Dan Williams [Fri, 20 Jul 2007 07:31:46 +0000 (00:31 -0700)]
async_tx: fix kmap_atomic usage in async_memcpy
Andrew Morton:
[async_memcpy] is very wrong if both ASYNC_TX_KMAP_DST and
ASYNC_TX_KMAP_SRC can ever be set. We'll end up using the same kmap
slot for both src add dest and we get either corrupted data or a BUG.
Evgeniy Polyakov:
Btw, shouldn't it always be kmap_atomic() even if flag is not set.
That pages are usual one returned by alloc_page().
So fix the usage of kmap_atomic and kill the ASYNC_TX_KMAP_DST and
ASYNC_TX_KMAP_SRC flags.
Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Evgeniy Polyakov <johnpol@2ka.mipt.ru> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Nick Piggin [Fri, 20 Jul 2007 07:31:45 +0000 (00:31 -0700)]
fix some conversion overflows
Fix page index to offset conversion overflows in buffer layer, ecryptfs,
and ocfs2.
It would be nice to convert the whole tree to page_offset, but for now
just fix the bugs.
Signed-off-by: Nick Piggin <npiggin@suse.de> Cc: Michael Halcrow <mhalcrow@us.ibm.com> Cc: Mark Fasheh <mark.fasheh@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Paul Mundt [Fri, 20 Jul 2007 07:31:44 +0000 (00:31 -0700)]
mm: fix memory hotplug oops from ZONE_MOVABLE changes.
zone_movable_pfn is presently marked as __initdata and referenced from
adjust_zone_range_for_zone_movable(), which in turn is referenced by
zone_spanned_pages_in_node(). Both of these are __meminit annotated. When
memory hotplug is enabled, this will oops on a hot-add, due to
zone_movable_pfn having been freed.
__meminitdata annotation gives the desired behaviour.
This will only impact platforms that enable both memory hotplug
and ARCH_POPULATES_NODE_MAP.
Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: Mel Gorman <mel@csn.ul.ie> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Roland McGrath [Fri, 20 Jul 2007 07:31:43 +0000 (00:31 -0700)]
xen: disable vdso "nosegneg" on native boot
One of the nice ideas behind paravirt is that CONFIG_XEN=y can be included
in a standard configuration and be no worse for native booting than as a
Xen guest. The glibc feature that supports the vDSO "nosegneg" note is
designed specifically to make this easy. You just have to flip one bit at
boot time. This patch makes Xen flip the bit, so a CONFIG_XEN=y kernel on
bare hardware does not make glibc use the less-optimized library builds.
Signed-off-by: Roland McGrath <roland@redhat.com> Acked-by: Jeremy Fitzhardinge <jeremy@xensource.com> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6
* 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6:
[SPARC64]: Fix two year old bug in early bootup asm.
[SPARC64]: Update defconfig.
[SPARC64]: Fix log message type in vio_create_one().
[SPARC64]: Tweak assertions in sun4v_build_virq().
[SPARC64]: Tweak kernel log messages in power_probe().
[SPARC64]: Fix handling of multiple vdc-port nodes.
[SPARC64]: Fix device type matching in VIO's devspec_show().
[SPARC64]: Fix MODULE_DEVICE_TABLE() specification in VDC and VNET.
[SPARC]: Add sys_fallocate() entries.
[SPARC64]: Use orderly_poweroff().
Al Viro [Fri, 20 Jul 2007 15:07:33 +0000 (16:07 +0100)]
Fix up sky2 breakage
Doing |= 1 << 19 to 16bit unsigned is not particulary useful;
that register is 32bit, unlike the ones dealt with in the rest of
function, so we need u32 variable here.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Acked-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
David S. Miller [Fri, 20 Jul 2007 04:50:09 +0000 (21:50 -0700)]
[SPARC64]: Fix two year old bug in early bootup asm.
