Currently looking up a structure definition in TAGS / tags takes one to
one of multiple "static struct X" definitions in arch sources, which makes
it for many structs practically impossible to get to the required header.
This patch changes the order of sources being tagged to first scan
architecture includes, then the top-level include/ directory, and only
then the rest. It also takes into account, that many architectures have
more than one include directory, i.e., not only arch/$ARCH/include, but
also arch/$ARCH/mach-X/include etc.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Reviewed-by: WANG Cong <xiyou.wangcong@gmail.com>
[mmarek@suse.cz: fix 'var+=text' bashism]
Signed-off-by: Michal Marek <mmarek@suse.cz>
Better dwarf2 unwind information is a good thing, it allows better
debugging with kgdb and crash and helps systemtap.
Commit 003086497f ("Build with
-fno-dwarf2-cfi-asm") disabled some CFI information globally to work
around a module loader bug on powerpc.
But this disables the better unwind tables for all architectures, not just
powerpc. Move the workaround to powerpc and also add a suitable comment
that's it really a workaround.
This improves dwarf2 unwind tables on x86 at least.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Michal Marek <mmarek@suse.cz>
1. Fix a little format issue.
2. Check the return of "Getopt::Long::GetOptions". Output usage and
exit if it get error.
3. Change $ARGV[$#ARGV] to $ARGV[0].
4. Change the code which get $modulefile from modinfo. Replace the
pipeline with `modinfo -F filename $module`.
4. Change usage from "Specify the module directory name" to "Specify the
module filename".
Signed-off-by: Hui Zhu <teawater@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
The markup_oops.pl have 3 troubles to support cross-compiler environment:
1. It use objdump directly.
2. It use modinfo to get the message of module.
3. It use hex function that cannot support 64-bit number in 32-bit arch.
This patch add 3 options to markup_oops.pl:
1. -c CROSS_COMPILE Specify the prefix used for toolchain.
2. -m MODULE_DIRNAME Specify the module directory name.
3. Change hex function to Math::BigInt->from_hex.
After this patch, parse the x8664 oops in x86, we can:
cat amd64m | perl ~/kernel/tmp/m.pl -c /home/teawater/kernel/bin/x8664- -m ./e.ko vmlinux
Thanks,
Hui
Signed-off-by: Hui Zhu <teawater@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: ozan@pardus.org.tr
Cc: Matthew Wilcox <willy@linux.intel.com>
Acked-by: WANG Cong <xiyou.wangcong@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Also, add a note that "unmaintained" files below scripts/ should go via
the kbuild tree (best current practice).
Signed-off-by: Michal Marek <mmarek@suse.cz>
The mkspec script hardcodes "/var/tmp" into the generated rpm spec file's
BuildRoot. The user, however, may have a custom setting for %_tmppath,
which should be used in BuildRoot. This patch changes mkspec's
BuildRoot output to appropriately use %_tmppath.
Signed-off-by: John Saalwaechter <saalwaechter@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
I got a "No matching code found" when I use markup_oops.pl parse a error
in a x86_64 module.
cat e.c
int init_module(void)
{
char *buf = 0;
buf[0] = 3;
return 0;
}
void cleanup_module(void)
{
//char *buf = 0;
//buf[0] = 3;
}
MODULE_AUTHOR("Hui Zhu");
MODULE_LICENSE("GPL");
0000000000000000 <init_module>:
init_module():
/home/teawater/study/kernel/stack2core/example/e.c:10
0: c6 04 25 00 00 00 00 movb $0x3,0x0
7: 03
/home/teawater/study/kernel/stack2core/example/e.c:13
8: 31 c0 xor %eax,%eax
a: c3 retq
b: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
0000000000000010 <cleanup_module>:
cleanup_module():
/home/teawater/study/kernel/stack2core/example/e.c:20
10: f3 c3 repz retq
12: 90 nop
13: 90 nop
Disassembly of section .modinfo:
This is because the faulting instruction "movb $0x3,0x0" is the first
line of the range.
