Linus Torvalds: Linux 2.6.19-rc6

there’s nothing earth-shattering here (and there shouldn’t be), but
we’ve hopefully made good progress on the regression list (and
thanks again to Adrian Bunk for reminding people, especially when
they thought cough that some particular regression had
already been fixed)..

So with -rc6, we hopefully should leave the irq-related
regressions behind us. There were issues both with devices that
started enabling MSI (which seem to trigger hardware bugs, although
there’s also been discussion about what we should do to make things
safer) and with the new genirq layer that showed problems with
edge-triggered irq’s (notably legacy ISA interrupts, or, more
commonly these days, the 16-bit PCMCIA interrupts that are
basically just ISA in another formfactor).

Thanks for everybody involved in whittling down that regression

Also, apart from the regression tracking, we’ve had some other
updates, eg infiniband and DVB fixes, network driver fixes, some
networking fixes etc.

The ShortLog is appended, and gives a mostly readable picture of
what has been going on. But the main thing to take away is:
regressions fixed, and not a whole lot of changes since -rc5 (it
may not look that way, but a lot of these things are essentially
one-liners or close to it, so the total diff between -rc5 and -rc6
is actually just about 5k lines, which is not a whole lot,

Aaron Durbin (1):
      x86-64: Fix partial page check
to ensure unusable memory is not being marked usable.

Adrian Bunk (2):
      bcm43xx: Add error checking in
      drivers/telephony/ixj: fix an
array overrun

Alan Cox (1):
      hpt37x: Check the

Alan Stern (1):
      SCSI core: always store >=
36 bytes of INQUIRY data

Alasdair G Kergon (2):
      dm: fix find_device race
      dm: suspend: fix error path

Alexey Dobriyan (4):
      ipmi_si_intf.c: fix “&&
0xff” typos
      V4L/DVB (4795): Tda826x: use
correct max frequency
      V4L/DVB (4818): Flexcop-usb:
fix debug printk
      pata_artop: fix “& (1
>>” typo

Andi Kleen (6):
      Revert “MMCONFIG and new Intel
      x86-64: Fix
PTRACE_[SG]ET_THREAD_AREA regression with ia32 emulation.
      x86-64: Handle
reserve_bootmem_generic beyond end_pfn
      x86: Add
acpi_user_timer_override option for Asus boards
      x86-64: Fix vgetcpu when
      x86-64: Fix race in

Andrew Morton (2):
      setup_irq(): better mismatch
      revert “PCI: quirk for IBM Dock
II cardbus controllers”

Arjan van de Ven (1):
      Regression in 2.6.19-rc
microcode driver

Benjamin Herrenschmidt (2):
      [POWERPC] Fix cell “new style”
mapping and add debug
      powerpc: windfarm shall request
it’s sub modules

Brian King (1):
      libata: Convert from
module_init to subsys_initcall

Bryan O’Sullivan (1):
      IB/ipath – program intconfig
register using new HT irq hook

Chris Lalancette (1):
      [NETPOLL]: Compute checksum
properly in netpoll_send_udp().

Corey Minyard (3):
      IPMI: Clean up the waiting
message queue properly on unload
      IPMI: retry messages on certain
error returns
      IPMI: Fix more &&

Daniel Ritz (1):
      fix via586 irq routing for pirq

Darrick J. Wong (1):
      libata: fix double-completion
on error

David Brownell (1):
      usb: MAINTAINERS updates

David Chinner (3):
      [XFS] Clean up i_flags and
i_flags_lock handling.
      [XFS] Prevent a deadlock when
xfslogd unpins inodes.
macros from xfs_dmapi.h

David Gibson (1):
      hugetlb: check for brk()
entering a hugepage region

David Miller (1):
      pci: don’t try to remove sysfs
files before they are setup.

David Rientjes (1):
      drivers cris: return on NULL

Eric Dumazet (1):
      vmalloc: optimization, cleanup,

Eric W. Biederman (4):
      sysctl: Undeprecate
      htirq: refactor so we only have
one function that writes to the chip
      htirq: allow buggy drivers of
buggy hardware to write the registers
      Use delayed disable mode of
ioapic edge triggered interrupts

Franck Bui-Huu (1):
      .gitignore: add miscellaneous

Geoff Levand (1):
      [POWERPC] cell: set

Herbert Xu (1):
      [NET]: Set truesize in

Hermann Pitton (1):
      V4L/DVB (4802): Cx88: fix
remote control on WinFast 2000XP Expert

Hoang-Nam Nguyen (3):
      IB/ehca: Assure 4K alignment
for firmware control blocks
      IB/ehca: Use named constant for
max mtu
      IB/ehca: Activate scaling code
by default

Hugh Dickins (2):
      hugetlb: prepare_hugepage_range
check offset too
      hugetlb: fix error return for
brk() entering a hugepage region