We try to fetch the CIF entry pointer from %o4, but that
can get clobbered by the early OBP calls. It is saved
in %l7 already, so actually this "mov %o4, %l7" can just
be completely removed with no other changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 18 Jul 2007 22:15:45 +0000 (15:15 -0700)]
[SPARC64]: Fix handling of multiple vdc-port nodes.
The "id" property in vdc-port nodes are not unique, they
are all zero. Therefore assign ID's using the parent's
"cfg-handle" property which will be unique.
Signed-off-by: David S. Miller <davem@davemloft.net>
Paul Mundt [Fri, 20 Jul 2007 01:11:58 +0000 (10:11 +0900)]
mm: Remove slab destructors from kmem_cache_create().
Slab destructors were no longer supported after Christoph's c59def9f222d44bb7e2f0a559f2906191a0862d7 change. They've been
BUGs for both slab and slub, and slob never supported them
either.
This rips out support for the dtor pointer from kmem_cache_create()
completely and fixes up every single callsite in the kernel (there were
about 224, not including the slab allocator definitions themselves,
or the documentation references).
Alan Stern [Wed, 18 Jul 2007 16:14:24 +0000 (12:14 -0400)]
USB: change name of spinlock in hcd.c
This patch (as940 renames hcd_data_lock in hcd.c to hcd_urb_list_lock,
which is more descriptive of the lock's job. It also introduces a
convenient inline routine for testing whether a particular USB device
is a root hub.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Wed, 18 Jul 2007 16:08:02 +0000 (12:08 -0400)]
USB: move routines in hcd.c
This patch (as939) moves a couple of routine in hcd.c around. The
purpose is to put all the general URB- and endpoint-related routines
(submit, unlink, giveback, and disable) together in one spot.
There are no functional changes.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch incorporates some updates from the review of the
Renesas m66592-udc driver. Updates include:
- Fix some locking bugs; and add a few sparse annotations
- Don't #define __iomem !
- Lots of whitespace fixes (most of the patch by volume)
- Some #include file trimmage
- Other checkpatch.pl and sparse updates
- Alphabetized and slightly-more-informative Kconfig
- Don't use the ID which was assigned to the amd5536udc driver.
- Remove pointless suspend/resume methods updating obsolete field.
- Some section fixups
- Fix some leak bugs
- Fix byteswapping
It's complaining that we dereference 'edge_port' in line 2799 which
makes the test of that pointer against NULL in 2800 pointless, since if
edge_port was actually NULL we'd have crashed already before reaching
line 2800.
Reading the edge_open() function it seems to me that the pointer
returned by usb_get_serial_port_data(serial->port[i]) and stored in
'edge_port' can never actually be NULL here, so the test is entirely
superfluous (even if it could be NULL it would be pointless here,
ignoring the then possible crash in that case, since both
edge_buf_free() and kfree() can handle being passed NULL pointers.
This patch removes the pointless conditional (and also makes a few
tiny style corrections now that I was in the area anyway).
Thomas Dahlmann [Tue, 17 Jul 2007 04:40:54 +0000 (21:40 -0700)]
USB: amd5536 UDC driver (in GEODE southbridge)
Driver for the AMD5536 UDC, as found in the AMD Geode CS5536 (southbridge).
This is a high speed DMA-capable controller, which can also be used in
OTG configurations (which are not supported by this patch).
Acked-by: Jordan Crouse <jordan.crouse@amd.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Fri, 13 Jul 2007 19:47:16 +0000 (15:47 -0400)]
USB: documentation update for usb_unlink_urb
This patch (as936) updates the kerneldoc for usb_unlink_urb. The
explanation of how endpoint queues are meant to work is now clearer
and in better agreement with reality.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Fri, 13 Jul 2007 19:46:29 +0000 (15:46 -0400)]
UHCI: short control URBs get a status stage
It has recently been pointed out that short control transfers should
have a status stage, even if they generate an error because
URB_SHORT_NOT_OK was set. This patch (as935) changes uhci-hcd to
enable the status stage when this happens.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Coverity (1709, 1710, 1711, 1712, 1713) actually flagged these as
REVERSE_INULLs (NULL check performed after dereference). But looking at
the other drivers I can't see any similar tests and the USB core already
makes sure urb is non-null - so might as well get rid of the checks.
