In next-20080530 and next-20080602 (and possibly earlier - I can't remember the linux-next tree before that I tried) I get the following: [ 12.885153] ------------[ cut here ]------------ [ 12.885203] WARNING: at kernel/lockdep.c:2680 check_flags+0x98/0x151() [ 12.885248] Pid: 4, comm: watchdog/0 Not tainted 2.6.26-rc4-next-20080602 #13 [ 12.885292] [ 12.885293] Call Trace: [ 12.885364] [<ffffffff8022bbd5>] warn_on_slowpath+0x58/0x8a [ 12.885410] [<ffffffff804c9cfe&gt...

Hello, I evaluated overhead of kernel marker using linux-2.6-sched-fixes git tree, which includes several markers for LTTng, using an ia64 server. While the immediate trace mark feature isn't implemented on ia64, there is no major performance regression. So, I think that we don't have any issues to propose merging marker point patches into Linus's tree from the viewpoint of performance impact. I prepared two kernels to evaluate. The first one was compiled without CONFIG_MARKERS. The second ...

Maybe we could settle for an intermediate solution : I agree with you that defining the trace points in headers, like you did for the scheduler, makes the code much cleaner and makes things much easier to maintain afterward. However, having the trace_mark mechanism underneath helps a lot in plugging a generic tracer (actually, if we can settle the marker issue, I've got a kernel tracer, LTTng, that I've been waiting for quite a while to push to mainline that I would like to post someday). So I wo...

This is an updated/split up version of the patch I sent earlier. Currently sched domains created by the cpusets are completely destroyed during CPU hotplug event handling. For each CPU hotplug event scheduler attaches all CPUs to the NULL domain and then puts them all into a single domain thereby destroying domains created by the cpusets. The solution is simple, when cpusets are enabled scheduler should not create the default domain and instead let cpusets rebuild the domains based on the current...

ide_find_port() now depends on ->chipset being set for occupied ide_hwifs[] slots so all host drivers have to initialize hwif->chipset properly. This patch fixes a regression on hosts with > 1 port or with a single port but no devices attached to it for an affected host drivers. Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> --- drivers/ide/arm/bast-ide.c | 1 + drivers/ide/arm/ide_arm.c | 1 + drivers/ide/ide-pnp.c | 1 + drivers/ide/id...

Convert the driver to use struct ide_port_info - as a nice side-effect this fixes hwif->channel initialization. Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> --- drivers/ide/legacy/ide-4drives.c | 27 ++++++++++++++++++--------- 1 file changed, 18 insertions(+), 9 deletions(-) Index: b/drivers/ide/legacy/ide-4drives.c =================================================================== --- a/drivers/ide/legacy/ide-4drives.c +++ b/drivers/ide/legacy/ide-4drives.c @@ -...

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> --- drivers/ide/pci/delkin_cb.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) Index: b/drivers/ide/pci/delkin_cb.c =================================================================== --- a/drivers/ide/pci/delkin_cb.c +++ b/drivers/ide/pci/delkin_cb.c @@ -134,14 +134,12 @@ static struct pci_driver driver = { .remove = delkin_cb_remove, }; -static int -delkin_cb_init (void) +static int __init delkin_cb_...

Convert the driver to use struct ide_port_info - as a nice side-effect this fixes racy setup of ->io_32bit/unmask settings (after ide_device_add() call device can be already in use). Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> --- drivers/ide/pci/delkin_cb.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) Index: b/drivers/ide/pci/delkin_cb.c =================================================================== --- a/drivers/ide/pci/delkin_cb.c ...

* Add 'parent' field to hw_regs_t for optional parent device pointer (needed by macio PMAC IDE controllers) and set hwif->dev in ide_init_port_hw(). * Update au1xxx-ide.c, sgiioc4.c, pmac.c and setup-pci.c accordingly. There should be no functional changes caused by this patch. Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> --- drivers/ide/ide.c | 3 ++- drivers/ide/mips/au1xxx-ide.c | 2 -- drivers/ide/pci/sgiioc4.c | 2 -- drivers/ide/ppc/...