Ian Kent (1):
      autofs4: panic after mount

J. Bruce Fields (3):
      nfsd4: reindent
      nfsd4: fix open-create
      nfsd: fix spurious error return
from nfsd_create in async case

Jean Delvare (2):
      V4L/DVB (4817): Fix uses of
“&&” where “&” was intended
      RDMA/amso1100: Fix &&

Jeff Garzik (1):
      [libata] sata_via: fix obvious

Jens Axboe (4):
      Fix bad data direction in
      ide-cd: only set rq->errors
SCSI style for block pc requests
      cciss: fix iostat
      cpqarray: fix iostat

Jes Sorensen (1):
      mspec driver build fix

Jiri Slaby (2):
      [NET]: kconfig, correct traffic
      Char: isicom, fix close bug

John Heffner (1):
      [TCP]: Don’t use highmem in tcp
hash size calculation.

John Rose (1):
      [POWERPC] pseries: Force 4k
update_flash block and list sizes

Jonathan E Brassow (2):
      dm: multipath: fix rr_add_path
      dm: raid1: fix waiting for io
on suspend

Julian Anastasov (1):
      [IPVS]: More endianness

Kalle Pokki (2):
non-console transmit
non-console initialisation

KAMEZAWA Hiroyuki (1):
      ia64: select ACPI_NUMA if

Linus Torvalds (6):
      Revert “i386: Add MMCFG
resources to i386 too”
      x86-64: clean up io-apic
      x86-64: write IO APIC irq
routing entries in correct order
      [dvb saa7134] Fix missing
‘break’ for avermedia card case
      Revert “fix Data Acess error in
      Linux 2.6.19-rc6

Magnus Damm (1):
      x86-64: setup saved_max_pfn
correctly (kdump)

Masami Hiramatsu (1):
      kretprobe: fix
kretprobe-booster to save regs and set status

Mauro Carvalho Chehab (1):
      V4L/DVB (4804): Fix missing i2c
dependency for saa7110

Michael Buesch (1):
      bcm43xx: Drain TX status before
starting IRQs

Michael Chan (1):
      [TG3]: Fix array overrun in

Nathan Lynch (1):
      nvidiafb: fix unreachable code
in nv10GetConfig

NeilBrown (2):
      md: change ONLINE/OFFLINE
events to a single CHANGE event
      md: fix sizing problem with
raid5-reshape and CONFIG_LBD=n

Nicolas Kaiser (1):
      drivers/ide: stray bracket

Oleg Nesterov (1):
      A minor fix for set_mb() in

pasky@ucw.cz (3):
      V4L/DVB (4814): Remote support
for Avermedia 777
      V4L/DVB (4815): Remote support
for Avermedia A16AR
      V4L/DVB (4816): Change tuner
type for Avermedia A16AR

Paul Mackerras (1):
      [POWERPC] Make sure initrd and
dtb sections get into zImage correctly

Pavel Emelianov (1):
      Fix misrouted interrupts

Peter Zijlstra (1):
      bonding: lockdep annotation

Rafael J. Wysocki (1):
      md: do not freeze md threads
for suspend

Randy Dunlap (1):
      com20020 build fix

Roland Dreier (1):
      IB/mad: Fix race between cancel
and receive completion

Russell King (1):
      Fix missing parens in

Sharyathi Nagesh (1):
      fix Data Acess error in

Simon Horman (1):
      [IPVS]: Compile fix for
annotations in userland.

Stephen Hemminger (1):
      [PKT_SCHED] sch_htb: Use

Stephen Rothwell (2):
      [POWERPC] Add the
thread_siblings files to sysfs
      [POWERPC] Wire up

Steve French (4):
      [CIFS] NFS stress test
generates flood of “close with pending write” messages
      [CIFS] Explicitly set
      [CIFS] Fix mount failure when
domain not specified
      [CIFS] Fix minor problem with
previous patch

Steven Rostedt (1):
      x86-64: shorten the x86_64 boot
setup GDT to what the comment says

Steven Whitehouse (1):
      [DECNET]: Endianess fixes (try

Takashi Iwai (1):
      ALSA: hda-intel – Disable MSI
support by default

Tigran Aivazian (1):
      Tigran has moved

Tim Shimmin (1):
      [XFS] Keep lockdep happy.

Timo Teras (2):
      MMC: Poll card status after
rescanning cards
      MMC: Do not set unsupported
bits in OCR response

Tom Tucker (1):
      RDMA/amso1100: Fix unitialized
pseudo_netdev accessed in c2_register_device

Vivek Goyal (1):
      i386: Force data segment to be
4K aligned

Vlad Apostolov (3):
      [XFS] 956618: Linux crashes on
boot with XFS-DMAPI filesystem when
      [XFS] rename uio_read() to
      [XFS] 956664: dm_read_invis()
changes i_atime

Wink Saville (1):
      Patch for nvidia divide by zero
error for 7600 pci-express card

Get the Free Newsletter!

Subscribe to Developer Insider for top news, trends, & analysis