13 years ago[MTD] [NAND] fix ifdef option in nand_ecc.c
Timo Lindhorst [Mon, 27 Nov 2006 12:35:49 +0000]
[MTD] [NAND] fix ifdef option in nand_ecc.c

Fix up the config option in the #ifdef statements in nand_ecc.c

Signed-off-by: Timo Lindhorst <lindhors@linux.vnet.ibm.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] [NAND] Update CAFÉ driver interrupt handler prototype
David Woodhouse [Wed, 29 Nov 2006 00:04:59 +0000]
[MTD] [NAND] Update CAFÉ driver interrupt handler prototype

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years agoMerge git://git.infradead.org/~kmpark/onenand-mtd-2.6
David Woodhouse [Wed, 29 Nov 2006 00:03:10 +0000]
Merge git://git.infradead.org/~kmpark/onenand-mtd-2.6

13 years ago[MTD] replace kmalloc+memset with kzalloc
Burman Yan [Wed, 15 Nov 2006 19:10:29 +0000]
[MTD] replace kmalloc+memset with kzalloc

Signed-off-by: Yan Burman <yan_952@hotmail.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] Fix printk format warning in physmap. (resources again)
Randy Dunlap [Tue, 28 Nov 2006 23:40:46 +0000]
[MTD] Fix printk format warning in physmap. (resources again)

Fix printk format warning:
drivers/mtd/maps/physmap.c:93: warning: long long unsigned int format, long unsigned int arg (arg 2)

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years agoMerge git://git.infradead.org/~dwmw2/cafe-2.6
David Woodhouse [Tue, 28 Nov 2006 22:43:10 +0000]
Merge git://git.infradead.org/~dwmw2/cafe-2.6

13 years ago[MTD] [NAND] remove len/ooblen confusion.
Vitaly Wool [Fri, 3 Nov 2006 15:20:38 +0000]
[MTD] [NAND] remove len/ooblen confusion.

As was discussed between Ricard Wanderlöf, David Woodhouse, Artem
Bityutskiy and me, the current API for reading/writing OOB is confusing.

The thing that introduces confusion is the need to specify ops.len
together with ops.ooblen for reads/writes that concern only OOB not data
area. So, ops.len is overloaded: when ops.datbuf != NULL it serves to
specify the length of the data read, and when ops.datbuf == NULL, it
serves to specify the full OOB read length.

The patch inlined below is the slightly updated version of the previous
patch serving the same purpose, but with the new Artem's comments taken
into account.

Artem, BTW, thanks a lot for your valuable input!

Signed-off-by: Vitaly Wool <vwool@ru.mvista.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] Allow variable block sizes in mtd_blkdevs
Richard Purdie [Fri, 27 Oct 2006 08:09:33 +0000]
[MTD] Allow variable block sizes in mtd_blkdevs

Currently, mtd_blkdevs enforces a block size of 512, even if the drivers
can seemingly request a different size. This patch fixes mtd_blkdevs so
block sizes other than 512 work correctly.

Signed-off-by: Richard Purdie <rpurdie@openedhand.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] [MAPS] Support for BIOS flash chips on the nvidia ck804 southbridge
Dave Olsen [Mon, 6 Nov 2006 23:33:57 +0000]
[MTD] [MAPS] Support for BIOS flash chips on the nvidia ck804 southbridge

Add support for accessing BIOS flash chips connected to the NVIDIA ck804 southbridge.

Signed-off-by: Ryan Jackson <rjackson@lnxi.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] [NAND] Fix endianess bug in ndfc.c
Stefan Roese [Fri, 10 Nov 2006 11:19:32 +0000]
[MTD] [NAND] Fix endianess bug in ndfc.c

The writel() call accidentally clears all bits in the NDFC_CCR
register (endianess problem). Now __raw_writel() is used instead.

Tested on Bamboo with NAND on chip select 0 and chip select 1.

Signed-off-by: Stefan Roese <sr@denx.de>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] make drivers/mtd/cmdlinepart.c:mtdpart_setup() static
Adrian Bunk [Sat, 25 Nov 2006 19:15:41 +0000]
[MTD] make drivers/mtd/cmdlinepart.c:mtdpart_setup() static

This patch makes the needlessly global mtdpart_setup() static.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] [NAND] rtc_from4.c: use lib/bitrev.c
Adrian Bunk [Wed, 22 Nov 2006 04:38:11 +0000]
[MTD] [NAND] rtc_from4.c: use lib/bitrev.c

This patch converts drivers/mtd/nand/rtc_from4.c to use the new
lib/bitrev.c

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] OneNAND: Single bit error detection
Kyungmin Park [Thu, 16 Nov 2006 03:03:56 +0000]
[MTD] OneNAND: Single bit error detection

Idea from Jarkko Lavinen

Signed-off-by: Jarkko Lavinen <jarkko.lavinen@nokia.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>