From: J. Bruce Fields <bfields@citi.umich.edu> Every now and then as I'm looking for something in Documentation/ I get annoyed at how cluttered that directory is, and go look for a few files that could be grouped into a subdirectory. This seemed like one candidate, though a lot of referenes to DMA-mapping.txt require fixing. Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu> --- A Documentation/arch directory would take out another dozen entries or so, but I haven't had the e...

1. kernel does not boot in SMP mode if I press any button on a keyboard before kernel loads (i.e. grub menu) 2. This is Dell Vostro 400 Core 2 Quad Q6600 with the latest bios 1.0.13. Ubuntu kernel 2.6.24-16-generic boots with ipqpoll. Gentoo 2.6.24-gentoo-r8 and vanilla 2.6.25.4 do not boot in SMP mode if I press any button on a keyboard before kernel loads (i.e. grub menu). I tried noacpi, noapic, acpi=off, irqpoll still does not boot. It boots if I add nosmp but the computer is very slow. 3. ...

Fairly simple. "dev_use" was being allocated as a zero length array because of bad math on 64-bit systems causing a crash in find_first_zero_bit(). One-liner follows: diff --git a/drivers/mmc/card/block.c b/drivers/mmc/card/block.c index 91ded3e..f9ad960 100644 --- a/drivers/mmc/card/block.c +++ b/drivers/mmc/card/block.c @@ -46,7 +46,7 @@ #define MMC_SHIFT 3 #define MMC_NUM_MINORS (256 >> MMC_SHIFT) -static unsigned long dev_use[MMC_NUM_MINORS/(8*sizeof(unsigned long))]; +static DEC...

Hi all, No comments on the previous version for two weeks... resending once again. Changes since v3: - None. Simply resending, plus adding Andrew Morton and linux-kernel to Cc, since I'm constantly getting bounces from Wim Van Sebroeck's email: - - - - <wim@iguana.be>: 213.249.96.99 does not like recipient. Remote host said: 450 Client host rejected: cannot find your hostname, [85.21.88.2] Giving up on 213.249.96.99. I'm not going to try again; this message has been...

mpc83xx_wdt is the OF driver now, so we don't need fsl_soc constructor. Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> --- arch/powerpc/sysdev/fsl_soc.c | 46 ----------------------------------------- 1 files changed, 0 insertions(+), 46 deletions(-) diff --git a/arch/powerpc/sysdev/fsl_soc.c b/arch/powerpc/sysdev/fsl_soc.c index a5ceeef..32a3ac8 100644 --- a/arch/powerpc/sysdev/fsl_soc.c +++ b/arch/powerpc/sysdev/fsl_soc.c @@ -545,52 +545,6 @@ err: arch_initcall(fsl_i2c_o...

mpc83xx_wdt.c renamed to mpc8xxx_wdt.c, now we can do various renames in the file itself. Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> --- drivers/watchdog/mpc8xxx_wdt.c | 104 ++++++++++++++++++++-------------------- 1 files changed, 52 insertions(+), 52 deletions(-) diff --git a/drivers/watchdog/mpc8xxx_wdt.c b/drivers/watchdog/mpc8xxx_wdt.c index 19e3082..2f0681f 100644 --- a/drivers/watchdog/mpc8xxx_wdt.c +++ b/drivers/watchdog/mpc8xxx_wdt.c @@ -1,5 +1,5 @@ /* - * m...

On MPC86xx the watchdog could be enabled only at power-on-reset, and could not be disabled afterwards. We must ping the watchdog from the kernel until the userspace handles it. MPC83xx CPUs are only differ in a way that watchdog could be disabled once, but after it was enabled via software it becomes just the same as MPC86xx. Thus, to support MPC86xx I added the kernel timer which pings the watchdog until the userspace opens it. Since we implemented the timer, now we're able to implement prope...

