Ok, let's calm down for a while before the final 2.6.7.
-rc3 does a lot of sparse type cleanup, mainly thanks to Al Viro (but his
work ended up getting some other people involved too, since the list of
sparse warnings isn't as daunting any more). Some of that has unearthed
real bugs which Al fixed.
But there are DRM, AGP, cpufreq, sparc64, and input updates there too. See
the appended shortlog for more details,
Linus
-----
Summary of changes from v2.6.7-rc2 to v2.6.7-rc3
============================================
Adam Kropelin:
o [FB]: Fix uninitialized fb_cmap member in sbuslib.c
Adrian Bunk:
o POSIX_MQUEUE depends on NET
o SECURITY_SELINUX depends on NET
Alex Tsariounov:
o ia64: fix comment-typo in entry.S
Alex Williamson:
o ia64: setup cpu_to_node for cpus not present
Alexander Viro:
o sparse: missed setsockopt wrappers
o sparse: dev_ioctl() callers
o sparse: missed piece of sock_get_timestamp() annotation
o sparse: trivial part of net/* annotation
o sparse: more net/* annotation
o sparse: net/* non-ANSI argument lists
o sparse: amd64 annotation - beginning
o sparse: amd64 annotations - trivial part
o sparse: amd64 - more trivial annotations
o sparse: amd64 - #if abuse fixes
o sparse: amd64 - long constants
o sparse: amd64 - rest
o sparse: fs/compat.c::copy_iocb() cleanup
o sparse: tty_io annotation
o sparse: tty_ioctl annotation
o sparse: vt and friends
o sparse: n_tty annotation
o sparse: tty_driver ->write_proc()
o sparse: SIOCGIFCONF handling - the rest of it
o sparse alpha: beginning of __user annotation
o sparse alpha: long constants
o sparse alpha: #if abuses
o sparse alpha: trivial parts of __user annotation
o sparse alpha: utimes()
o sparse alpha: the rest of it
o trivial annotation for arch/i386/kernel/*
o sparse: reiserfs annotation
o sparse: asus_acpi dereference of userland pointers
o sparse: sound sb fix
o sparse: msnd sound fix
o sparse: sound/core/timer.c fix
o sparse: sound/core annotation - trivial part
o sparse: annotation of include/sound (generic parts)
o sparse: sound/core/pcm* annotation
o sparse: opl annatation
o sparse: rest of sound/core
o sparse: sound/isa/gus annotation
o sparse: sscape annotation
o sparse: wavefront annotation
o sparse: sb annotation
o sparse: vx annotation
o sparse: trident annotation
o sparse: rme9652 annotation
o sparse: emu10k1 annotation
o sparse: rest of ALSA annotation
o sparse: cs4281 annotation
o sparse: cmpci annotation
o sparse: es1371 annotation
o sparse: es1370 annotation
o sparse: esssolo annotation
o sparse: rme96xx annotation
o sparse: i810 annotation
o sparse: cs46xx annotation
o sparse: ali5455 annotation
o sparse: sonicvibes annotation
o sparse: maestro3 annotation
o sparse: the rest of sound/* annotation
o sparse: ad1889 annotation
o sparse: btaudio annotation
o sparse: maestro annotation
o sparse: ymfpci annotation
o sparse: msnd annotation
o sparse: forte annotation
o sparse: drivers/cpufreq annotation
o sparse: drivers/cdrom annotation
o sparse: drivers/md annotation
o sparse: drivers/pnp annotation
o sparse: DAC960 annotation
o sparse: cciss annotation
o sparse: cpqarray annotation
o sparse: paride annotation
o sparse: rest of drivers/block annotation
o drivers/atm/ambassador.c delousing
o drivers/atm/horizon.c delousing
o sparse: atm annotation (core)
o sparse: atm annotation - drivers
o sparse: coda annotation
o sparse: decnet annotation
o sparse: ax25 annotation
o sparse: netrom annotation
o sparse: rose annotation
o sparse: x25 annotation
o sparse: ipx annotation
o sparse: appletalk annotation
o sparse: econet annotation
o sparse: wanrouter annotation
o sparse: net/bridge annotation
o sparse: netlink annotation
o sparse: irda annotation
o sparse: vlan annotation
o sparse: bluetooth annotation
o sparse: sctp annotation
o sparse: more netfilter annotation
o sparse: mroute annotation
o sparse: ->ifr_data annotation
o sparse: ->ifr_data fixes
o sparse: more wan annotations
o sparse: selinux annotation
o sparse: rest of net/* annotations (in this patchset, that is ;-)
o sparse: hugetlb sysctl annotation
o sparse: gdth.c annotation
o sparse: misc scsi annotation
o sparse: rest of drivers/net annotations in the patchset
o sparse: misc headers annotations
o sparse: ide-floppy annotation
o sparse: synclink_cs annotation
o sparse: synclink annotation
o sparse: trivial fs annotations
o sparse: usb/core/devio annotation
o sparse: sonypi annotiation
o sparse: ppdev annotation
o sparse: misc drivers annotations
o sparse: lp annotation
o sparse: drivers/char/i8 annotation
o sparse: tpqic annotation
o sparse: if_mii() helper (from jgarzik)
o sparse: the rest of ifr_data cleanups and annotations
o sparse: drivers/net/tun.c annotation
o sparse: depca annotation
o sparse: slip.c annotation
Andi Kleen:
o x86-64 update
o Really fix empty node 0 on x86-64
o Fix compilation on x86-64
o Fix x86-64 compilation without CONFIG_NUMA
o i386: add missing bitop.h memory clobbers
o More x86-64 bugfixes
o Use KERN_ALERT more for oopses
Andreas Dilger:
o ext3: htree rename fix
Andrew Morton:
o s2io section fix
o ramfs o_sync oops fix
o shrink_all_memory() fixes
o Add reference_init.pl to `make buildcheck' target
o direct-io invalidation fix
o Add the sixth arg to the sys_futex() prototype
o nfs-direct warning fix
Andries E. Brouwer:
o partition table validity checking
Anton Altaparmakov:
o NTFS: Add a new address space operations struct, ntfs_mst_aops, for
mst protected attributes. This is because the default ntfs_aops do
not make sense with mst protected data and were they to write
anything to such an attribute they would cause data corruption so
we provide ntfs_mst_aops which does not have any write related
operations set.