13 years ago[MTD] OneNAND: lock support
Kyungmin Park [Thu, 16 Nov 2006 02:29:39 +0000]
[MTD] OneNAND: lock support

Now you can use mtd lock inferface on OneNAND

The idea is from Nemakal, Vijaya, thanks

Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>

13 years agoMTD: OneNAND: interrupt based wait support
Kyungmin Park [Thu, 16 Nov 2006 02:23:48 +0000]
MTD: OneNAND: interrupt based wait support

  We can use the two methods to wait.
  1. polling: read interrupt status register
  2. interrupt: use kernel ineterrupt mechanism

  To use interrupt method, you first connect onenand interrupt pin to your
platform and configure interrupt properly

Signed-off-by: Kyungmin Park <kyungmin.park at samsung.com>

13 years ago[MTD] NAND: Fix ECC settings in CAFÉ controller driver.
David Woodhouse [Wed, 1 Nov 2006 00:19:20 +0000]
[MTD] NAND: Fix ECC settings in CAFÉ controller driver.

We were resetting cafe->ctl2 to zero after an erase (and also during a
write, but it was correctly reset after that). This meant that ECC reads
after an erase were failing. Doh.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Add register debugging spew option to CAFÉ driver
David Woodhouse [Tue, 31 Oct 2006 04:36:04 +0000]
[MTD] NAND: Add register debugging spew option to CAFÉ driver

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Use register #defines throughout CAFÉ driver, not numbers
David Woodhouse [Tue, 31 Oct 2006 04:30:11 +0000]
[MTD] NAND: Use register #defines throughout CAFÉ driver, not numbers

Also use cafe_readl() and cafe_writel() abstraction to make code
slightly cleaner -- especially if we want to use it in PIO mode.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Fix timing calculation in CAFÉ debugging message
David Woodhouse [Sat, 28 Oct 2006 14:08:38 +0000]
[MTD] NAND: Fix timing calculation in CAFÉ debugging message

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Remove empty block ECC workaround
David Woodhouse [Fri, 27 Oct 2006 19:12:02 +0000]
[MTD] NAND: Remove empty block ECC workaround

They fixed the hardware so that ECC doesn't fail on reading an empty
block.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Add ECC debugging for CAFÉ
David Woodhouse [Fri, 27 Oct 2006 12:40:51 +0000]
[MTD] NAND: Add ECC debugging for CAFÉ

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] CAFÉ NAND: Add 'slowtiming' parameter, default usedma and checkecc on
David Woodhouse [Fri, 27 Oct 2006 11:50:04 +0000]
[MTD] CAFÉ NAND: Add 'slowtiming' parameter, default usedma and checkecc on

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Reset Café controller before initialising.
David Woodhouse [Fri, 27 Oct 2006 06:55:34 +0000]
[MTD] NAND: Reset Café controller before initialising.

Fixes http://dev.laptop.org/ticket/237

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Café ECC -- remove spurious BUG_ON() in err_pos()
David Woodhouse [Fri, 27 Oct 2006 06:53:08 +0000]
[MTD] NAND: Café ECC -- remove spurious BUG_ON() in err_pos()

Being a value which isn't in the table is a case we explicitly check for
in the caller. Don't BUG_ON() because it does actually happen in
practice.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Fix nand_default_mark_blockbad() when flash-based BBT disabled
Ricard Wanderlöf [Mon, 23 Oct 2006 07:33:34 +0000]
[MTD] NAND: Fix nand_default_mark_blockbad() when flash-based BBT disabled

When a flash-based BBT is not used, nand_default_mark_blockbad() is supposed
to mark the block bad in the oob. However, it sets the wrong length variable
so that no bad block marker is in fact written. This patch attempts to
rectify that.

(As note, it seems to be that logically, it shouldn't be necessary to set
both length variables, as one appears to be for the main buffer, and
one for the oob buffer, but this is how it is done in several places,
including the code for the mtd character device MEMWRITEOOB and MEMREADOOB
ioctls. I'm not sure if this is a temporary solution during some rework of
the mtd infrastructure, or whether there is a deeper thought here.)

