing cross-references, and had
some other minor issues with the markups. Add such things.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
There are two additional subdev-specific sections at the
v4l2-framework file. Move them to the subdev chapter, in order
to better organize the book.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
There are one enum and 4 functions undocumented there.
Document them. That will fix the broken links at the
v4l2-subdev.rst file.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Enrich the subdevice description by linking it to the
functions and structs from v4l2-subdev.h.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
This header were poorly documented, and weren't using the
kernel-doc format. Document everything but the macros using
the right format.
While here, also fix the other comments to match the
Linux CodingStyle.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
There are lots of documentation about V4L2 subdevices at
v4l2-framework.rst. Move them to its specific chapter at
v4l2-subdev.rst.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
This document describes the main kAPI interfaces for the
v4l2-device.h header. Add cross references to the documentation
produced via kernel-doc.
While here, also use monotonic font for constants.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Part of the contents of v4l2-framework is related to the
kAPI defined by v4l2-device. Move such contents to the
v4l2-device.rst.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Sphinx produce a 1:1 mapping between a rst file and an html file.
So, we need to split the kernel-doc tags on multiple documents.
A side effect is that we're now having a better name for each
section of the kAPI documentation.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The functions at v4l2-device.h are not using the proper
markups. Add it, and include at the v4l2-core.rst.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The kAPI, v4l-drivers and dvb-drivers never used the
GNU FDL license. The addition of such license header were
just due to copy-and-paste. So, let's fix it.
As the media_kapi were part of device-drivers.tmp, it is
under GPL v2+.
The other two books is an agregation of files without any
license explicitly specified. So, they're all bound to the
Kernel's COPYING license. So, they're GPL v2 only.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The Sphinx note markup for media_remove_intf_links() is wrong:
there's a missing space.
While here, let's auto-numerate the two notes.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The uAPI book has 5 parts, but they lost numeration after
conversion to rst. Manually number those parts, and make
the main index with 1 depth, to only show the parts and
the annexes.
At each part, use :maxwidth: 5, in order to show a more
complete index.
While here, fix the cross-references between different
books.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Fix all remaining media warnings with ReST that are fixable
without changing at the Sphinx code.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
* patchwork: (1492 commits)
[media] cec: always check all_device_types and features
[media] cec: poll should check if there is room in the tx queue
[media] vivid: support monitor all mode
[media] cec: fix test for unconfigured adapter in main message loop
[media] cec: limit the size of the transmit queue
[media] cec: zero unused msg part after msg->len
[media] cec: don't set fh to NULL in CEC_TRANSMIT
[media] cec: clear all status fields before transmit and always fill in sequence
[media] cec: CEC_RECEIVE overwrote the timeout field
[media] cxd2841er: Reading SNR for DVB-C added
[media] cxd2841er: Reading BER and UCB for DVB-C added
[media] cxd2841er: fix switch-case for DVB-C
[media] cxd2841er: fix signal strength scale for ISDB-T
[media] cxd2841er: adjust the dB scale for DVB-C
[media] cxd2841er: provide signal strength for DVB-C
[media] cxd2841er: fix BER report via DVBv5 stats API
[media] mb86a20s: apply mask to val after checking for read failure
[media] airspy: fix error logic during device register
[media] s5p-cec/TODO: add TODO item
[media] cec/TODO: drop comment about sphinx documentation
...
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
* 'docs-next' of git://git.lwn.net/linux:
doc-rst: kernel-doc: fix handling of address_space tags
Revert "doc/sphinx: Enable keep_warnings"
doc-rst: kernel-doc directive, fix state machine reporter
docs: deprecate kernel-doc-nano-HOWTO.txt
doc/sphinx: Enable keep_warnings
Documentation: add watermark_scale_factor to the list of vm systcl file
kernel-doc: Fix up warning output
docs: Get rid of some kernel-documentation warnings
Sparse warns:
drivers/dma/ioat/init.c:1215:6: warning: symbol 'ioat_resume' was not declared. Should it be static?
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Acked-by: Dave Jiang <dave.jiang@intel.com>
