| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Moved i_next_alloc_block and i_next_goal_block out from ext3_inod_info, and
put it together with the reservation structure into the
ext3_block_alloc_info structure, and dynamically allocate that structure
whenever need to allocation a block. This is also apply for noreservation
mount. Also cleanup ext3_find_goal() code.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Since now the ei->truncate_sem is guarding the concurrent allocation and
the deallocation, there is no need to use the the rsv_seqlock lock in the
ext3_reserve_window_node, which was there to protect using/allocating
reservation window race between two threads allocating blocks at the same
time.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Right now the ext3 reservation structure(ext3_reserve_window_node) is part of
the ext3 inode itself. This part of information is only needed for files that
need allocate blocks on disk. So, the attached patches reduce the ext3 inode
size by dynamically allocating the block allocation/reservation info
structure(called struct ext3_block_alloc_info) when it is needed(i.e. only
for files who need to allocate blocks)
The reservation structure is being allocated and linked to the ext3 inode at
ext3_get_block_handle(), and being freed and unlinked at the
iput_final->ext3_clear_inode().
The ei->truncate_sem which is currently used to protect concurrent
ext3_get_block() and ext3_truncate is used to protect reservation structure
allocation and deallocation.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
The attached patch improves nommu mmap support; particularly in terms on
supporting private mappings. It does this by examining the device capability
mask now in the backing_dev_info structure.
Private mappings will now be backed by the underlying device directly if
possible, where "possible" is constrained by the protection mask parameter
and the device capabilities mask.
I've also split the do_mmap_pgoff() function contents into a number of
auxilliary functions to make it easier to understand.
The documentation is also updated; including the addition of a warning
about permitting direct mapping of flash chips and the problems of XIP vs
write.
Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
into ppc970.osdl.org:/home/torvalds/v2.6/linux
|
|
* move ->disk from ide_drive_t to driver specific objects
* make drivers allocate struct gendisk and setup rq->rq_disk
(there is no need to do this for REQ_DRIVE_TASKFILE requests)
* add ide_init_disk() helper and kill alloc_disks() in ide-probe.c
* kill no longer needed ide_open() and ide_fops[] in ide.c
ide_init_disk() fixed by Andrew Morton <akpm@osdl.org>.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
|
|
Add ide_{un}register_region() and fix ide-{tape,scsi}.c to register
block device number ranges. In ata_probe() only probe for modules.
Behavior is unchanged because:
* if driver is already loaded and attached to drive ata_probe()
is not called et all
* if driver is loaded by ata_probe() it will register new number range
for a drive and this range will be found by kobj_lookup()
If this is not clear please read http://lwn.net/Articles/25711/
and see drivers/base/map.c.
This patch makes it possible to move drive->disk allocation from
ide-probe.c to device drivers.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
|
|
into ppc970.osdl.org:/home/torvalds/v2.6/linux
|
|
|
|
So far, the functionality handling of i2c adapters was done in i2c-core
by two exported functions: i2c_get_functionality and
i2c_check_functionality. I found that both functions could be reduced to
one line each, and propose that we turn them into inline function in the
i2c.h header file, much like other i2c helper functions (e.g.
i2c_get_clientdata, i2c_set_clientdata and i2c_clientname).
The conversion of i2c_get_functionality suppresses a legacy check which
shouldn't be needed anymore. Only one driver (s3c2410) was still relying
on it, and was fixed some days ago.
The conversion lets us get rid of two exports. Not only i2c-core gets
smaller (by 458 bytes), but the client drivers using these functions get
smaller too (typically by 48 bytes). And of course the new way is likely
to be faster too, even if it wasn't my primary objective.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
|
|
into ppc970.osdl.org:/home/torvalds/v2.6/linux
|
|
into ppc970.osdl.org:/home/torvalds/v2.6/linux
|
|
Remove i_sock from struct inode. Also remove some checks for SOCKET_I()
returning NULL -- it can never return NULL for a valid inode.
Signed-off-by: Matthew Wilcox <matthew@wil.cx>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
into sunset.davemloft.net:/home/davem/src/BK/net-2.6
|
|
Some defines in i2c.h (I2C_CLIENT_MODPARM and friends) are now useless.
They should have been removed when the i2c client parameters were
converted from MODULE_PARAM to module_parm_array, but where not. This
patch removes them now.