Alan Stern [Mon, 16 Jul 2007 19:28:19 +0000 (15:28 -0400)]
USB: fix warning caused by autosuspend counter going negative
This patch (as937) fixes a minor bug in the autosuspend usage-counting
code. Each hub's usage counter keeps track of the number of
unsuspended children. However the current driver increments the
counter after registering a new child, by which time the child may
already have been suspended and caused the counter to go negative.
The obvious solution is to increment the counter before registering
the child.
Alan Stern [Thu, 12 Jul 2007 21:06:23 +0000 (17:06 -0400)]
USB: add "descriptors" binary sysfs attribute
This patch (as934) adds a new readonly binary sysfs attribute file
called "descriptors" for each USB device. The attribute contains the
device descriptor followed by the raw descriptor entry (config plug
subsidiary descriptors) for the current configuration.
Having this information available in fixed-format binary makes life a
lot easier for user programs by avoiding the need to open, read, and
parse multiple sysfs text files.
The information in this attribute file is much like that in usbfs's
device file, but there are some significant differences:
The 2-byte fields in the device descriptor are left in
little-endian byte order, as they appear on the bus and
in the kernel.
Only one raw descriptor set is presented, that of the
current configuration.
Opening this file will not cause a suspended device to be
autoresumed.
The last item in particular should be a big selling point for libusb,
which currently forces all USB devices to be resumed as it scans the
device tree.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Cc: Dave Mielke <dave@mielke.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Alan Stern [Thu, 12 Jul 2007 21:03:01 +0000 (17:03 -0400)]
isp116x-hcd: prepare for urb->status
This patch (as931b), adapted from a patch by Olav Kongas, makes a small
set of conservative changes to the isp116x-hcd driver in preparation
for the removal of urb->status.
finish_request() is moved up in the source and is called
as soon as the URB is known to have completed, rather than
after all the active endpoints have been scanned.
The status of a completed URB is kept in a local variable
and copied to urb->status only when the URB is about to be
given back.
-EREMOTEIO error status for control transfers is set after
the status stage rather than when the short packet arrives.
Some unnecessary uses of urb->lock are removed.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Cc: Olav Kongas <ok@artecdesign.ee> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Pete Zaitcev [Wed, 11 Jul 2007 03:09:58 +0000 (20:09 -0700)]
USB: usblp: "Big cleanup" breaks O_NONBLOCK
I found the first regresson in the rewritten ("all dynamic" and "no races")
driver. If application uses O_NONBLOCK, I return -EAGAIN despite the URB
being submitted successfuly. This causes the application to resubmit the
same data erroneously.
The fix is to pretend that the transfer has succeeded even if URB was
merely queued. It is the same behaviour as with the old version.
Signed-off-by: Pete Zaitcev <zaitcev@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Kevin Lloyd [Mon, 16 Jul 2007 20:49:29 +0000 (13:49 -0700)]
USB: sierra: Add new devices
This patch adds new devices to the Sierra Wireless driver. This is being
resubmitted because the dependent patch (patch 01/02) needed to be
resubmitted.
Signed-off-by: Kevin Lloyd <linux@sierrawireless.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Kevin Lloyd [Mon, 16 Jul 2007 20:49:27 +0000 (13:49 -0700)]
USB: sierra: Add TRU-Install (c) Support
This patch adds compatibility with Sierra Wireless' new TRU-Install
feature. Future devices that use this feature will not work unless this
patch has been applied.
Signed-off-by: Kevin Lloyd <linux@sierrawireless.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Al Viro [Thu, 19 Jul 2007 23:23:31 +0000 (00:23 +0100)]
coda breakage
a) switch by loff_t == __cmpdi2 use. Replaced with a couple
of obvious ifs; update of ->f_pos in the first one makes sure that we
do the right thing in all cases.
b) block_signals() and unblock_signals() are globals on UML.