o NTFS: Cleanup dirty ntfs inode handling (fs/ntfs/inode.[hc]) which
also includes an adapted ntfs_commit_inode() and an implementation
of ntfs_write_inode() which for now just cleans dirty inodes
without writing them (it does emit a warning that this is
happening).
o NTFS: 2.1.12 release - Fix the second fix to the decompression
engine
Anton Blanchard:
o ppc64: add eeh_add_device_early/late
o ppc64: reset iseries progress indicator on boot
o ppc64: bolt first vmalloc segment into SLB
o ppc64: SLB accounting fix
o ppc64: iseries bolted SLB fix
Arjan van de Ven:
o [libata] Use standard headers from include/scsi, not drivers/scsi
Arun Sharma:
o ia64: tighten FPH state context switch check
o ia64: disable in-kernel ia32 emulation post Madison
Bartlomiej Zolnierkiewicz:
o ide: fix for generic IDE PCI module
o ide: ide_pci_device_t sanitization
o ide: merge amd74xx.h into amd74xx.c
o ide: use generic ide_init_hwif_ports() on m68k
o ide: use <asm-i386/ide.h> as <asm-x86_64/ide.h>
o ide: add IDE_ARCH_OBSOLETE_DEFAULTS
o ide: remove useless /proc/ide/siimage from siimage.c
o ide: simplify CONFIG_IDEDMA_ONLYDISK logic a bit
o DMA mode setup fixes for piix.c/ata_piix.c
o apply nForce2 fixup only if C1 Halt Disconnect is enabled
o ide: remove unused ide_remove_setting() from ide.c
o ide: use try_to_flush_leftover_data() in idedisk_error()
o ide: remove needless exports
o ide: sanitize __ide_do_rw_disk() debugging code
o ide: remove dead code from ide-taskfile.c
Ben Dooks:
o [ARM PATCH] 1906/1: S3C2410: serial minor number fix
Benjamin Herrenschmidt:
o Fix rivafb's OpenFirmware parsing
o ppc32: Fix CPUs with soft loaded TLB
o ppc32: missed flush_tlb_page_nohash() cases
o ppc64: make the TLB flush logic match ppc32
Bert Hubert:
o s/tkill/tgkill/ in /** documentation */
Bjorn Helgaas:
o ia64: fix hotplug config placement
o active_load_balance() deadlock
Brian Gerst:
o Mark cache_names __initdata
Brian Lazara:
o ide: add new nForce IDE/SATA device IDs to amd74xx.c
Catalin Marinas:
o [ARM PATCH] 1912/1: Wrong cache aliasing bit check
Chris Mason:
o direct-io hole fix
Christoph Hellwig:
o farsync needs <asm/io.h>
o pnpbios only makes sense for X86
o linux/timer.h needs linux/stddef.h
o fix mca procfs stub
o [SBUS]: Define dma direction bits in terms of enum
dma_data_direction
o [ATM]: Simplify fore200e DMA macros even further
o [SBUS]: Further cleanup of scsi driver header usage
o ide: change CONFIG_IDEDISK_STOKE to a runtime option
o [WATCHDOG] linux/watchdog.h include types.h patch
Colin Gibbs:
o Fix readahead handling in knfsd
Colin Leroy:
o Fix DRM mismerge(?)
