Dave Jones [Thu, 26 Feb 2004 15:38:33 +0000 (15:38 +0000)]
[CPUFREQ] scaling_available_frequencies work for remaining x86 drivers.
Export scaling_available_frequencies on the remaining
x86 freq_table-based cpufreq drivers. powernow-k7, acpi and speedstep-centrino
are already queued.
Please note that this cannot be done in the cpufreq core as the cpufreq core
tries very hard to be independent of the freq_table helpers.
Andrew Morton [Thu, 26 Feb 2004 00:25:59 +0000 (16:25 -0800)]
[PATCH] C99 patch for fs/afs/inode.c
From: "Art Haas" <ahaas@airmail.net>
This patch converts the file to use C99 initializers. The file had this
patch applied before, but a subsequent patch reversed the change and
readded the GNU-style initializer. This patch brings back the C99 one.
Andrew Morton [Thu, 26 Feb 2004 00:25:35 +0000 (16:25 -0800)]
[PATCH] don't use floating point in tdfxfb
From: Jakub Bogusz <qboosh@pld-linux.org>
This patch removes using of floating point operations in tdfxfb - they
are really not needed here (these consts are used only in substitutions
and comparisons with integers) are cause unresolved symbols on some
archs - e.g. on alpha:
Andrew Morton [Thu, 26 Feb 2004 00:25:10 +0000 (16:25 -0800)]
[PATCH] radeon config fix
From: Herbert Xu <herbert@gondor.apana.org.au>
RADEON is selecting I2C_ALGOBIT but as kconfig currently does not propagate
selects up the dependency chain (that is according to Roman), this can
leave I2C as m while I2C_ALGOBIT is y.
Andrew Morton [Thu, 26 Feb 2004 00:22:38 +0000 (16:22 -0800)]
[PATCH] dvb: tda1004x DVB frontend update
From: Michael Hunold <hunold@linuxtv.org>
- tda1004x: standardised where the firmware should be.
- tda1004x: need to re-invert inversion for tda10046 in get_fe()
- tda1004x: reset chip before uploading firmware
- tda1004x: split firmware upload off from frontend init. the initial tune
attempt was taking too long. provide explanation of tuner frequency
calculations
- tda1004x: Fixed signal strength reading for tda10046h
Andrew Morton [Thu, 26 Feb 2004 00:19:49 +0000 (16:19 -0800)]
[PATCH] s390: DASD device driver.
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
DASD driver fixes:
- Fix generic_set_online if diag discipline is not availab.e
- Fix reserve on already reserved device.
- Use default-erp for unit check without sence information.
- Revert dasd naming scheme change from dasd<xyz> to dasd_<busid>_. This
breaks too many user space packages.
- Extend dasd naming scheme to four letters dasd<aaaa>-dasd<zzzz>.
- Fix formatting of dasds.
Andrew Morton [Thu, 26 Feb 2004 00:19:37 +0000 (16:19 -0800)]
[PATCH] s390: IUCV network driver.
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
IUCV interface fixes:
- VM workaround: redirect setmask call to cpu 0.
- Disable message interrupts during connection setup.
- Honor incoming connection severed during connection setup.
- Add connect retry when remote severed connection.
- Add symlinks between net device and iucv device.
- More fixes for proper net device allocation/deallocation.
Andrew Morton [Thu, 26 Feb 2004 00:19:23 +0000 (16:19 -0800)]
[PATCH] s390: LCS network driver.
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
LCS network driver fixes:
- Fix ccwgroup behaviour, remove should offline cards still online.
- Better not write to the debug feature before it is registers.
- Don't free card structure on offline but on remove.
- Check for deferred condition code 1 in lcs_irq for halt requests.
- Call free_netdev only if there is a net device.
Andrew Morton [Thu, 26 Feb 2004 00:19:10 +0000 (16:19 -0800)]
[PATCH] s390: CTC network driver.
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
CTC network driver fixes:
- Fixed some debug messages (did not compile with DEBUG defined).
- Added sanity checks in ctc_proto_store() & ctc_shutdown_device().
- Added some bugfixes from 2.4.
- Tweaked logging.
- Removed syntax error in pr_debug call.
- Changed do_IO (not existing any more) to ccw_device in messages.
- Corrected format failure in pr_debug of channel_get.
- Add symlinks between net device and ctc device.
- Add tiocmset/tiocmget API change patch from Russell King
- Fix locking problem in ctc_tty_cleanup.
Andrew Morton [Thu, 26 Feb 2004 00:18:59 +0000 (16:18 -0800)]
[PATCH] s390: compat_timer_settime.
