- 30 Dec, 2003 15 commits
-
-
bk://linuxusb.bkbits.net/i2c-devel-2.6Linus Torvalds authored
into home.osdl.org:/home/torvalds/v2.5/linux
-
bk://linux-scsi.bkbits.net/scsi-for-linus-2.6Linus Torvalds authored
into home.osdl.org:/home/torvalds/v2.5/linux
-
James Bottomley authored
-
James Bottomley authored
This should ensure it doesn't ordinarily break the builds, but will error out if the builder requests it to.
-
James Bottomley authored
From: "Justin T. Gibbs" <gibbs@scsiguy.com>
-
James Bottomley authored
From: Willem Riede <wrlk@riede.org> Brings 2.6.x version of osst up to par with the 2.4.y version. Except for the /proc changes Tested against released 2.6.0 kernel. Changes from what's in the kernel tree today: - Fixes bug that files shorter than one 32K frame don't get written. - Fix a memory alloc/free mismatch that could have made your kernel unstable after rmmod osst. - Fix a number of tape (re)positioning bugs around filemarks that affected Amanda, Arkeia and Storix backup software. - Rationalize module parameters. - Fix time-out skipping to EOD - Write FM+EOD+Header-update when file write terminated by ioctl - Follow standard Unix behavior for read at EOD (return zero bytes read twice then error) - Implement SETBLK ioctl (allowed before first write only)
-
Kai Mäkisara authored
This patch adds support for cdevs to the st driver. The changes are based on Doug Gilbert's corresponding changes to the sg driver. Using cdevs brings the following advantanges: - support for many drives, currently the maximum is set to 128; the minor number assignment is explained in the patch to Documentation/scsi/README.st - the tape devices appear in /sys/cdev/major/st*, as an example here are the entries for the first drive (4 modes, rewind and non-rewind devices): /sys/cdev/major/st0m0 /sys/cdev/major/st0m1 /sys/cdev/major/st0m2 /sys/cdev/major/st0m3 /sys/cdev/major/st0m0n /sys/cdev/major/st0m1n /sys/cdev/major/st0m2n /sys/cdev/major/st0m3n
-
James Bottomley authored
into raven.il.steeleye.com:/home/jejb/BK/scsi-for-linus-2.6
-
Greg Kroah-Hartman authored
into kroah.com:/home/greg/linux/BK/i2c-new_drivers-2.6
-
Jean Delvare authored
Here is a patch for the lm83 driver, to be applied on top of your pending patches stack. What it does: * Remove limit initialisation by the driver. This is a backport from CVS. * A few whitespace changes inspired by my recent porting of the lm90 driver.
-
Jean Delvare authored
This patch replaces "K9000" with "K8000" everywhere (5 occurences) since it seems that the "K9000" was a typo in the first place. It also rewords the i2c-velleman doc. I have fixed it in our CVS repository too, and have been sending a similar patch to Marcelo for Linux 2.4.
-
Jean Delvare authored
The following patch fixes incorrect dependencies (as far as I can see) for the SCx200 modules. A similar patch (with even more fixes) is also needed for Linux 2.4, and will be part of my next wave to Marcelo. Note that I don't have the necessary hardware to actually test anything, but a quick look at the code doesn't leave much place for doubt IMHO.
-
Jean Delvare authored
This patch drops bus scan from i2c-algo-ite and i2c-ibm_iic. It also removes the incomplete and broken SLO_IO stuff from i2c-algo-ite.