In the markup_oops.pl:
main::(./scripts/markup_oops.pl:245):
245: if (InRange($1, $target)) {
DB<2> p $line
ffffffffa001b000: c6 04 25 00 00 00 00 movb $0x3,0x0
DB<3> p $counter
0
It just set $center in next loop. So it cannot get the $center.
And even if $center is set to the right value 0.
if ($center == 0) {
print "No matching code found \n";
exit;
}
The first line $center will be 0, so I change the default value to -1.
Signed-off-by: Hui Zhu <teawater@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Just a small change to a couple of scripts to go from
#!/usr/bin/env python
to
#!/usr/bin/python
This shouldn't effect anyone, unless they don't install python there.
In preparation for python3, Fedora is doing a big push to change the scripts
to use the system python. This allows developers to put the python3 in
their path without fear of breaking existing scripts.
Now I am pretty sure anyone using python3 for testing purposes will probably
not run any of the scripts I changed, but Fedora has this automated tool
that checks for this stuff so I thought I would try to push it upstream.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Acked-by: WANG Cong <xiyou.wangcong@gmail.com>
Acked-by: Jarod Wilson <jarod@redhat.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Suppress a warn_unused_result warning.
fgets is called as a part of error handling. It is called just to drop a
line and return immediately. read_map is reading the file in a loop and
read_symbol reads line by line. So I think there is no point in using
return value for useful checking. Other checks like 3 items were returned
or !EOF have already been done.
Signed-off-by: Himanshu Chauhan <hschauhan@nulltrace.org>
Cc: WANG Cong <xiyou.wangcong@gmail.com>
Cc: Michal Marek <mmarek@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Don't test for /bin/{dnsdomainname,domainname}, simply try to execute
the command and check if it returned something.
Reported-by: Glenn Sommer <glemsom@gmail.com>
Acked-by: WANG Cong <xiyou.wangcong@gmail.com>
Tested-by: Glenn Sommer <glemsom@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
While looking for something else I noticed that the symbol
hash function used by kconfig is quite poor. It doesn't
use any of the standard hash techniques but simply
adds up the string and then uses power of two masking,
which is both known to perform poorly.
The current x86 kconfig has over 7000 symbols.
When I instrumented it showed that the minimum hash chain
length was 16 and a significant number of them was over
30.
It didn't help that the hash table size was only 256 buckets.
This patch increases the hash table size to a larger prime
and switches to a FNV32 hash. I played around with a couple of hash
functions, but that one seemed to perform best with reasonable
hash table sizes.
Increasing the hash table size even further didn't
seem like a good idea, because there are a couple of global
walks which walk the complete hash table.
I also moved the unnamed bucket to 0. It's still the longest
of all the buckets (44 entries), but hopefully it's not
often hit except for the global walk which doesn't care.
The result is a much nicer distribution:
(first column bucket length, second number of buckets with that length)
1: 3505
2: 1236
3: 294
4: 52
5: 3
47: 1 <--- this is the unnamed symbols bucket
There are still some 5+ buckets, but increasing the hash table
even more would be likely not worth it.
This also cleans up the code slightly by removing hard coded
magic numbers.
I didn't notice a big performance difference either way
on my Nehalem system, but I presume it'll help somewhat
on slower systems.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
The LOCALVERSION= string passed to "make" will now always be appended to
the kernel version after CONFIG_LOCALVERSION, if it exists, regardless of
whether CONFIG_LOCALVERSION_AUTO is set or not. This allows users to
uniquely identify their kernel builds with a string.
If CONFIG_LOCALVERSION_AUTO is enabled, the unique SCM tag reported by
setlocalversion (or .scmversion) is appended to the kernel version, if it
exists. When CONFIG_LOCALVERSION_AUTO is not enabled, a `+' is appended
to the kernel version to represent that the kernel has been revised since
the last release unless "make LOCALVERSION=" was used to uniquely identify
the build.
The end result is this:
- when LOCALVERSION= is passed to "make", it is appended to the kernel
version,
- when CONFIG_LOCALVERSION_AUTO is enabled, a unique SCM identifier is
appended if the respository has been revised beyond a tagged commit,
and
- when CONFIG_LOCALVERSION_AUTO is disabled, a `+' is appended if the
repository has been revised beyond a tagged commit and LOCALVERSION=
was not passed to "make".