Corey Minyard:
o Fix procfs warnings when removing ipmi_si module
o Fix ipmi compile failure
Darren Williams:
o ia64: define pending_irq_cpumask in irq.c
Dave Airlie:
o fix missing DRM_ERR()s - Eric Anholt
o add R200_EMIT_RB3D_BLENDCOLOR state packet to support
GL_EXT_blend_color, GL_EXT_blend_func_separate and
GL_EXT_blend_equation_separate on r200 from Roland Scheidegger
o MTD: add st m50fw0* to jedec_probe.c
o Remove check that is needed only for 2.4 kernel fixes bug reports
on lk
Dave Jones:
o [AGPGART] If ati_create_gatt_pages() fails, don't propagate an
address we've freed
o [AGPGART] Various Intel/EM64T AGP fixes
o [AGPGART] Don't abort if Intel-agp can't find AGP capability
o [AGPGART] Make the driver only announce the PCI IDs it actually
supports
o [AGPGART] kill trailing whitespace & indentation fixes
o [AGPGART] Make the VIA AGP driver only announce PCI IDs it actually
supports
o [AGPGART] Make the Nvidia AGP driver only announce PCI IDs it
actually supports
o [AGPGART] Add missing SIS755 ID to AMD64 AGP driver From: Andi
Kleen
o [AGPGART] Remove lots of trailing whitespace from amd64 gart driver
o [AGPGART] Fix a potential leak in intel driver
o [CPUFREQ] Silence noisy debugging printk in longhaul driver
o [CPUFREQ] Convert longhaul driver to use module_param
o [CPUFREQ] Remove some unneeded includes
o [CPUFREQ] Use correct printk prefix in p4-clockmod driver
o [CPUFREQ] Quieten the powernow-k7 init printk a little
o [CPUFREQ] Remove a bunch of trailing whitespace from the
powernow-k8 driver
o [CPUFREQ] Fix leak in powernow-k8 Spotted by Yury Umanets
o [CPUFREQ] Make longhaul debug a module option
o [CPUFREQ] Scaling on VIA C3 Nehemiah works now
o [CPUFREQ] fix panic in powernow_k8_verify on MP but PSB systems
o [CPUFREQ] Make powernow-k7 module parameter not need powernow
prefix
o [AGPGART] Fix broken serverworks tlb flush routine
o [AGPGART] Make agp=off work again
o [AGPGART] Improve the resume functions for Intel AGP bridges by
restoring config space (the bios might not have done that).
o [CPUFREQ] Fix build of longhaul
o Cset exclude: [email protected]|ChangeSet|20040523113850|65135
o [AGPGART] intel-agp: skip non-AGP devices From: Matt Domsch.
o [CPUFREQ] Consolidate version checking in longhaul_get_cpu_mult
reduce code duplication.
o [CPUFREQ] Remove duplicate declaration of debug var from longhaul
o [CPUFREQ] Change longhaul debugging info printk's to dprintk's
o [CPUFREQ] Fix powernow-k7 when ACPI_PROCESSOR built as module
o [CPUFREQ] Remove lots of redundant code from longhaul driver
o [CPUFREQ] Document some oddness in the longhaul driver
o [CPUFREQ] Unify the EBLCR parsing code in longhaul
o [CPUFREQ] Remove duplicate debug printk from longhaul driver
o [CPUFREQ] Move longhaul multiplier debug printk to somewhere more
useful
o [CPUFREQ] Remove bogus longhaul v4 The code only supports 3
versions, so numbering them 1,2 and 4 doesn't make a lot of sense.
o [CPUFREQ] Convert longhaul debug printk to use varargs
o fix warning in synclink.c
o Fix missing padding in DMI table
David Gibson:
o Add watchdog timer to iseries_veth driver
o hugetlbpage msync() fix
o hugetlbpage: reinitialise compound page destructor
David Goodenough:
o Support for SC1100
David Mosberger:
o ia64: Reserve syscall # for vserver()
o Update ia64 linux mailing-list and web-site addresses
o ia64: Add defconfig for Ski simulator (patch by Darren Williams)
o move #endif to correct place
o ia64: Avoid intermediate-overflows in sched_clock()
o ia64: Kill unused external declaration of "acpi_legacy_devices"
David S. Miller:
o [SPARC64]: Update defconfig
o [SPARC]: Do tty_flip_buffer_push outside of port lock
o [PKT_SCHED]: Missing rta_len init in sch_delay
o [SPARC64]: Fix NR_IRQS check in hardirq.h
o [SPARC64]: Annotate 64-bit constants with 'UL'
o [SPARC64]: __volatile --> __volatile__
o [SPARC64]: Lots of sparse work for arch/sparc64
o [SPARC]: Set sparse arch defines explicitly
o [SPARC]: First stage of sparc32 sparse work
o [IPV6]: ndisc_dst_alloc can never get a NULL dev
o [SPARC64]: Compat syscall overhaul
o [SPARC64]: Kill _exit kernelsyscall stub from unistd.h
o [ATM]: fore200e dma direction macro tests no longer needed
o [SPARC64]: New 6th arg of sys_futex needs sign extension
o [SPARC64]: Update defconfig
o [NETFILTER]: In ipt_TCPMSS, SYN packets are never hw checksummed
o [SPARC64]: Fix compat_sys_wait4 extern decl
o [SPARC64]: Tidy asm macros in sfp-util.h
o [COMPAT]: Add __user attributes for pointers passed while KERNEL_DS
o [AIO]: Split ki_user_obj into a 2 member union
o [COMPAT]: Fix some sparse annotations in fs/compat_ioctl.c
o [KERNEL]: Some sparse fixes for init/main.c
o [TCP]: Need tcp_get_info decl in net/tcp.h
Diego Calleja Garc?a:
o Document checkstacks
Dmitry Torokhov:
o Input: synaptics driver cleanup
o Input: support Synaptics touchpads that have separate middle button
o Input: pass maximum allowed protocol to psmouse_extensions instead
of accessing psmouse_max_proto directly allowing to avoid changing
the global parameter when synaptics initialization fails
o Input: Change spurious ACK warning in atkbd to soften accusation
against XFree86
o Input: fix trailing whitespace in atkbd
o Input: remove unneeded fields in atkbd structure, convert to
bitfields
o Input: Do not generate events from atkbd until keyboard is
completely initialized. It should suppress messages about suprious
NAKs when controller's timeout is longer than one in atkbd
o Input: when getting a new device announce (0xAA 0x00) in psmouse
try reconnecting instead of rescanning to preserve (if possible)
the same input device.