Additionally, it moves the definition of I2C_CLIENT_MAX_OPTS next to
where it is used rather than 220 lines before, which is preferable IMHO.
As a side note, I think that there is a bug in the way these options are
handled. The i2c code looks for I2C_CLIENT_END as a list terminator, but
if the maximum number of parameters are actually provided, no terminator
will be left. It's rather unlikely to happen because nobody will
probably ever provide that many parameters, but this should probably be
fixed. I'll address this issue later, since I plan to completely rewrite
the way these parameters are handled anyway.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
|
|
into sunset.davemloft.net:/home/davem/src/BK/net-2.6
|
|
|
|
The use of auth_domains is somewhat confusing, in part because they were
originally intended to be used in a more general way than they currently are.
Update the documentation a little with an eye towards how it's currently used.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@cse.unsw.edu.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This patch contains cleanups under drivers/video/ including:
- make some needlessly global code static
- the following was needlessly EXPORT_SYMBOL'ed:
- fbcon.c: fb_con
- fbmon.c: get_EDID_from_firmware (completely unused)
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Antonino Daplas <adaplas@pol.net>
Signed-off-by: Benoit Boissinot <benoit.boissinot@ens-lyon.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
initialized
Accessing the hardware before it is properly initialized can lead to crashes
or screen corruption. This happens when switching to X then back to console.
When console comes back from X, the device is in an undefined state. During
this window, accessing the hardware is disallowed.
A new field in fbcon_par is added (graphics), which will be set to nonzero
just before initialization of the framebuffer and when coming back from
KD_GRAPHICS, then unset when an fb_set_var/fb_set_par is done. While this
field is set, no accesses to the hardware is done. The consequence of this
change is, hopefully, more robust switching between KD_GRAPHICS<-> KD_TEXT.
An added benefit coming from this change is that the MODESWITCHLATE hack is
not needed anymore and thus removed. This hack is used by savagefb, rivafb
and nvidiafb.
Signed-off-by: Antonino Daplas <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This patch against 12-rc1 adds seccomp to the ppc64 arch. I tested it
successfully with the seccomp_test. I didn't bother to change the syscall
exit not to check for TIF_SECCOMP, in theory that bit could be optimized
but it's an optimization in the slow path, and current code is a bit
simpler. I also verified it still compiles and works fine on x86 and
x86-64.
Instead of the TIF_32BIT redefine, if you want to change x86-64 to use
TIF_32BIT too (instead of TIF_IA32), let me know.
Signed-off-by: Andrea Arcangeli <andrea@cpushare.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
The attached patch replaces backing_dev_info::memory_backed with capabilitied
bitmap. The capabilities available include:
(*) BDI_CAP_NO_ACCT_DIRTY
Set if the pages associated with this backing device should not be
tracked by the dirty page accounting.
(*) BDI_CAP_NO_WRITEBACK
Set if dirty pages associated with this backing device should not have
writepage() or writepages() invoked upon them to clean them.
(*) Capability markers that indicate what a backing device is capable of
with regard to memory mapping facilities. These flags indicate whether a
device can be mapped directly, whether it can be copied for a mapping,
and whether direct mappings can be read, written and/or executed. This
information is primarily aimed at improving no-MMU private mapping
support.
The patch also provides convenience functions for determining the dirty-page
capabilities available on backing devices directly or on the backing devices
associated with a mapping. These are provided to keep line length down when
checking for the capabilities.
Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
into ppc970.osdl.org:/home/torvalds/v2.6/linux
|
|
1. Change of Copyright to reflect S2io's new name "Neterion Inc."
2. Updated driver version number.
3. Add an additional PCI device id to support Neterion's new Xframe II.
Some background info on Xframe-II, just in case -
The NIC was announced back in January (see
http://www.serverwatch.com/news/article.php/3464871)
Xframe II is a PCI-X 2.0 DDR adapter designed to work in PCI-X 2.0 servers
(it is also backwards compatible with current pci-x and pci slots); it's a
first card to overcome pci-x 133 throughput bottleneck. Some of these pci-x
2.0 servers are available now and more are coming later in the year.
Xframe II is completely backward compatible and hence the current driver
will work as-is, except the device id change.
There are some additional features/stateless offloads in Xframe-II
as well; we plan to submit patches for these soon.