Signed-off-by: Ricard Wanderlöf <ricardw@axis.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Disable ECC checking on CAFÉ since it's broken for now
David Woodhouse [Mon, 23 Oct 2006 13:29:04 +0000]
[MTD] NAND: Disable ECC checking on CAFÉ since it's broken for now

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: CAFÉ NAND driver cleanup, fix ECC on reading empty flash
David Woodhouse [Sun, 22 Oct 2006 14:09:33 +0000]
[MTD] NAND: CAFÉ NAND driver cleanup, fix ECC on reading empty flash

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Add hardware ECC correction support to CAFÉ NAND driver
David Woodhouse [Sun, 22 Oct 2006 01:18:48 +0000]
[MTD] NAND: Add hardware ECC correction support to CAFÉ NAND driver

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years agoMerge git://git.infradead.org/~dwmw2/cafe-2.6
David Woodhouse [Sun, 22 Oct 2006 01:17:05 +0000]
Merge git://git.infradead.org/~dwmw2/cafe-2.6

Conflicts:

drivers/mtd/nand/Kconfig

13 years ago[MTD] NAND: Correct setting of chip->oob_poi OOB buffer
David Woodhouse [Sun, 22 Oct 2006 00:47:45 +0000]
[MTD] NAND: Correct setting of chip->oob_poi OOB buffer

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: Combined oob buffer so it's contiguous with data
David Woodhouse [Sat, 21 Oct 2006 16:09:53 +0000]
[MTD] NAND: Combined oob buffer so it's contiguous with data

Ditch the separate oobrbuf and oobwbuf fields from the chip buffers,
and use only a single buffer immediately after the data. This accommodates
NAND controllers such as the OLPC CAFÉ chip, which can't do scatter/gather
DMA so needs the OOB buffer to be contiguous with the data, for both read
and write.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6
David Woodhouse [Sat, 21 Oct 2006 15:46:04 +0000]
Merge branch 'master' of git://git./linux/kernel/git/torvalds/linux-2.6

13 years ago[MTD] NAND: nandsim: support subpage write
Artem Bityutskiy [Fri, 20 Oct 2006 08:23:56 +0000]
[MTD] NAND: nandsim: support subpage write

As flash cannot do 0->1 bit transitions when programming, do not do this in
the simulator too. This makes nandsim able to accept subpage writes.

Signed-off-by: Artem Bityutskiy <dedekind@infradead.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NOR: leave Intel chips in read-array mode on suspend
David Anders [Thu, 19 Oct 2006 16:33:19 +0000]
[MTD] NOR: leave Intel chips in read-array mode on suspend

During some testing with several samsung s3c24xx based
devices it was discovered that often the
cfi_cmdset_0001.c would not leave the chip in
read-array mode on suspend. this is an issue if the
same flash chip is used for the bootloader that needs
to be read on resume.

Signed-off-by: David Anders <danders@amltd.com>
Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] core: trivial comments fix
Artem Bityutskiy [Thu, 19 Oct 2006 13:04:02 +0000]
[MTD] core: trivial comments fix

Signed-off-by: Artem Bityutskiy <dedekind@infradead.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: nandsim coding style fix
Vijay Kumar [Sat, 14 Oct 2006 16:03:34 +0000]
[MTD] NAND: nandsim coding style fix

Removes line break after return type in function definitions, to be
consistent with the Linux coding style.

Signed-off-by: Vijay Kumar <vijaykumar@bravegnu.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: nandsim page-wise allocation (2/2)
Vijay Kumar [Sun, 8 Oct 2006 16:32:31 +0000]
[MTD] NAND: nandsim page-wise allocation (2/2)

For page wise allocation, an array of flash page pointers is allocated
during initialization. The flash pages are themselves allocated when a
write occurs to the page. The flash pages are deallocated when they
are erased.

Signed-off-by: Vijay Kumar <vijaykumar@bravegnu.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: nandsim page-wise allocation (1/2)
Vijay Kumar [Sun, 8 Oct 2006 16:30:37 +0000]
[MTD] NAND: nandsim page-wise allocation (1/2)

This patch removes code that does chip mapping. The chip mapping code
is no longer used.

Signed-off-by: Vijay Kumar <vijaykumar@bravegnu.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] MAPS: Remove ITE 8172G and Globespan IVR MTD support
Yoichi Yuasa [Thu, 12 Oct 2006 08:38:15 +0000]
[MTD] MAPS: Remove ITE 8172G and Globespan IVR MTD support

This patch has removed ITE 8172G and Globespan IVR MTD support.
These boards support have already been removed.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Acked-by: Ralf Bächle <ralf@linux-mips.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] mtdchar: Fix MEMGETOOBSEL and ECCGETLAYOUT ioctls
Ricard Wanderlöf [Tue, 17 Oct 2006 15:27:11 +0000]
[MTD] mtdchar: Fix MEMGETOOBSEL and ECCGETLAYOUT ioctls

1. The ECCGETLAYOUT ioctl copy_to_user() call has a superfluous '&'
causing the resulting information to be garbage rather than the intended
mtd->ecclayout.

2. The MEMGETOOBSEL misses copying mtd->ecclayout->eccbytes so the
resulting field of the returned structure contains garbage.

Signed-off-by: Ricard Wanderlöf <ricardw@axis.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] NAND: AT91 NAND driver
Andrew Victor [Thu, 19 Oct 2006 16:24:35 +0000]
[MTD] NAND: AT91 NAND driver

This version only differs from version posted by Savin Zlobec (20 Jun
2006) in that the AT91RM9200-specific chip-select / bus setup code has
been moved from the at91_nand.c driver into the processor-specific file.