o Input: move "reconnect after so many errors" handling from
synaptics driver to psmouse so it can be used by other PS/2 protcol
drivers (but so far
o Input: add protocol_handler to psmouse structure to ease adding new
protocols to psmouse module
o Input: add psmouse_sliced_command (passes extended commands encoded
with 0xE8 to the mouse) and use it in Synaptics and Logitech
drivers
o Input: do not modify device's properties when probing for protocol
extensions on reconnect as it may interfere with reconnect process
o Input: allow disabling legacy psaux device even for non-embedded
systems
o Input: serio trailing whitespace fixes
o Input: make serio open and close methods optional
o Input: trailing whitespace fixes
o Input: - move set_abs_params from synaptics driver to input and
rename to input_set_abs_params
o Input: trailing whitespace fixes in drivers/input/serio
o Input: kbd98io_interrupt should return irqreturn_t
o Input: kbd98_interrupt should return irqreturn_t
o Input: various fixes for H3600 touchscreen driver
o Input: twidjoy module
o Input: trailing whitespace fixes in drivers/input/keyboard
o Input: power - add MODULE_LICENSE
o Input: trailing whitespace fixes in drivers/input/joystick
o Input: trailing whitespace fixes in drivers/input/gameport
o Input: trailing whitespace fixes in drivers/input
o Input: do not call synaptics_init unless we are ready to do full
mouse setup
o Input: split i8042 interrupt handling into an IRQ handler and a
tasklet
o Patch from Sau Dan Lee Input: i8042 - kill the timer only after
removing interrupt handler,
o Input: mousedev - better multiplex absolute and relative devices;
cleanups
Eugene Surovegin:
o ppc32: reorg DMA API, add coherent alloc in irq
Gary Hade:
o ia64: Fix early serial console setup regression on the IBM x450 and
x455
George France:
o cpqarray.c: seed the random number pool
Herbert Xu:
o [IPSEC]: Fix xfrm_tunnel leak
o Check cmd in plip_ioctl
Hideaki Yoshifuji:
o [IPV6] use appropriate __alignof__() for mibs
o [NET] Introduce and use several common stuff for snmp item list
o [IPV{4,6}] introduce ip-independent ipstats_mib based on ipv6_mib
o [IPV6] rename snmp6_ipv6_list to snmp6_ipstats_list
o [IPV4] use ip-independent ipstats_mib to store IPv4 statistics
o [IPV6]: Store idev in routes
o [IPV4]: Store idev in routes
o [UDPv4]: Pass correct socket to ip_mc_sf_allow
Hugh Dickins:
o mm: swapper_space.i_mmap_nonlinear
o mm: follow_page invalid pte_page
o mm: vma_adjust adjust_next wrap
o mm: vma_adjust insert file earlier
o mm: get_user_pages vs. try_to_unmap
o mm: kill missed pte warning
Ian Campbell:
o [ARM PATCH] 1890/2: Consolidate CPUFREQ handling in SOC PCMCIA
driver
Ingo Molnar:
o md.c message during quiet boot
o sched: remove noinline workaround
o sched: honor the "sync" wakeup bit
Ivan Kokshaysky:
o ide: don't put disks in standby mode on halt on Alpha
J. Bruce Fields:
o kNFSd: exp_find(): remove null pointer check
o kNFSd: nfsd_acceptable() typo fix
o kNFSd: nfsd4 xdr name encoding improvements
o kNFSd: gss_svc locking and refcounting fixes
o kNFSd: rsc_lookup simplification
o kNFSd: nfsd4_release_lockowner() oops fix
o kNFSd: nfsd getattr fix
o kNFSd: nfsd4 setclientid fix
o kNFSd: nfsd4 file creation fix
o kNFSd: documentation typo fixes
Jack Steiner:
o sched: balance-on-exec fix
Jakub Jel?nek:
o Add FUTEX_CMP_REQUEUE futex op
Jamal Hadi Salim:
o [NET]: Add ARPHRD_NONE and use it in tun driver
o [NETFILTER]: Small cleanup for {ipt,ip6t,arpt}_find_target
James Morris:
o [NETFILTER]: Fix checksum bug for multicast/broadcast packets on
postrouting hook
o [IPVS]: IPVS needs checksum fixups
Jan Kara:
o quota: fix writing of quota info
o quota: fix for old quota format
Jeff Garzik:
o [netdrvr e1000] use generic ethtool_ops provided in
net/core/ethtool.c
o [netdrvr 8139too] remove bogus config option test