Signed-off-by: Ravinandan Arakali <ravinandan.arakali@neterion.com>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
|
|
into suse.de:/home/greg/linux/BK/usb-2.6
|
|
This fixes u32 vs. pm_message_t confusion in MMC layer.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This fixes u32 vs. pm_message_t in generic code. No code changes.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
|
|
Add a `nobh' mount option to ext3 in writeback mode: avoid attaching
buffer_head to data pages, like ext2.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Add writepages support for ext3 writeback mode.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Fix destruction of in-use journal_head
journal_put_journal_head() can destroy a journal_head at any time as
long as the jh's b_jcount is zero and b_transaction is NULL. It has no
locking protection against the rest of the journaling code, as the lock
it uses to protect b_jcount and bh->b_private is not used elsewhere in
jbd.
However, there are small windows where b_transaction is getting set
temporarily to NULL during normal operations; typically this is
happening in
__journal_unfile_buffer(jh);
__journal_file_buffer(jh, ...);
call pairs, as __journal_unfile_buffer() will set b_transaction to NULL
and __journal_file_buffer() re-sets it afterwards. A truncate running
in parallel can lead to journal_unmap_buffer() destroying the jh if it
occurs between these two calls.
Fix this by adding a variant of __journal_unfile_buffer() which is only
used for these temporary jh unlinks, and which leaves the b_transaction
field intact so that we never leave a window open where b_transaction is
NULL.
Additionally, trap this error if it does occur, by checking against
jh->b_jlist being non-null when we destroy a jh.
Signed-off-by: Stephen Tweedie <sct@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
There's a bugzilla entry for this here:
http://bugzilla.kernel.org/show_bug.cgi?id=4334
This should fix all the problems I know about with Netmos combo cards:
- 9735, 9835, and 9855 are not supported
- combo cards with parallel are erroneously claimed by serial driver
- serial and parport_serial blindly probe for ports
parport_pc:
Sort Netmos device IDs, no functional change.
parport_serial:
Previously supported 9735 and 9835. Add 9745, 9845, 9855, and
add init hooks to discover how many serial/parallel ports are
actually present (the boards are available in various configs).
Add protection for overflow of static tables.
quirks:
Detect Netmos combo (parallel + serial) cards and change class from
SERIAL to OTHER to prevent serial driver from claiming them.
8250:
Add init hook to discover the number of serial ports present.
This prevents us from poking at unused BARs.
pci_ids:
Add Netmos 9745, 9845, and sort.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This device driver is for the Digi Neo serial adapter. We got several rounds
of detail comments from linux community and updated our earlier patches
according to comments including Jeff Garzik and Greg KH's.
Signed-off-by: Wen Xiong <wendyx@us.ltcfwd.linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This patch converts all save_flags/restore_flags to the new
spin_lock_irqsave/spin_unlock_irqrestore calls, as well as some other 2.6.X
cleanups. This prepares the way for the "io8+", "sx" and "rio" drivers to
become SMP safe. Patches for those drivers follow.
Signed-off-by: Patrick vd Lageweg <patrick@bitwizard.nl>
Signed-off-by: Rogier Wolff <R.E.Wolff@BitWizard.nl>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Eliminates "warning: type qualifiers ignored on function return type".
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Needed this to build Fedora rawhide kernel (2.6.12-rc1 + some patches) on
alpha. This is the upstream portion of the build fixes.
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
CON_BOOT is like early printk in that it allows for output really early on.
It's better than early printk because it unregisters automatically when a
real console is initialised. So if you don't get consoles registering in
console_init, there isn't a huge delay between the boot console
unregistering and the real console starting. This is the case on PA-RISC
where we have serial ports that aren't discovered until the PCI bus has
been walked.
I think all the current early printk users could be converted to this
scheme with a minimal amount of effort.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Extend the gcc printk format-string checking.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
I made a patch to update m32r_sio driver.
Compile checked and tested on a M32700UT eva board with SMP kernel.
It looks working...
__register_m32r_sio, register_m32r_sio, unregister_m32r_sio are still
remained. I'm going to take a look them after.
Signed-off-by: Hirokazu Takata <takata@linux-m32r.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This kills unused macro and write-only variable, and adds messages where
something goes wrong with suspending devices.