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
sys_timer_settime returns EINVAL if *new is NULL, compat_timer_settime
returns EFAULT. Add a check to compat_timer_settime to return EINVAL
if *new is NULL.
Andrew Morton [Thu, 26 Feb 2004 00:18:34 +0000 (16:18 -0800)]
[PATCH] s390: common i/o layer.
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Common i/o layer fixes:
- Add atomic onoff variable to ccw devices and ccw-group devices to
avoid races during online/offline.
- Fix pr_debug calls.
- A lot of path fixes:
+ Set device to disconnected state after no path event.
+ Fix chpid vary on/off for single path devices.
+ Make logical vary on/off consistent with physical vary on/off.
+ Don't update subchannel schib if the device is gone (dnv not set).
+ Add code to recover lost chpids after machine checks.
+ Avoid processing link incidents, resource accessability events and
chpid machine checks for logically offline chpids.
+ Recover disconnected devices after chsc machine checks.
+ Delay de-registering of no path devices to avoid deadlocks.
+ Don't redo ssd for known subchannels - the info is static.
+ Introduce a second, "slow" machine check handler thread for new devices.
The "fast" machine check handler only recovers disconnected devices.
- Deregister subchannel rather than ccw device on not oper events.
- Fix calling sequence of notify function vs. path verification.
- Reset timeout for disconnected devices.
- Fix problem with debug feature and %s arguments.
- Fix __get_subchannel_by_stsch to deal with "zombie" subchannels.
- Avoid "zombie" subchannels if device is not operational during sense id.
- Handle call to the io_subchannel remove function if the ccw device
is not registered yet.
- Add availability attribute for ccw devices: "good", "no device",
"no path", "boxed".
- Export ccw_device_work for qdio as module.
- Retry sense id for tape devices which present intervention required.
- Don't check the activity control to decide if the device driver interrupt
handler needs to be called but use the bits in status control.
- Fix race in ccw_device_stlck.
- Accumulate deferred condition code.
- Fix setting_up_sema locking.
- Call qdio_shutdown instead of qdio_cleanup on failed establish.
- Fix problem when 64 FCP adapters are initialized simultaneously.
- Fix problem with >64 adapter interrupt capable devices.
- Reduce stack usage in qdio.
Andrew Morton [Thu, 26 Feb 2004 00:18:22 +0000 (16:18 -0800)]
[PATCH] s390: general update.
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Core s/390 changes:
- Add -mbackchain to CFLAGS to prepare for gcc 3.4 with comes with a
default setting of -mno-backchain
- Add implementation of strcpy.
- Pad bytes after string end in strncpy.
- Fix __sem_update_count inline assembly for gcc 3.4.
- Export smp_ptlb_all for tlb flushing in module code.
- Fix sched_clock.
- Remove the last KERNEL_VERSION #if in s390 code.
- Add dummy implementation for missing dma_{alloc,free}_coherent.
- Avoid cast of lvalue in idal_buffer_{from,to}_user.
- Remove _exit definition from unistd.h.
Andrew Morton [Thu, 26 Feb 2004 00:18:11 +0000 (16:18 -0800)]
[PATCH] fix module reference counting in zoran driver
From: Christoph Hellwig <hch@lst.de>
Take a reference before calling into the module and release it after we're
done. Also remove the useless (and wrong) refcounting in videocodec -
symbols from this module are used by other modules if we call into those
functions so it can't be unloaded anyway.
We really need to add a debug check to tip all those
try_module_get(THIS_MODULE) callers..
Andrew Morton [Thu, 26 Feb 2004 00:17:36 +0000 (16:17 -0800)]
[PATCH] HFS rewrite
From: Roman Zippel <zippel@linux-m68k.org>
This is a complete rewrite of the HFS driver, it gets rid of a all the
special conversion options, which belong in user space. The driver uses now
a btree support very similiar to HFS+, so that both could be merged at some
point.
Thanks to Ethan Benson <erbenson@alaska.net> for a number of patches to make
the driver more compliant with the spec and Christoph Hellwig <hch@lst.de>
for fixing up the documentation.