Examples:
With CONFIG_LOCALVERSION_AUTO: "make" results in
v2.6.32-rc4-00149-ga3ccf63. If there are uncommited changes to the
respository, it results in v2.6.32-rc4-00149-ga3ccf63-dirty. If
"make LOCALVERSION=kbuild" were used, it results in
v2.6.32-rc4-kbuild-00149-ga3ccf63-dirty.
Without CONFIG_LOCALVERSION_AUTO, "make" results in v2.6.32-rc4+
unless the repository is at the Linux v2.6.32-rc4 commit (in which
case the version would be v2.6.32-rc4). If "make LOCALVERSION=kbuild"
were used, it results in v2.6.32-rc4-kbuild.
Also renames variables such as localver-auto and _localver-auto to more
accurately describe what they represent: localver-extra and
scm-identifier, respectively.
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
This patch fixes two problems reported by Jan Engelhardt:
1) Border is now properly placed, to always be visible
2) Long menu items are properly displayed
Reported-by: Jan Engelhardt <jengelh@medozas.de>
Signed-off-by: Nir Tzachar <nir.tzachar@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
scripts/kconfig/nconf.gui.c:23: warning: no previous prototype for 'set_normal_colors'
scripts/kconfig/nconf.gui.c:68: warning: no previous prototype for 'normal_color_theme'
scripts/kconfig/nconf.gui.c💯 warning: no previous prototype for 'no_colors_theme'
scripts/kconfig/nconf.c:455: warning: no previous prototype for 'process_special_keys'
scripts/kconfig/nconf.c:487: warning: no previous prototype for 'get_next_hot'
scripts/kconfig/nconf.c:506: warning: no previous prototype for 'canbhot'
scripts/kconfig/nconf.c:514: warning: no previous prototype for 'is_hot'
scripts/kconfig/nconf.c:522: warning: no previous prototype for 'make_hot'
scripts/kconfig/nconf.c:582: warning: no previous prototype for 'item_make'
scripts/kconfig/nconf.c:626: warning: no previous prototype for 'item_add_str'
scripts/kconfig/nconf.c:656: warning: no previous prototype for 'item_tag'
scripts/kconfig/nconf.c:668: warning: no previous prototype for 'curses_item_index'
scripts/kconfig/nconf.c:673: warning: no previous prototype for 'item_data'
scripts/kconfig/nconf.c:684: warning: no previous prototype for 'item_is_tag'
scripts/kconfig/nconf.c:691: warning: no previous prototype for 'set_config_filename'
Signed-off-by: Michal Marek <mmarek@suse.cz>
This patch was inspired by the kernel projects page, where an ncurses
replacement for menuconfig was mentioned (by Sam Ravnborg).
Building on menuconfig, this patch implements a more modern look
interface using ncurses and ncurses' satellite libraries (menu, panel,
form). The implementation does not depend on lxdialog, which is
currently distributed with the kernel.
Signed-off-by: Nir Tzachar <nir.tzachar@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
It is the last place when the file is read, so close it.
Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Help text for certain config options is very extensive (the text
includes the names of all other options the option in question depends
on). Long lines are not wrapped, making it impossible to see the list
without scrolling horizontally.
This patch adds some logic which wraps help screen lines at word
boundaries to prevent truncating.
Tested by running
ARCH=powerpc make menuconfig O=/tmp/build
which shows that the long lines are now wrapped, and
ARCH=powerpc make xconfig O=/tmp/build
to demonstrate that it still compiles and operates as expected.
Signed-off-by: Vadim Bendebury <vbendeb@google.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
This patch adds support for decoding ARM oopses to scripts/decodecode.
The following things are handled:
- ARCH and CROSS_COMPILE environment variables are respected.
- The Code: in x86 oopses is in bytes, while it is in either words (4
bytes) or halfwords for ARM.
- Some versions of ARM objdump refuse to disassemble instructions
generated by literal constants (".word 0x..."). The workaround is to
strip the object file first.