Quite tired of these warnings ;-), checkpatch spitting them when seeing the rename patch. WARNING: Use #include <linux/io.h> instead of <asm/io.h> #25: FILE: watchdog/mpc83xx_wdt.c:25: +#include <asm/io.h> WARNING: Use #include <linux/uaccess.h> instead of <asm/uaccess.h> #26: FILE: watchdog/mpc83xx_wdt.c:26: +#include <asm/uaccess.h> WARNING: line over 80 characters #45: FILE: watchdog/mpc83xx_wdt.c:45: +MODULE_PARM_DESC(timeout, "Watchdog timeout in tick...

Oh, I see it now, thanks. I've just read Wim's comments there, and I think I'll have to wait for git-watchdog update and then will rebase my work on top. Thanks, -- Anton Vorontsov email: cbouatmailru@gmail.com irc://irc.freenode.net/bd2 --

Hi, as far as I understand, the current GPIO API only presents very basic GPIO functionality: direction and level reading and writing. Whereas many GPIO controllers have many further configurable parameters: pull-ups and pull-downs, drive strength, slew rate, etc. And it is desirable to be able to access those features too. Of course, we cannot extent the API with all these possible functions. Would a generic GPIO parameter handling API be desirable? Like struct gpio_parameter { char *na...

I like the idea in general. The biggest worry I have is trying to find the parameter for you to fiddle with. The driver which is going to want to set the parameters is going to have the gpio number, not the gpio_chip. Also, the fact that the parameters are uniquely identified by strings is a bit awkward. I can see people registering the same kind of parameter for different chips like "pullup", "Pullup", "pu" etc making the driver's task even harder. So, I reckon if we're to do this we shoul...

Signed-off-by: Jan Kara <jack@suse.cz> --- include/linux/quota.h | 53 +++++++++++++++++++++++++++++++++++------------- 1 files changed, 38 insertions(+), 15 deletions(-) diff --git a/include/linux/quota.h b/include/linux/quota.h index 841c4ae..04ec8c5 100644 --- a/include/linux/quota.h +++ b/include/linux/quota.h @@ -125,10 +125,20 @@ extern spinlock_t dq_data_lock; #define DQUOT_DEL_REWRITE max(V1_DEL_REWRITE, V2_DEL_REWRITE) void mark_info_dirty(struct super_block *sb, int type)...

Cleanup quotaops.h: Rename functions from uppercase to lowercase (and define backward compatibility macros), move larger functions to dquot.c and make them non-inline. Signed-off-by: Jan Kara <jack@suse.cz> --- fs/dquot.c | 53 +++++++++++ include/linux/quotaops.h | 226 ++++++++++++++++++++++------------------------ 2 files changed, 160 insertions(+), 119 deletions(-) diff --git a/fs/dquot.c b/fs/dquot.c index 5ac77da..2664602 100644 --- a/fs/dquot.c +++ b/fs/dquot.c ...

Hi, This is really too hideous in my opinion and looks like a candidate for its own static inline function. Or you can try to rewrite the boolean expression on multiple lines using continue, something like: - for (cnt = 0, dirty = 0; cnt < MAXQUOTAS; cnt++) - if ((type == cnt || type == -1) && sb_has_quota_enabled( - && info_any_dirty(&sb_dqopt(sb)->info[cnt])) - dirty = 1; ...

We need definitions from quota.h and so far we depended on fs.h to include this file for us. This will be no longer true. Signed-off-by: Jan Kara <jack@suse.cz> --- fs/ext2/super.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/fs/ext2/super.c b/fs/ext2/super.c index ef50cbc..609df85 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -31,6 +31,7 @@ #include <linux/seq_file.h> #include <linux/mount.h> #include <linux/log2.h> +#include ...

Em Mon, 02 Jun 2008 10:04:32 -0700 Casey Schaufler <casey@schaufler-ca.com> escreveu: | From: Casey Schaufler <casey@schaufler-ca.com> | | The d_instantiate hook for Smack can hang on the root | inode of a filesystem if the file system code has not | really done all the set-up. Fuse is known to encounter | this problem. This change detects an attempt to | instantiate a root inode and addresses it early in the | processing, before any attempt is made to do something | that might hang...