Don Fry [Wed, 25 Feb 2004 19:37:05 +0000 (14:37 -0500)]
[PATCH] pcnet32 non-mii link state fix
s patch correctly returns the link state of an adapter which does not
have an MII. Tested by myself and the patch author Thomas Munck Steenholdt
<tmus@tmus.dk>. (This version has the correct file name)
Jakub Bogusz [Wed, 25 Feb 2004 19:36:42 +0000 (14:36 -0500)]
[PATCH] 2.6.3 - fix for undefined mdelay in 3c505
This patch fixes undefined mdelay() in 3c505 driver - at least on alpha
(maybe on other archs <linux/delay.h> is included by some other headers,
but on alpha it isn't) there was warning:
Andi Kleen [Wed, 25 Feb 2004 02:01:35 +0000 (18:01 -0800)]
[PATCH] Support AGP bridge on Nvidia Nforce3 + cleanup
For some unknown reasons Nvidia NForce3 doesn't use the standard Hammer AGP architecture,
but requires set up of some shadow registers. This patch adds that to the K8 AGP driver.
Based on an old 2.4 patch from someone at Nvidia.
Also includes another bug fix for the K8 AGP handler, from Brad House.
We should not assume that there is only one northbridge in a Uniprocessor system.
Always flush all.
Andi Kleen [Wed, 25 Feb 2004 01:58:41 +0000 (17:58 -0800)]
[PATCH] New machine check handler for x86-64
This adds a new completely rewritten machine check handler for x86-64.
The old one never worked on 2.6.
The new handler has many improvements. It closely follows the Intel and AMD
recommendations on MCE handlers now (the old one had many violations). It handles
unrecoverable errors in user space better now - it will only kill the process now
if possible instead of panicing.
This one is CPU independent now - it should work on any CPU that supports the standard
x86 MCA architecture.
This new handler only logs fatal errors that lead to kernel panic to the console.
Non fatal errors are logged race free into a new (non ring) buffer now
and supplied to the user using a new character device. The old one could
deadlock on console and printk locks. This also separates machine check errors
from real kernel errors better. The new buffer has been also designed to
be easily accessible from external debugging tools: it has a signature
and could be even recovered after reboot. It is not organized as a ring buffer -
this means the first errors are kept unless explicitely cleared.
The new error formats can be parsed using ftp://ftp.suse.com/pub/people/ak/x86-64/mcelog.c
The new character device for it can be created with mknod /dev/mcelog c 10 227
There is a new sysfs interface to configure the machine check handler.
It has a "tolerant" parameter that defines the aggressiveness of the machine check:
0: always panic
1: panic if deadlock possible (e.g. MCE happened in the kernel)
2: try to avoid panic
Default is 2
Despite of having more features the new handler is shorter.
Andi Kleen [Wed, 25 Feb 2004 01:56:12 +0000 (17:56 -0800)]
[PATCH] x86-64 merge for 2.6.3
Bring the x86-64 port up to date. Lots of smaller bug fixes that have accumulated.
Also fixes another nasty bug introduced by the IA32e changes that causes BUGs at
boot for some people.
Only changes x86-64 specific files. There are some other changes that I'm sending
separately.
- Some cleanup in NMI watchdog code
- Fix HyperThreading CPU setup race (Suresh B. Siddha)
- Update defconfig
- Add a comment on why iommu_fullflush is disabled.
- Export sys_ioctl again
- Fix build with IA32_EMULATION=y and SYSVIPC=n
- Remove noisy boot printks in the mptable scan.
- Implement automatic NMI watchdog switching for real now
- Remove redundant 32bit ioctl handlers for autofs
- Remove CONFIG ifdefs around rtc 32bit ioctl handlers
- Remove useless nfsctl ifdef in syscall.c (Al Viro)
- Increase padding for prefetchw alternative
- Check for NX bit early before setting up memory maps (Suresh B. Siddha)
- Change Intel IA32e config description and fix help texts (Jun Nakajima)
- Fix microcode driver build really now (Dave Jones)
- Add nohpet option to disable HPET timer
- Fix double semicolon in aperture.c
- Add cmpxchg16b cpuid entry
- Fix return value of read_pci_config_16 (Paul Menage)
- Fix __KERNEL_COMPAT32_CS (Zachary Amsden)
- Disable the infamous 30 minutes check in CMOS time setting
- Update URLs in Kconfig (Petri T. Koistinen)
- Fix ACPI interrupt source parsing for Nforce3 (Maciej W. Rozycki)
- Fix 32bit ipc version parsing.
- Run local APIC NMI watchdog only once a second (or less often on idle boxes)
- Merge ACPI APIC SCI functions from i386
- Add i8254 timer suspend code from i386
- Merge with 2.6.2-rc3 + minor changes from i386
- Fix empty_zero_page declaration (Greg Johnson)
- Readd sysctls for exception/page fault trace and vsyscall32
- Fix WCHAN
- Fix STACK_TOP usage. Stack for 64bit processes should be at the
top of memory now again. Also set it correctly for LINUX32_3GB.
- Add warning fixes for gcc 3.4 and -Wdeclaration-after-statement