From: Savin Zlobec <savin@epico.si>
Signed-off-by: Andrew Victor <andrew@sanpeople.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[JFFS2] Fix jffs2_follow_link() typo
Qi Yong [Mon, 16 Oct 2006 11:42:23 +0000]
[JFFS2] Fix jffs2_follow_link() typo

typo fix: noticed this typo while reading the patch
"jffs2: fix symlink error handling"

Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] MAPS: esb2rom: use hotplug safe interfaces
Alan Cox [Fri, 20 Oct 2006 21:41:06 +0000]
[MTD] MAPS: esb2rom: use hotplug safe interfaces

Fairly self explanatory.  Keep a reference initially, drop it when we free up
the driver resources.

Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] JEDEC probe: fix comment typo (devic)
Alexey Dobriyan [Fri, 20 Oct 2006 21:41:05 +0000]
[MTD] JEDEC probe: fix comment typo (devic)

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[JFFS2] Use rb_first() and rb_last() cleanup
Akinobu Mita [Fri, 20 Oct 2006 21:41:05 +0000]
[JFFS2] Use rb_first() and rb_last() cleanup

Use rb_first() and rb_last() to implement frag_first() and frag_last().

Signed-off-by: Akinbou Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] MAPS: Support for BIOS flash chips on Intel ESB2 southbridge
Lew Glendenning [Fri, 20 Oct 2006 21:41:04 +0000]
[MTD] MAPS: Support for BIOS flash chips on Intel ESB2 southbridge

Add MTD map driver for BIOS flash chips connected to the Intel ESB2
southbridge.

[akpm@osdl.org: coding-style fixes, build fix]
Signed-off-by: Ryan Jackson <rjackson@lnxi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] CHIPS: Support for SST 49LF040B flash chip
Ryan Jackson [Fri, 20 Oct 2006 21:41:03 +0000]
[MTD] CHIPS: Support for SST 49LF040B flash chip

Add chip driver and JEDEC probe support for the SST 49LF040B flash chip.

Signed-off-by: Ryan Jackson <rjackson@lnxi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[MTD] MAPS: Add parameter to amd76xrom to override rom window size
Ryan Jackson [Fri, 20 Oct 2006 21:41:01 +0000]
[MTD] MAPS: Add parameter to amd76xrom to override rom window size

The 2 bits controlling the window size are often set to allow reading the
BIOS, but too small to allow writing, since the lock registers are 4MiB
lower in the address space than the data.  This is intended to prevent
flashing the bios, perhaps accidentally.

The bits are 6 and 7.  If both bits are set, it is a 5MiB window. If only
the 7 Bit is set, it is a 4MiB window.  Otherwise, it is a 64KiB window.

This parameter allows the driver to override the BIOS settings.

Signed-off-by: Ryan Jackson <rjackson@lnxi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>

13 years ago[PATCH] fix PXA2xx UDC compilation error
Nicolas Pitre [Fri, 20 Oct 2006 21:20:17 +0000]
[PATCH] fix PXA2xx UDC compilation error

This was apparently missed by the move to the generic IRQ code.

Signed-off-by: Nicolas Pitre <nico@cam.org>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] Fix up rpaphp driver for pci hotplug header move
Olaf Hering [Fri, 20 Oct 2006 07:23:05 +0000]
[PATCH] Fix up rpaphp driver for pci hotplug header move

Use grep instead of make during interface changes.

Signed-off-by: Olaf Hering <olaf@aepfle.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years agoconfigfs: handle kzalloc() failure in check_perm()
Chandra Seetharaman [Tue, 10 Oct 2006 22:15:55 +0000]
configfs: handle kzalloc() failure in check_perm()

check_perm() does not drop the reference to the module when kzalloc()
failure occurs.

Signed-Off-By: Chandra Seetharaman <sekharan@us.ibm.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>

13 years agoocfs2: cond_resched() in ocfs2_zero_extend()
Mark Fasheh [Wed, 4 Oct 2006 00:53:05 +0000]
ocfs2: cond_resched() in ocfs2_zero_extend()

The loop within ocfs2_zero_extend() can execute for a long time, causing
spurious soft lockup warnings.

Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>

13 years agoocfs2: fix page zeroing during simple extends
Mark Fasheh [Wed, 4 Oct 2006 00:44:42 +0000]
ocfs2: fix page zeroing during simple extends

The page zeroing code was missing the region between old i_size and new
i_size for those extends that didn't actually require a change in space
allocation.

Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>

13 years agoocfs2: remove spurious d_count check in ocfs2_rename()
Sunil Mushran [Wed, 11 Oct 2006 19:23:02 +0000]
ocfs2: remove spurious d_count check in ocfs2_rename()

This was causing some folks to incorrectly get -EBUSY during rename.