o Cset exclude: [email protected]|ChangeSet|20040527204246|14084
o [libata promise] revert broken taskfile delivery change
o [libata scsi] ack SYNCHRONIZE CACHE command
Jeff Mahoney:
o ext3_orphan_del may double-decrement bh->b_count
Jens Axboe:
o fix cdrom length check
o kill drivers/ide TCQ support
Jeremy Kerr:
o Fix signal race during process exit
Jerzy Szczepkowski:
o bug in sys_io_setup
Jesse Barnes:
o ia64: don't udelay() in sn_mmiob
Jorge Hernandez-Herrero:
o [SCTP] Fix to not setup a new association if the endpoint is in
SHUTDOWN_ACK_SENT state and recognizes that the peer has restarted.
o [SCTP] Fix to not start a new association on a 1-many style
sendmsg() with MSG_EOF/MSG_ABORT flag and no data.
Jurriaan:
o radeonfb fix (non-8bpp clear doesn't use palette)
J?rn Engel:
o Improve `make checkstack'
Keith Owens:
o ia64: work around linker bug
Kevin Corry:
o dm: dm-ioctl.c: change an int* to a size_t*
Krzysztof Halasa:
o Re: [Fwd: [PATCH] Stop queue on close in hdlcdrv]
o 2.6 Generic HDLC update
Linus Torvalds:
o Add __user annotations to sock_get_timestamp()
o ppc64: fix untyped large constants so that they don't cause sparse
to warn about implicit typing.
o ppc64: use "ASM_CONST()" to give proper C type to constants that
can also be used in assembly language context.
o ppc64: more explicitly typed constants
o ppc64: mark the "regshere" marker with proper type information
o sparse: add "__force" type attribute for forcing a cast
o ppc64: check more of the user access functions for proper arguments
o ppc64: missing __user annotations noticed by stricter checks
o sparse: make x86 user pointer checks stricter
o ppc64: fix more user pointers in signal handling
o ppc64: add more user annotations to ptrace.c
o ppc64: more user address fixups
o sparse: make x86 and ppc64 set the architecture-specific #define's
explcitly.
o sparse: use new generic __chk_user_ptr() macro in x86/ppc64/sparc*
o Start documenting the sign-off procedure in SubmittingPatches
o Add comments on load balancing special cases
o sparse: fix up futex address space warning
o sparse: annotate (and comment on) kmod.c user pointer usage
o rpc: Use proper printk format for size_t argument
o Linux 2.6.7-rc3
Luiz Capitulino:
o fix net/ixgb/ixgb_main.c warning
o fix possible NULL pointer in fs/ext3/super.c
o mm/oom_kill.c trivial cleanup
Maeda Naoaki:
o ia64: fix /proc/ioports regression in 2.6.6 on Tiger4
Marc-Christian Petersen:
o vesafb: vram boot option the same as 2.4.x
Marcel Holtmann:
o [Bluetooth] Move function exports out of syms.c
o [Bluetooth] Remove CVS tags and cleanup the code
o [Bluetooth] Update Kconfig help entries
o [Bluetooth] Allocate protocol number for HIDP support
o [Bluetooth] Add dynamic PSM allocation for L2CAP server sockets
o [Bluetooth] Add dynamic channel allocation for RFCOMM server
sockets
o [Bluetooth] Add quirk for broken RTX Telecom based dongles
Marcus Meissner:
o dm: add DM_REMOVE_ALL_32 compat ioctl
Margit Schubert-While:
o prism54: delete cvs tags
o prism54: Add new private ioctls
o prism54: Reset card on tx timeout
o prism54: Add iwspy support
o prism54: Add avs header support
Martin Schwidefsky:
o s390: core
o s390: common i/o layer
o s390: block device driver
o s390: network device driver
Matt Porter:
o ppc32: add "indirect" DCR access, pass 2
Matt Tolentino:
o kill off efi_dir in efi.h
o update elilo loader location in Kconfig
Matthew Wilcox:
o Better tulip handling on PA-RISC
Mikael Starvik:
o CRIS architecture update
Nathan Lynch:
o ppc64 gives up too quickly on hotplugged cpu
o use c99 struct initializer in hotcpu_notifier
Neil Brown:
o md: rationalise device selection in md/multipath
o md: make sure md_check_recovery will remove a faulty device when
->nr_pending hits 0
o md: allow an md personality to refuse a hot-remove request
o md: make sure the size of a raid5/6 array is a multiple of the
chunk size