Renamed coda ones; in principle UML probably ought to do rename as
well, but that's another story.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that ide-floppy supports SG_IO we can add support for SCSI ioctls
(except deprecated SCSI_IOCTL_SEND_COMMAND and legacy CDROM_SEND_PACKET
ones - we can add them later iff really needed).
While at it remove handling of CDROMEJECT and CDROMCLOSETRAY ioctls from
generic_ide_ioctl():
- This prevents ide-{disk,tape,scsi} device drivers from obtaining
REQ_TYPE_BLOCK_PC type requests which are currently unsupported by
these drivers and which are potentially harmful (as reported by Andrew).
- There is no functionality loss since aforementioned ioctls will now be
handled by idefloppy_ioctl()->scsi_cmd_ioctl() (for devices using
ide-floppy driver) and by idecd_ioctl->cdrom_ioctl()->scsi_cmd_ioctl()
(for devices using ide-cd driver).
Cc: Jens Axboe <jens.axboe@oracle.com> Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: James Bottomley <James.Bottomley@steeleye.com> Cc: Jeff Garzik <jeff@garzik.org> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* Add ->host_flags to ide_hwif_t to store ide_pci_device_t.host_flags,
assign it in setup-pci.c:ide_pci_setup_ports().
* Add IDE_HFLAG_PIO_NO_{BLACKLIST,DOWNGRADE} to ide_pci_device_t.host_flags
and teach ide_get_best_pio_mode() about them. Also remove needless
!drive->id check while at it (drive->id is always present).
* Convert amd74xx, via82cxxx and ide-timing.h to use ide_get_best_pio_mode()
and then remove no longer needed ide_find_best_pio_mode().
There should be no functionality changes caused by this patch.
* Some devices claim maximum PIO mode > 2 in id->tPIO, they were punished too
severly for this by being limited to PIO_SLOW. Limit them to PIO2 instead.
v2:
* Fix PIO number being returned incorrectly instead of PIO mode
(Noticed by Sergei).
Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com> Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* Use it in ali14xx/ht6560b/qd65xx/cmd64{0,x}/sl82c105 and pmac host drivers
(previously cycle time given by the device was only used for "pio" == 255).
* Remove no longer needed ide_pio_data_t.cycle_time field.
v2:
* Fix "ata_" prefix (Noticed by Jeff).
Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* Nowadays core IDE code handles restoring of PIO and DMA modes
(ide-io.c:ide_start_power_step() etc) so remove open-coded version
from sc1200_resume().
There should be no change in behavior because settings done by
sc1200_resume() were always overridden by generic_ide_resume()
and ide_{start,stop}_power_step().
* Bump driver version.
Cc: Mark Lord <mlord@pobox.com> Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* Do the same thing as probe_hwif() and always disable DMA so chipset DMA
enabled bit gets cleared (if the drive doesn't support DMA ide_set_dma()
won't try to tune it anyway).
* Add TODO comment about respecting ->using_dma setting.
The driver used to depend on BIOS settings for deciding whether it is OK
to use DMA. However it seems that BIOS doesn't always handle all cases
correctly so just let IDE core to decide about this. It should be a safe
thing to do now, after the driver went through heavy bugfixing.
Thanks for bugreport and testing the patch goes out to Sven Niedner.
* Always set ->autotune so PIO gets correctly auto-tuned (previously
->autotune was only set when ->dma_base wasn't available, however
->ide_dma_check()/->speedproc() was always trying to tune PIO when
tuning DMA).
* Move code responsible for programming chipset for PIO mode from
svwks_tune_chipset() to svwks_tune_pio(). Don't tune PIO when tuning
DMA (this is no longer needed since ->autotune is always set now).
* Handle PIO modes early in svwks_tune_chipset() so DMA configuration
registers don't get cleared when programming PIO mode.
projects / linux-2.6 / log