-
Greg Kroah-Hartman authored
into kroah.com:/home/greg/linux/BK/i2c-new_drivers-2.6
-
Matthew Wilcox authored
Highlights: - Switch to generic ioctl32 handling - Use the new *_defconfig mechanism - Use drivers/Kconfig - Big signal cleanups and support for restartable syscalls
-
- 29 Dec, 2003 25 commits
-
-
David S. Miller authored
-
David S. Miller authored
-
David S. Miller authored
into nuts.ninka.net:/disk1/davem/BK/sparc-2.6
-
bk://gkernel.bkbits.net/net-drivers-2.5Linus Torvalds authored
into home.osdl.org:/home/torvalds/v2.5/linux
-
Jeff Garzik authored
-
Andrew Morton authored
-
Daniele Venzano authored
The attached patch adds support for suspend/resume to the sis900 driver. With this patch on resume the NIC is fully configured and operational, before a module reload was needed because of the complete lack of suspend/resume callbacks. I added two functions, sis900_suspend and sis900_resume, with their pointers in struct pci_driver. A vector of 16 u32 was then needed to the to keep PCI data during suspend. I added it in struct sis900_private. I updated the revision number to reflect my changes. Looking at the code I also killed three typos. The patch doesn't touch any other code. Since I don't know anything on ethernet drivers the rule 'works for me' is fully valid.
-
Linus Torvalds authored
-
bk://bk.arm.linux.org.uk/linux-2.6-serialLinus Torvalds authored
into home.osdl.org:/home/torvalds/v2.5/linux
-
bk://kernel.bkbits.net/davem/net-2.6Linus Torvalds authored
into home.osdl.org:/home/torvalds/v2.5/linux
-
Andrew Morton authored
From: Joe Thornber <thornber@sistina.com> You can no longer call dm_table_event() from interrupt context.
-
Andrew Morton authored
From: Joe Thornber <thornber@sistina.com> Make sure that a target has a sensible set of default io restrictions.
-
Andrew Morton authored
From: Joe Thornber <thornber@sistina.com> Make the version-4 ioctl interface the default kernel configuration option. If you have out of date tools you will need to use the v1 interface.
-
Andrew Morton authored
From: Joe Thornber <thornber@sistina.com> The dm table size is always known in advance, so we can specify it in dm_table_create(), rather than relying on dynamic resizing.
-
Andrew Morton authored
From: Joe Thornber <thornber@sistina.com> When setting the size of a Device-Mapper device in the gendisk entry, also try to set the size of the corresponding block_device entry's inode. This is necessary to allow online device/filesystem resizing to work correctly. [Kevin Corry]
-
Andrew Morton authored
From: viro@parcelfarce.linux.theplanet.co.uk When we register disks, we mangle the disk names that contain slashes (e.g. cciss/c0d0) replacing them with '!' in corresponding sysfs names. So name_to_dev_t() should mangle the name in the same way before looking for it in /sys/block.
-
Andrew Morton authored
From: Rusty Russell <rusty@rustcorp.com.au> Entirely revised, and largely rewritten. Has a continuing example now, which I think makes things clearer. Also covers Read Copy Update. This version further deprecates rwlock_t, shuffles sections for better organization.
-
Andrew Morton authored
From: Alan Cox <alan@redhat.com> VIA audio had a fix from 2.4 missing so any user could spam the system log. Also include a fix for a bug which is pending 2.4 fixing too and causes a bogus warning to be displayed on close of audio file handle.
-
Andrew Morton authored
From: Alan Cox <alan@redhat.com> Type errors, just fixes a warning
-
Andrew Morton authored
From: Alan Cox <alan@redhat.com> Just a warning fix and behaviour tidy. Changing the kiss.mintime variable isn't going to work as its exposed to user space
-
Andrew Morton authored
From: jbarnes@sgi.com (Jesse Barnes) Just a quick patch to fix MAINTAINERS for sn2.
-
Andrew Morton authored
From: Pádraig Brady <P@draigBrady.com> Watchdog driver for the Winbond w83627hf which is on the last 3 motherboards I got here for test (tyan, advantech, force).
-
Andrew Morton authored
From: Johannes Stezenbach <js@convergence.de> the patch below removes warnings like: warning: signed and unsigned type in conditional expression when compiling userspace applications against a glibc built with 2.6 kernel headers (like on Debian unstable).