o md: abort the resync of raid1 there is only one device
o md: allow md arrays to be resized if devices are large enough
o md: support reshaping raid1 arrays - adding or removing drives
o md: fix BUG in raid5 resync code
Nick Piggin:
o sched: improve wakeup-affinity
Nicolas Pitre:
o [ARM PATCH] 1896/2: distinguish between memory and LCD clock on PXA
o [ARM PATCH] 1910/1: rework Mainstone and IDP include files
Olaf Hering:
o ppc64: kernel Makefile options for $(AS)
o ppc64: update info about available iseries_veth interfaces
o [NET]: Fix compat bug in recvmsg/sendmsg wrt MSG_CMSG_COMPAT
Olaf Kirch:
o kNFSd: Fix nfs3 dentry encoding
Oliver Heilmann:
o [AGPGART] SIS AGP updates
Pat Gefre:
o ia64: for SN2, disable MST errors on probe
Patrick J. LoPresti:
o Better names for EDD legacy_* fields
o Use decimal instead of hex for EDD values
Paul Clements:
o md: handle hot-add for arrays with non-persistent superblocks
Paul Fulghum:
o synclinkmp.c driver init modifications
o synclink_cs.c driver init modifications
o synclink.c driver init modifications
Paul Jackson:
o ia64: uninline find_next_bit on ia64
Paul Mackerras:
o ppc64: fix missing RELOCs, add linux,phandle property
o ppp ldisc close deadlock prevention
o ppc32: Add _raw_write_trylock
o ppc32: Don't synchronize in disable_irq() if no handler
o ppc32: Suppress bogus info in /proc/ppc_htab on 64-bit cpus
o ppc32: Use -fPIC instead of -mrelocatable-lib
o ppc32: Make ppc32 PCI code more robust
o ppc32: Fix preemptible check
o ppc32: Reduce WARN_ON(0) to nothing
o ppc32: Fix locks.c properly this time
o ppc64: don't clear MSR.RI in do_hash_page_DSI
o ppc32: Fix typo in ppc32 spinlock.h
Paul Slootman:
o floppy minor number fix
Peter Osterlund:
o Avoid excessive stack usage in NFS
Piotr Kaczuba:
o make vram boot option actually work
P?draig Brady:
o [WATCHDOG] v2.6.6 w83627hf_wdt.c-patch
Rob Melby:
o evdev.c
Russell King:
o [ARM] Add two more missing __user annotations for sparse
o [PCMCIA] Fix up SOC PCMCIA socket timing calculations
o [ARM] Eliminate meminfo 'end' element
o [ARM] Correct permissions on several ARMv6 files
o [ARM] Update ARM memory region reservations
o [ARM] Fix missing spinlock initialisation
o [ARM] asm_do_IRQ takes an unsigned IRQ number
o Fix loop device cache handling
o Export swapper_space
o [ARM] Don't include lubbock.h in asm/arch/hardware.h
o [PCMCIA] Add sparse annotations to ds.c
o [ARM] Uninline flush_dcache_page()
o [ARM] Update mach-types
Rusty Russell:
o Export kthread primitives
o fix sysfs node cpumap for large NR_CPUS
Sridhar Samudrala:
o [SCTP] Fix the use of cached non-zero vtag in an INIT that is
resent after a stale cookie error.
o [SCTP] Fix missing VTAG validation on certain incoming packets
o [SCTP] Fix to wakeup blocking connect() after max INIT retries
failed
o [SCTP] Fix poll() on a 1-1 style socket so that it returns when the
association is aborted by peer.
Stephen D. Smalley:
o selinux: check processed security context length
Stephen Hemminger:
o [TCP]: Common code for generating tcp_info
Stephen Rothwell:
o ppc64: iSeries default config update
o ppc64: iSeries virtual ethernet minor optimisation
o ppc64: iSeries fix virtual ethernet transmit block
Takashi Iwai:
o input: Fix oops in hiddev
Todd E. Johnson:
o input: Microtouch USB driver update Changed reset from standard USB
dev reset to vendor reset Changed data sent to host from
compensated to raw coordinates Eliminated vendor/product module
params Performed multiple successfull tests with an EXII-5010UC
Tom 'spot' Callaway:
o [SBUS]: Stop using drivers/scsi/{scsi,hosts}.h in drivers
Tony Lindgren:
o [ARM PATCH] 1902/1: OMAP update 1/3: entry-armv.S
Toshihiro Kobayashi:
o [ARM PATCH] 1911/1: fix of odd PMD handling in
do_translation_fault()
Vojtech Pavlik:
o input: Chips passing MUX detection incorrectly due to USB Legacy
support report MUX version 10.12, not 12.10. Fixed.