Signed-off-by: Sunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>

13 years agoocfs2: delete redundant memcmp()
Akinobu Mita [Thu, 12 Oct 2006 05:29:33 +0000]
ocfs2: delete redundant memcmp()

This patch deletes redundant memcmp() while looking up in rb tree.

Signed-off-by: Akinbou Mita <akinobu.mita@gmail.com>
Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>

13 years agoMerge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
Linus Torvalds [Fri, 20 Oct 2006 17:44:52 +0000]
Merge branch 'upstream' of git://ftp.linux-mips.org/upstream-linus

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] Fix iounmap argument to const volatile.
  [MIPS] Reserve syscall numbers for kexec_load.
  [MIPS] Update tb0287_defconfig
  [MIPS] Update pnx8550-v2pci_defconfig
  [MIPS] Update pnx8500-jbs_defconfig
  [MIPS] More vr41xx pt_regs fixups
  [MIPS] save_context_stack fix
  [MIPS] Use compat_sys_mount.
  [MIPS] Fix O32 personality(2) call with 0xffffffff argument.
  [MIPS] A few more pt_regs fixups.
  [MIPS] Malta: Fix uninitialized regs pointer.
  [MIPS] Delete unneeded pt_regs forward declaration.
  [MIPS] Use kallsyms_lookup_size_offset() instead of kallsyms_lookup()

13 years agoRevert "[mv643xx] Add pci device table for auto module loading."
Linus Torvalds [Fri, 20 Oct 2006 17:40:48 +0000]
Revert "[mv643xx] Add pci device table for auto module loading."

This reverts commit 4596c75c23dde2623cbeec69357d5eb13d28387e as
requested by Olaf Hering. It causes compile errors, and says Olaf:

  "This change is also wrong, the autoloading works perfect with 2.6.18,
   no need to add random PCI ids.

   See commit a0245f7ad5214cb00131d7cd176446e067c913dc, platform devices
   have now a modalias entry in sysfs.  The network card is not a PCI
   device."

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Fri, 20 Oct 2006 17:28:48 +0000]
Merge git://git./linux/kernel/git/herbert/crypto-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  [CRYPTO] api: Select cryptomgr where needed
  [CRYPTO] api: fix crypto_alloc_base() return value

13 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-fixes
Linus Torvalds [Fri, 20 Oct 2006 17:28:14 +0000]
Merge git://git./linux/kernel/git/steve/gfs2-2.6-fixes

* git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-fixes:
  [GFS2] gfs2_dir_read_data(): fix uninitialized variable usage
  [GFS2] fs/gfs2/ops_fstype.c:fill_super_meta(): fix NULL dereference
  [GFS2] fs/gfs2/dir.c:gfs2_dir_write_data(): don't use an uninitialized variable
  [GFS2] fs/gfs2/ops_fstype.c:gfs2_get_sb_meta(): remove unused variable
  [GFS2] fs/gfs2/dir.c:gfs2_dir_write_data(): remove dead code
  [GFS2] gfs2 endianness bug: be16 assigned to be32 field
  [GFS2] Fix bmap to map extents properly
  [DLM] fix iovec length in recvmsg

13 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6
Linus Torvalds [Fri, 20 Oct 2006 17:27:55 +0000]
Merge /pub/scm/linux/kernel/git/davem/sparc-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6:
  [SPARC]: Clean up asm-sparc/elf.h pollution in userspace.
  [SPARC64]: Fix of_ioremap().
  [SPARC64]: Compute dma_end argument to sabre_pbm_init() correctly.

13 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Fri, 20 Oct 2006 17:27:38 +0000]
Merge /pub/scm/linux/kernel/git/davem/net-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6: (36 commits)
  [Bluetooth] Fix HID disconnect NULL pointer dereference
  [Bluetooth] Add missing entry for Nokia DTL-4 PCMCIA card
  [Bluetooth] Add support for newer ANYCOM USB dongles
  [NET]: Can use __get_cpu_var() instead of per_cpu() in loopback driver.
  [IPV4] inet_peer: Group together avl_left, avl_right, v4daddr to speedup lookups on some CPUS
  [TCP]: One NET_INC_STATS() could be NET_INC_STATS_BH in tcp_v4_err()
  [NETFILTER]: Missing check for CAP_NET_ADMIN in iptables compat layer
  [NETPOLL]: initialize skb for UDP
  [IPV6]: Fix route.c warnings when multiple tables are disabled.
  [TG3]: Bump driver version and release date.
  [TG3]: Add lower bound checks for tx ring size.
  [TG3]: Fix set ring params tx ring size implementation
  [NET]: reduce per cpu ram used for loopback stats
  [IPv6] route: Fix prohibit and blackhole routing decision
  [DECNET]: Fix input routing bug
  [TCP]: Bound TSO defer time
  [IPv4] fib: Remove unused fib_config members
  [IPV6]: Always copy rt->u.dst.error when copying a rt6_info.
  [IPV6]: Make IPV6_SUBTREES depend on IPV6_MULTIPLE_TABLES.
  [IPV6]: Clean up BACKTRACK().
  ...