- The faulting instruction is marked (liked so) in ARM, but <like so>
in x86.
- ARM mnemonics may include characters such as [] which need to be
escaped before being passed to sed for the "<- trapping instruction"
substitution.
Signed-off-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Michal Marek <mmarek@suse.cz>
This adds CROSS_COMPILE as a kconfig string so you can store it in
.config. Then you can use plain "make" in the configured kernel build
directory to do the right cross compilation without setting the
command-line or environment variable every time.
With this, you can set up different build directories for different kernel
configurations, whether native or cross-builds, and then use the simple:
make -C /build/dir M=module-source-dir
idiom to build modules for any given target kernel, indicating which one
by nothing but the build directory chosen.
I tried a version that defaults the string with env="CROSS_COMPILE" so
that in a "make oldconfig" with CROSS_COMPILE in the environment you can
just hit return to store the way you're building it. But the kconfig
prompt for strings doesn't give you any way to say you want an empty
string instead of the default, so I punted that.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Anibal Monsalve Salazar <anibal@debian.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Michal Marek <mmarek@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Michal Marek <mmarek@suse.cz>
This patch fixes the link error "built-in.o: no such file or directory".
The problem happens if "dirx/Makefile" contains only "obj-m += diry/
dirz/" and the empty "dirx/built-in.o" is missing. Adding $(subdir-m)
into check for builtin-target fixes this error.
Signed-off-by: Jiafu He <jay@goldhive.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Either the functions referred to in a driver struct should live in
.devinit or the driver should be registered using platform_driver_probe
(or equivalent for different driver types) with ->probe being NULL.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
asic3 also needs tmio_core or otherwise will fail to build.
Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
After memory pressure has forced it to dip into the reserves, 2.6.32's
5f8dcc2121 "page-allocator: split per-cpu
list into one-list-per-migrate-type" has been returning MIGRATE_RESERVE
pages to the MIGRATE_MOVABLE free_list: in some sense depleting reserves.
Fix that in the most straightforward way (which, considering the overheads
of alternative approaches, is Mel's preference): the right migratetype is
already in page_private(page), but free_pcppages_bulk() wasn't using it.
How did this bug show up? As a 20% slowdown in my tmpfs loop kbuild
swapping tests, on PowerMac G5 with SLUB allocator. Bisecting to that
commit was easy, but explaining the magnitude of the slowdown not easy.
The same effect appears, but much less markedly, with SLAB, and even
less markedly on other machines (the PowerMac divides into fewer zones
than x86, I think that may be a factor). We guess that lumpy reclaim
of short-lived high-order pages is implicated in some way, and probably
this bug has been tickling a poor decision somewhere in page reclaim.
But instrumentation hasn't told me much, I've run out of time and
imagination to determine exactly what's going on, and shouldn't hold up
the fix any longer: it's valid, and might even fix other misbehaviours.
Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
Btrfs: check total number of devices when removing missing
Btrfs: check return value of open_bdev_exclusive properly
Btrfs: do not mark the chunk as readonly if in degraded mode
Btrfs: run orphan cleanup on default fs root
Btrfs: fix a memory leak in btrfs_init_acl
Btrfs: Use correct values when updating inode i_size on fallocate
Btrfs: remove tree_search() in extent_map.c
Btrfs: Add mount -o compress-force
Here are the sparc bits to remove TIF_ABI_PENDING now that
set_personality() is called at the appropriate place in exec.
Signed-off-by: David S. Miller <davem@davemloft.net>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that the previous commit made it possible to do the personality
setting at the point of no return, we do just that for ELF binaries.
And suddenly all the reasons for that insane TIF_ABI_PENDING bit go
away, and we can just make SET_PERSONALITY() just do the obvious thing
for a 32-bit compat process.
Everything becomes much more straightforward this way.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
'flush_old_exec()' is the point of no return when doing an execve(), and
it is pretty badly misnamed. It doesn't just flush the old executable
environment, it also starts up the new one.
Which is very inconvenient for things like setting up the new
personality, because we want the new personality to affect the starting
of the new environment, but at the same time we do _not_ want the new
personality to take effect if flushing the old one fails.