o input: Profusion/ServerWorks chipset workaround in i8042.c for Ingo
Molnar
o input: Fix emulation of mouse reset (0xff) command
o input: Check for IM Explorer mice even if IMPS check failed
o input: Exclude tasklet changes to i8042.c
William Lee Irwin III:
o use aio workqueue in fs/aio.c
o correct use_mm()/unuse_mm() to use task_lock() to protect ->mm
o use const in time.h unit conversion functions
o aio: fix io_getevents() timer expiry setting
On Mon, Jun 07, 2004 at 12:32:45PM -0700, Linus Torvalds wrote:
>
> Ok, let's calm down for a while before the final 2.6.7.
>
> -rc3 does a lot of sparse type cleanup, mainly thanks to Al Viro (but his
> work ended up getting some other people involved too, since the list of
> sparse warnings isn't as daunting any more). Some of that has unearthed
> real bugs which Al fixed.
>
> But there are DRM, AGP, cpufreq, sparc64, and input updates there too. See
> the appended shortlog for more details,
This one here:
diff -Nru a/include/linux/netfilter_arp.h b/include/linux/netfilter_arp.h
--- a/include/linux/netfilter_arp.h 2004-06-07 21:58:09 +02:00
+++ b/include/linux/netfilter_arp.h 2004-06-07 21:58:09 +02:00
@@ -17,4 +17,5 @@
#define NF_ARP_FORWARD 2
#define NF_ARP_NUMHOOKS 3
+static DECLARE_MUTEX(arpt_mutex);
#endif /* __LINUX_ARP_NETFILTER_H */
looks perfectly fucked up. This way every file including netfilter_arp.h
will get it's own copy of arpt_mutex which certainly doesn't help
synchronization.
But in fact I can only see a single file actually using it, that's
arp_tables.c where it was defined previously. What's going on here?
On Mon, 7 Jun 2004, Christoph Hellwig wrote:
>
> This one here:
>
> diff -Nru a/include/linux/netfilter_arp.h b/include/linux/netfilter_arp.h
> --- a/include/linux/netfilter_arp.h 2004-06-07 21:58:09 +02:00
> +++ b/include/linux/netfilter_arp.h 2004-06-07 21:58:09 +02:00
> @@ -17,4 +17,5 @@
> #define NF_ARP_FORWARD 2
> #define NF_ARP_NUMHOOKS 3
>
> +static DECLARE_MUTEX(arpt_mutex);
> #endif /* __LINUX_ARP_NETFILTER_H */
>
> looks perfectly fucked up.
Agreed. David? Jamal?
(Hey, first time when I used the "Signed-off-by:" info. Very good).
Linus
ACPI seems to be broken on Intel D865PERL board, I can only boot with
acpi=off, otherwise the kernel either hangs after the "ACPI: Subsystem
revision 20040326" message or shows a seemingly random oops message.
Tom
On Mon, 7 Jun 2004 14:09:04 -0700 (PDT)
Linus Torvalds <[email protected]> wrote:
> On Mon, 7 Jun 2004, Christoph Hellwig wrote:
> >
> > This one here:
> >
> > diff -Nru a/include/linux/netfilter_arp.h b/include/linux/netfilter_arp.h
> > --- a/include/linux/netfilter_arp.h 2004-06-07 21:58:09 +02:00
> > +++ b/include/linux/netfilter_arp.h 2004-06-07 21:58:09 +02:00
> > @@ -17,4 +17,5 @@
> > #define NF_ARP_FORWARD 2
> > #define NF_ARP_NUMHOOKS 3
> >
> > +static DECLARE_MUTEX(arpt_mutex);
> > #endif /* __LINUX_ARP_NETFILTER_H */
> >
> > looks perfectly fucked up.
>
> Agreed. David? Jamal?
It happens to be OK since arp_tables.c is the only includer of that header
but I agree it's gross and I'll put it back into arp_tables.c
I am running the sparse regressions on -rc3 as well. Will post them on
the compile regression site when they are complete.