Signed-off-by: Pavel Machek <pavel@suse.cz>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This patch extracts all the operations on counters protected by the page
table lock (currently rss and anon_rss) into definitions in
include/linux/sched.h. All rss operations are performed through the
following macros:
get_mm_counter(mm, member) -> Obtain the value of a counter
set_mm_counter(mm, member, value) -> Set the value of a counter
update_mm_counter(mm, member, value) -> Add to a counter
inc_mm_counter(mm, member) -> Increment a counter
dec_mm_counter(mm, member) -> Decrement a counter
With this patch it becomes easier to add new counters and it is possible to
redefine the method of counter handling. The counters are an issue for
scalability since they are used in frequently used code paths and may cause
cache line bouncing.
F.e. One may not use counters at all and count the pages when needed, switch
to atomic operations if the mm_struct locking changes or split the rss
into counters that can be locally incremented.
The relevant fields of the task_struct are renamed with a leading underscore
to catch out people who are not using the acceessor macros.
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
When testing AIO on PPC64 (a power5 machine) running 2.6.11 with
CONFIG_HUGETLB_PAGE=y, I ran into a kernel panic when a process exits that has
done AIO (io_queue_init()) but has not done the io_queue_release(). The
exit_aio() code is cleaning up and panicking when trying to free the aio ring
buffer.
I tracked this down to is_hugepage_only_range() (include/asm-ppc64/page.h)
which is doing a touches_hugepage_low_range() which is checking
current->mm->context.htlb_segs. The problem is that exit_mm() cleared tsk->mm
before doing the mmput() which leads to the exit_aio() and then the panic.
Looks like is_hugepage_only_range() is only used in ia64 and ppc64. Fix is to
change is_hugepage_only_range() to take an 'mm' as a parameter as well as
'addr' and 'len' and then the ppc64 code could change to use 'mm'. It looks
like it has been broken for quite a while.
Signed-off-by: Daniel McNeil <daniel@osdl.org>
Acked-by: David Gibson <dwg@au1.ibm.com>
Acked-by: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This makes it hard(er) to mix argument orders by mistake for things like
kmalloc() and friends, since silent integer promotion is now caught by
sparse.
|
|
where implicit casts are not appropriate.
This can be used to mark unsigned integers as being uncastable
to signed, for example. Or enums as not degrading to integers
and vice versa.
|
|
sk_alloc_slab becomes proto_register, that receives a struct proto not necessarily
completely filled, but at least with the proto name, owner and obj_size (aka proto
specific sock size), with this we can remove the struct sock sk_owner and sk_slab,
using sk->sk_prot->{owner,slab} instead.
This patch also makes sk_set_owner not necessary anymore, as at sk_alloc time we
have now access to the struct proto onwer and slab members, so we can bump the
module refcount exactly at sock allocation time.
Another nice "side effect" is that this patch removes the generic sk_cachep slab
cache, making the only last two protocols that used it use just kmalloc, informing
a struct proto obj_size equal to sizeof(struct sock).
Ah, almost forgot that with this patch it is very easy to use a slab cache, as it is
now created at proto_register time, and all protocols need to use proto_register,
so its just a matter of switching the second parameter of proto_register to '1', heck,
this can be done even at module load time with some small additional patch.
Another optimization that will be possible in the future is to move the sk_protocol
and sk_type struct sock members to struct proto, but this has to wait for all protocols
to move completely to sk_prot.
This changeset also introduces /proc/net/protocols, that lists the registered protocols
details, some may seem excessive, but I'd like to keep them while working on further
struct sock hierarchy work and also to realize which protocols are old ones, i.e. that
still use struct proto_ops, etc, yeah, this is a bit of an exaggeration, as all protos
still use struct proto_ops, but in time the idea is to move all to use sk->sk_prot and
make the proto_ops infrastructure be shared among all protos, reducing one level of
indirection.
Signed-off-by: Arnaldo Carvalho de Melo <acme@conectiva.com.br>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
into ppc970.osdl.org:/home/torvalds/v2.6/linux
|
|
into sunset.davemloft.net:/home/davem/src/BK/net-2.6
|
|
into ppc970.osdl.org:/home/torvalds/v2.6/linux
|
|
This removes the duplicate of uart_match_port in 8250.c
Signed-off-by: Russell King <rmk@arm.linux.org.uk>
|
|
Action stats were being sent on the wrong TLV. Patch attached.
Thanks to Andy Furniss for finding and helping debug this.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