13 years ago[PATCH] one more ARM IRQ fix
Adrian Bunk [Fri, 20 Oct 2006 06:29:14 +0000]
[PATCH] one more ARM IRQ fix

Fix one more compile breakage caused by the post -rc1 IRQ changes.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] knfsd: fix race that can disable NFS server
NeilBrown [Fri, 20 Oct 2006 06:29:13 +0000]
[PATCH] knfsd: fix race that can disable NFS server

This patch is suitable for just about any 2.6 kernel.  It should go in
2.6.19 and 2.6.18.2 and possible even the .17 and .16 stable series.

This is a long standing bug that seems to have only recently become
apparent, presumably due to increasing use of NFS over TCP - many
distros seem to be making it the default.

The SK_CONN bit gets set when a listening socket may be ready
for an accept, just as SK_DATA is set when data may be available.

It is entirely possible for svc_tcp_accept to be called with neither
of these set.  It doesn't happen often but there is a small race in
svc_sock_enqueue as SK_CONN and SK_DATA are tested outside the
spin_lock.  They could be cleared immediately after the test and
before the lock is gained.

This normally shouldn't be a problem.  The sockets are non-blocking so
trying to read() or accept() when ther is nothing to do is not a problem.

However: svc_tcp_recvfrom makes the decision "Should I accept() or
should I read()" based on whether SK_CONN is set or not.  This usually
works but is not safe.  The decision should be based on whether it is
a TCP_LISTEN socket or a TCP_CONNECTED socket.

Signed-off-by: Neil Brown <neilb@suse.de>
Cc: Adrian Bunk <bunk@stusta.de>
Cc: <stable@kernel.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] kernel/nsproxy.c: use kmemdup()
Alexey Dobriyan [Fri, 20 Oct 2006 06:29:12 +0000]
[PATCH] kernel/nsproxy.c: use kmemdup()

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] ext3/4: fix J_ASSERT(transaction->t_updates > 0) in journal_stop()
OGAWA Hirofumi [Fri, 20 Oct 2006 06:29:11 +0000]
[PATCH] ext3/4: fix J_ASSERT(transaction->t_updates > 0) in journal_stop()

A disk generated some I/O error, after it, I hitted
J_ASSERT(transaction->t_updates > 0) in journal_stop().

It seems to happened on ext3_truncate() path from stack trace. Then,
maybe the following case may trigger J_ASSERT(transaction->t_updates > 0).

ext3_truncate()
    -> ext3_free_branches()
        -> ext3_journal_test_restart()
    -> ext3_journal_restart()
                -> journal_restart()
                transaction->t_updates--;
                /* another process aborted journal */
                    -> start_this_handle()
    returns -EROFS without transaction->t_updates++;

    -> ext3_journal_stop()
        -> journal_stop()
J_ASSERT(transaction->t_updates > 0)

If journal was aborted in middle of journal_restart(), ext3_truncate()
may trigger J_ASSERT().

Signed-off-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] mm: more commenting on lock ordering
Nick Piggin [Fri, 20 Oct 2006 06:29:10 +0000]
[PATCH] mm: more commenting on lock ordering

Clarify lockorder comments now that sys_msync dropps mmap_sem before
calling do_fsync.

Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] firmware/dcdbas: add size check in smi_data_write
Doug Warzecha [Fri, 20 Oct 2006 06:29:09 +0000]
[PATCH] firmware/dcdbas: add size check in smi_data_write

Add a size check in smi_data_write to prevent possible wrapping problems
with large pos values when calling smi_data_buf_realloc on 32-bit.

Signed-off-by: Doug Warzecha <Douglas_Warzecha@dell.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] ipmi: fix return codes in failure case
Dave Jones [Fri, 20 Oct 2006 06:29:09 +0000]
[PATCH] ipmi: fix return codes in failure case

These returns should be negative, like the others in this function.