As a result, the x86-64 '32-bit' personality is actually done using this
insane "I'm going to change the ABI, but I haven't done it yet" bit
(TIF_ABI_PENDING), with SET_PERSONALITY() not actually setting the
personality, but just the "pending" bit, so that "flush_thread()" can do
the actual personality magic.
This patch in no way changes any of that insanity, but it does split the
'flush_old_exec()' function up into a preparatory part that can fail
(still called flush_old_exec()), and a new part that will actually set
up the new exec environment (setup_new_exec()). All callers are changed
to trivially comply with the new world order.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch documents a new ABS_MT parameter and adds further text to
clarify some points around the MT protocol.
Requested-by: Yoonyoung Shim <jy0922.shim@samsung.com>
Requested-by: Mika Kuoppala <mika.kuoppala@nokia.com>
Requested-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Henrik Rydberg <rydberg@euromail.se>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
For pressure-based multi-touch devices, a direct way to send sensor
intensity data per finger is needed. This patch adds the ABS_MT_PRESSURE
event to the MT protocol.
Requested-by: Yoonyoung Shim <jy0922.shim@samsung.com>
Requested-by: Mika Kuoppala <mika.kuoppala@nokia.com>
Requested-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Henrik Rydberg <rydberg@euromail.se>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
I missed converting one dev_info call to deb_dbg before submitting the driver.
Without this change, a message will be printed to dmesg for each button press
if a RC6 remote is used.
Signed-off-by: David Härdeman <david@hardeman.nu>
Cc: stable <stable@kernel.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
Fix failure exit in ipathfs
fix oops in fs/9p late mount failure
fix leak in romfs_fill_super()
get rid of pointless checks after simple_pin_fs()
Fix failure exits in bfs_fill_super()
fix affs parse_options()
Fix remount races with symlink handling in affs
Fix a leak in affs_fill_super()
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6:
x86/PCI: remove IOH range fetching
PCI: fix nested spinlock hang in aer_inject
* master.kernel.org:/home/rmk/linux-2.6-arm:
[ARM] Update mach-types
[ARM] orion5x: D-link DNS-323 rev. B1 power-off
[ARM] Orion5x: add GPIO LED and buttons for wrt350n v2
[ARM] pxa: fix irq suspend/resume for pxa25x
[ARM] pxa: fix the incorrect naming of AC97 reset pin config for pxa26x
[ARM] pxa/corgi: fix incorrect default GPIO for UDC Vbus
[ARM] Kirkwood: drive USB VBUS pin on rd88f6192-nas high on boot
[ARM] Orion: fix PCIe inbound window programming when RAM size is not a power of two
If you have a disk failure in RAID1 and then add a new disk to the
array, and then try to remove the missing volume, it will fail. The
reason is the sanity check only looks at the total number of rw devices,
which is just 2 because we have 2 good disks and 1 bad one. Instead
check the total number of devices in the array to make sure we can
actually remove the device. Tested this with a failed disk setup and
with this test we can now run
btrfs-vol -r missing /mount/point
and it works fine.
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Hit this problem while testing RAID1 failure stuff. open_bdev_exclusive
returns ERR_PTR(), not NULL. So change the return value properly. This
is important if you accidently specify a device that doesn't exist when
trying to add a new device to an array, you will panic the box
dereferencing bdev.
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
If a RAID setup has chunks that span multiple disks, and one of those
disks has failed, btrfs_chunk_readonly will return 1 since one of the
disks in that chunk's stripes is dead and therefore not writeable. So
instead if we are in degraded mode, return 0 so we can go ahead and
allocate stuff. Without this patch all of the block groups in a RAID1
setup will end up read-only, which will mean we can't add new disks to
the array since we won't be able to make allocations.
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This patch revert's commit
6c090a11e1
Since it introduces this problem where we can run orphan cleanup on a
volume that can have orphan entries re-added. Instead of my original
fix, Yan Zheng pointed out that we can just revert my original fix and
then run the orphan cleanup in open_ctree after we look up the fs_root.
I have tested this with all the tests that gave me problems and this
patch fixes both problems. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