-
Andrew Morton authored
From: Jeff Garzik <jgarzik@pobox.com> I've seen this patch floating around. Not sure the origin, but it's surfaced on lkml and also when I was poking around handhelds.org CVS for iPAQ patches: on non-PCs, particularly system-on-chip devices but not just there, you have a custom "platform bus" that is the root of pretty much all other devices and buses. It's something I wanted to make sure people didn't forget; to make sure the legacy_bus didn't get "legacied out of existence." ;-)
-
Andrew Morton authored
From: Roland McGrath <roland@redhat.com> I believe I have identified a failure mode that Linus saw a couple weeks back when tracking down some other fork/exit sorts of races. We saw this come up on rare occasions with the RHEL3 kernel's backport of the new code (while trying to track down other race failure modes we have yet to fix, sigh). I am talking about the following scenario: > Btw, even with the fix, doing a "while : ; ./crash t 10 ; done" will > eventually result in a stuck process: > > 1415 tty1 D 0:00 ./crash > > This is some kind of deadlock: most of the fifty threads are in "D" > state, with a trace something like > > [<c011fbe3>] schedule+0x360/0x7f8 > [<c0120539>] wait_for_completion+0xd4/0x1c3 > [<c0128c9e>] do_exit+0x627/0x6a4 > [<c0128ddd>] do_group_exit+0x3d/0x177 > [<c0130c13>] dequeue_signal+0x2d/0x84 > [<c0133911>] get_signal_to_deliver+0x390/0x575 > [<c010a541>] do_signal+0x6c/0xf1 > [<c01200be>] default_wake_function+0x0/0x12 > [<c01200be>] default_wake_function+0x0/0x12 > [<c013d50f>] do_futex+0x6d/0x7d > [<c013d635>] sys_futex+0x116/0x12f > [<c010a601>] do_notify_resume+0x3b/0x3d > [<c010a82e>] work_notifysig+0x13/0x15 > > except for one that is trying to core-dump: > > [<c0120539>] wait_for_completion+0xd4/0x1c3 > [<c01200be>] default_wake_function+0x0/0x12 > [<c01200be>] default_wake_function+0x0/0x12 > [<c02101aa>] rwsem_wake+0x86/0x12d > [<c01738af>] coredump_wait+0xa8/0xaa > [<c0173a26>] do_coredump+0x175/0x26c > > and three that are just doing a regular "exit()" system call: > > [<c011fbe3>] schedule+0x360/0x7f8 > [<c011e19a>] recalc_task_prio+0x90/0x1aa > [<c0120539>] wait_for_completion+0xd4/0x1c3 > [<c01200be>] default_wake_function+0x0/0x12 > [<c01200be>] default_wake_function+0x0/0x12 > [<c0210207>] rwsem_wake+0xe3/0x12d > [<c0128c9e>] do_exit+0x627/0x6a4 > [<c0128d4d>] next_thread+0x0/0x53 > [<c010a7e3>] syscall_call+0x7/0xb > > However, the rest of the system is totally unaffected by this deadlock: > it's only deadlocked withing the thread group itself, nobody else cares. What happens here is a race between an exiting thread checking mm->core_waiters in __exit_mm, and the thread taking the core-dump signal (in coredump_wait) examining the first thread's ->mm pointer and incrementing mm->core_waiters to account for it. There is no synchronization at all in __exit_mm's use of mm->core_waiters. If the coredump_wait thread reads tsk->mm when tsk is in __exit_mm between checking mm->core_waiters and clearing tsk->mm, then it will increment mm->core_waiters and the total count will later exceed the number of threads that will ever decrement it and synchronize. Hence it blocks forever. The following patch fixes the problem by using mm->mmap_sem in __exit_mm. The read lock must be held around checking mm->core_waiters and clearing tsk->mm so that coredump_wait (which gets the write lock) cannot come in between and do bogus bookkeeping.
-