Signed-off-by: Dave Jones <davej@redhat.com>
Acked-by: Corey Minyard <minyard@acm.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] mm: D-cache aliasing issue in cow_user_page
Dmitriy Monakhov [Fri, 20 Oct 2006 06:29:08 +0000]
[PATCH] mm: D-cache aliasing issue in cow_user_page

--=-=-=

 from mm/memory.c:
  1434  static inline void cow_user_page(struct page *dst, struct page *src, unsigned long va)
  1435  {
  1436          /*
  1437           * If the source page was a PFN mapping, we don't have
  1438           * a "struct page" for it. We do a best-effort copy by
  1439           * just copying from the original user address. If that
  1440           * fails, we just zero-fill it. Live with it.
  1441           */
  1442          if (unlikely(!src)) {
  1443                  void *kaddr = kmap_atomic(dst, KM_USER0);
  1444                  void __user *uaddr = (void __user *)(va & PAGE_MASK);
  1445
  1446                  /*
  1447                   * This really shouldn't fail, because the page is there
  1448                   * in the page tables. But it might just be unreadable,
  1449                   * in which case we just give up and fill the result with
  1450                   * zeroes.
  1451                   */
  1452                  if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE))
  1453                          memset(kaddr, 0, PAGE_SIZE);
  1454                  kunmap_atomic(kaddr, KM_USER0);
  #### D-cache have to be flushed here.
  #### It seems it is just forgotten.

  1455                  return;
  1456
  1457          }
  1458          copy_user_highpage(dst, src, va);
  #### Ok here. flush_dcache_page() called from this func if arch need it
  1459  }

Following is the patch  fix this issue:

Signed-off-by: Dmitriy Monakhov <dmonakhov@openvz.org>
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] doc: fixing cpu-hotplug documentation
Satoru Takeuchi [Fri, 20 Oct 2006 06:29:06 +0000]
[PATCH] doc: fixing cpu-hotplug documentation

Fixing cpu-hotplug documentation as follows:

 - moving confusing asterisk on additional_cpus descrition
 - fixing some typos
 - unifying indentation for source code and command line example

Signed-off-by: Satoru Takeuchi <takeuchi_satoru@jp.fujitsu.com>
Cc: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] drivers/isdn: ioremap balanced with iounmap
Amol Lad [Fri, 20 Oct 2006 06:29:06 +0000]
[PATCH] drivers/isdn: ioremap balanced with iounmap

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Acked-by: Karsten Keil <kkeil@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] highest_possible_node_id() linkage fix
Andrew Morton [Fri, 20 Oct 2006 06:29:05 +0000]
[PATCH] highest_possible_node_id() linkage fix

Qooting Adrian:

- net/sunrpc/svc.c uses highest_possible_node_id()

- include/linux/nodemask.h says highest_possible_node_id() is
  out-of-line #if MAX_NUMNODES > 1

- the out-of-line highest_possible_node_id() is in lib/cpumask.c

- lib/Makefile: lib-$(CONFIG_SMP) += cpumask.o
  CONFIG_ARCH_DISCONTIGMEM_ENABLE=y, CONFIG_SMP=n, CONFIG_SUNRPC=y

-> highest_possible_node_id() is used in net/sunrpc/svc.c
   CONFIG_NODES_SHIFT defined and > 0

-> include/linux/numa.h: MAX_NUMNODES > 1

-> compile error

The bug is not present on architectures where ARCH_DISCONTIGMEM_ENABLE
depends on NUMA (but m32r isn't the only affected architecture).

So move the function into page_alloc.c

Cc: Adrian Bunk <bunk@stusta.de>
Cc: Paul Jackson <pj@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] Fix potential interrupts during alternative patching
Zachary Amsden [Fri, 20 Oct 2006 06:29:04 +0000]
[PATCH] Fix potential interrupts during alternative patching

Interrupts must be disabled during alternative instruction patching.  On
systems with high timer IRQ rates, or when running in an emulator, timing
differences can result in random kernel panics because of running partially
patched instructions.  This doesn't yet fix NMIs, which requires extricating
the patch code from the late bug checking and is logically separate (and also
less likely to cause problems).

Signed-off-by: Zachary Amsden <zach@vmware.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsd: nfs_replay_me
Al Viro [Fri, 20 Oct 2006 06:29:03 +0000]
[PATCH] nfsd: nfs_replay_me

We are using NFS_REPLAY_ME as a special error value that is never leaked to
clients.  That works fine; the only problem is mixing host- and network-
endian values in the same objects.  Network-endian equivalent would work just
as fine; switch to it.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsd: misc endianness annotations
Al Viro [Fri, 20 Oct 2006 06:29:02 +0000]
[PATCH] nfsd: misc endianness annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: nfsd callback*
Al Viro [Fri, 20 Oct 2006 06:29:01 +0000]
[PATCH] xdr annotations: nfsd callback*

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsd: NFSv4 errno endianness annotations
Al Viro [Fri, 20 Oct 2006 06:29:01 +0000]
[PATCH] nfsd: NFSv4 errno endianness annotations

don't use the same variable to store NFS and host error values

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsd: NFSv{2,3} trivial endianness annotations for error values
Al Viro [Fri, 20 Oct 2006 06:29:00 +0000]
[PATCH] nfsd: NFSv{2,3} trivial endianness annotations for error values

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsd: nfs4 code returns error values in net-endian
Al Viro [Fri, 20 Oct 2006 06:28:59 +0000]
[PATCH] nfsd: nfs4 code returns error values in net-endian

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsd: vfs.c endianness annotations
Al Viro [Fri, 20 Oct 2006 06:28:58 +0000]
[PATCH] nfsd: vfs.c endianness annotations

don't use the same variable to store NFS and host error values

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: NFSv4 server
Al Viro [Fri, 20 Oct 2006 06:28:58 +0000]
[PATCH] xdr annotations: NFSv4 server

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: NFSv3 server
Al Viro [Fri, 20 Oct 2006 06:28:57 +0000]
[PATCH] xdr annotations: NFSv3 server

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: NFSv2 server
Al Viro [Fri, 20 Oct 2006 06:28:56 +0000]
[PATCH] xdr annotations: NFSv2 server

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: nfsd_dispatch()
Al Viro [Fri, 20 Oct 2006 06:28:55 +0000]
[PATCH] xdr annotations: nfsd_dispatch()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsfh simple endianness annotations
Al Viro [Fri, 20 Oct 2006 06:28:55 +0000]
[PATCH] nfsfh simple endianness annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfsd: nfserrno() endianness annotations
Al Viro [Fri, 20 Oct 2006 06:28:54 +0000]
[PATCH] nfsd: nfserrno() endianness annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfs_common endianness annotations
Al Viro [Fri, 20 Oct 2006 06:28:53 +0000]
[PATCH] nfs_common endianness annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: mount_clnt
Al Viro [Fri, 20 Oct 2006 06:28:52 +0000]
[PATCH] xdr annotations: mount_clnt

[pulled from Alexey's patch]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] nfs: verifier is network-endian
Al Viro [Fri, 20 Oct 2006 06:28:51 +0000]
[PATCH] nfs: verifier is network-endian

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: fs/nfs/callback*
Al Viro [Fri, 20 Oct 2006 06:28:51 +0000]
[PATCH] xdr annotations: fs/nfs/callback*

on-the-wire data is big-endian

[mostly pulled from Alexey's patch]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] fs/nfs/callback* passes error values big-endian
Al Viro [Fri, 20 Oct 2006 06:28:50 +0000]
[PATCH] fs/nfs/callback* passes error values big-endian

[pulled from Alexey's patch]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: NFS readdir entries
Al Viro [Fri, 20 Oct 2006 06:28:49 +0000]
[PATCH] xdr annotations: NFS readdir entries

on-the-wire data is big-endian

[in large part pulled from Alexey's patch]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: NFSv4
Al Viro [Fri, 20 Oct 2006 06:28:48 +0000]
[PATCH] xdr annotations: NFSv4

on-the-wire data is big-endian

[in large part pulled from Alexey's patch]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: NFSv3
Al Viro [Fri, 20 Oct 2006 06:28:48 +0000]
[PATCH] xdr annotations: NFSv3

on-the-wire data is big-endian

[in large part pulled from Alexey's patch]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] xdr annotations: NFSv2
Al Viro [Fri, 20 Oct 2006 06:28:47 +0000]
[PATCH] xdr annotations: NFSv2

on-the-wire data is big-endian

[in large part pulled from Alexey's patch]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] lockd endianness annotations
Al Viro [Fri, 20 Oct 2006 06:28:46 +0000]
[PATCH] lockd endianness annotations

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] fix svc_procfunc declaration
Al Viro [Fri, 20 Oct 2006 06:28:45 +0000]
[PATCH] fix svc_procfunc declaration

svc_procfunc instances return __be32, not int

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] bug: nfsd/nfs4xdr.c misuse of ERR_PTR()
Al Viro [Fri, 20 Oct 2006 06:28:44 +0000]
[PATCH] bug: nfsd/nfs4xdr.c misuse of ERR_PTR()

a) ERR_PTR(nfserr_something) is a bad idea;
IS_ERR() will be false for it.
b) mixing nfserr_.... with -EOPNOTSUPP is
even worse idea.

nfsd4_path() does both; caller expects to get NFS protocol error out it if
anything goes wrong, but if it does we either do not notice (see (a)) or get
host-endian negative (see (b)).

IOW, that's a case when we can't use ERR_PTR() to return error, even though we
return a pointer in case of success.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] SUNRPC: fix a typo
Chuck Lever [Fri, 20 Oct 2006 06:28:44 +0000]
[PATCH] SUNRPC: fix a typo

Yes, this actually passed tests the way it was.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

13 years ago[PATCH] SUNRPC: fix race in in-kernel RPC portmapper client
Chuck Lever [Fri, 20 Oct 2006 06:28:43 +0000]
[PATCH] SUNRPC: fix race in in-kernel RPC portmapper client

When submitting a request to a fast portmapper (such as the local rpcbind
daemon), the request can complete before the parent task is even queued up on
xprt->binding.  Fix this by queuing before submitting the rpcbind request.

Test plan:
Connectathon locking test with UDP.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>