----------------------------------------
Linux 2.6 Compile Statistics (gcc 3.2.2)
Warnings/Errors Summary
Kernel bzImage bzImage bzImage modules bzImage modules
(defconfig) (allno) (allyes) (allyes) (allmod) (allmod)
----------- ----------- -------- -------- -------- -------- ---------
2.6.7-rc3 0w/0e 0w/0e 108w/ 0e 5w/0e 2w/0e 104w/0e
2.6.7-rc2 0w/0e 0w/0e 110w/ 0e 5w/0e 2w/0e 106w/0e
2.6.7-rc1 0w/0e 0w/0e 111w/ 0e 6w/0e 2w/0e 107w/0e
2.6.6 0w/0e 0w/0e 123w/ 0e 7w/0e 4w/0e 121w/0e
2.6.6-rc3 0w/0e 0w/0e 124w/ 0e 7w/0e 5w/0e 121w/0e
2.6.6-rc2 0w/0e 0w/0e 122w/ 0e 7w/0e 4w/0e 121w/0e
2.6.6-rc1 0w/0e 0w/0e 125w/ 0e 7w/0e 4w/0e 123w/0e
2.6.5 0w/0e 0w/0e 134w/ 0e 8w/0e 4w/0e 132w/0e
2.6.5-rc3 0w/0e 0w/0e 135w/ 0e 8w/0e 4w/0e 132w/0e
2.6.5-rc2 0w/0e 0w/0e 135w/ 0e 8w/0e 3w/0e 132w/0e
2.6.5-rc1 0w/0e 0w/0e 138w/ 0e 8w/0e 3w/0e 135w/0e
2.6.4 1w/0e 0w/0e 145w/ 0e 7w/0e 3w/0e 142w/0e
2.6.4-rc2 1w/0e 0w/0e 148w/ 0e 7w/0e 3w/0e 145w/0e
2.6.4-rc1 1w/0e 0w/0e 148w/ 0e 7w/0e 3w/0e 145w/0e
2.6.3 1w/0e 0w/0e 142w/ 0e 9w/0e 3w/0e 142w/0e
2.6.3-rc4 1w/0e 0w/0e 142w/ 0e 9w/0e 3w/0e 142w/0e
2.6.3-rc3 1w/0e 0w/0e 145w/ 7e 9w/0e 3w/0e 148w/0e
2.6.3-rc2 1w/0e 0w/0e 141w/ 0e 9w/0e 3w/0e 144w/0e
2.6.3-rc1 1w/0e 0w/0e 145w/ 0e 9w/0e 3w/0e 177w/0e
2.6.2 1w/0e 0w/0e 152w/ 0e 12w/0e 3w/0e 187w/0e
2.6.2-rc3 0w/0e 0w/0e 152w/ 0e 12w/0e 3w/0e 187w/0e
2.6.2-rc2 0w/0e 0w/0e 153w/ 8e 12w/0e 3w/0e 188w/0e
2.6.2-rc1 0w/0e 0w/0e 152w/ 0e 12w/0e 3w/0e 187w/0e
2.6.1 0w/0e 0w/0e 158w/ 0e 12w/0e 3w/0e 197w/0e
2.6.1-rc3 0w/0e 0w/0e 158w/ 0e 12w/0e 3w/0e 197w/0e
2.6.1-rc2 0w/0e 0w/0e 166w/ 0e 12w/0e 3w/0e 205w/0e
2.6.1-rc1 0w/0e 0w/0e 167w/ 0e 12w/0e 3w/0e 206w/0e
2.6.0 0w/0e 0w/0e 170w/ 0e 12w/0e 3w/0e 209w/0e
Web page with links to complete details:
http://developer.osdl.org/cherry/compile/
Daily compiles (ia32):
http://developer.osdl.org/cherry/compile/2.6/linus-tree/running.txt
Daily compiles (ia64):
http://developer.osdl.org/cherry/compile/2.6/linus-tree/running64.txt
Latest changes in Linus' bitkeeper tree:
http://linux.bkbits.net:8080/linux-2.5
John
On Mon, Jun 07, 2004 at 05:28:54PM -0700, John Cherry wrote:
> I am running the sparse regressions on -rc3 as well. Will post them on
> the compile regression site when they are complete.
>
> ----------------------------------------
>
> Linux 2.6 Compile Statistics (gcc 3.2.2)
> Warnings/Errors Summary
>
> Kernel bzImage bzImage bzImage modules bzImage modules
> (defconfig) (allno) (allyes) (allyes) (allmod) (allmod)
> ----------- ----------- -------- -------- -------- -------- ---------
> 2.6.7-rc3 0w/0e 0w/0e 108w/ 0e 5w/0e 2w/0e 104w/0e
...
> 2.6.0 0w/0e 0w/0e 170w/ 0e 12w/0e 3w/0e 209w/0e
A quite impressive improvement.
Nice going, everyone!
Regards: David Weinehall
--
/) David Weinehall <[email protected]> /) Northern lights wander (\
// Maintainer of the v2.0 kernel // Dance across the winter sky //
\) http://www.acc.umu.se/~tao/ (/ Full colour fire (/
This failed to build for me:
% make oldconfig; make
<snip>
CHK include/linux/compile.h
UPD include/linux/compile.h
CC init/version.o
LD init/built-in.o
LD .tmp_vmlinux1
arch/i386/mach-generic/built-in.o(.init.text+0x39e): In function `parse_unisys_oem':
: undefined reference to `platform_rename_gsi'
make: *** [.tmp_vmlinux1] Error 1
So I switched from the default CONFIG_X86_GENERICARCH to
CONFIG_X86_PC, which is probably the more correct option for my system
(dual Pentium-III, 2 GB memory). Then it built just fine.
-jf
I am not sure whether it is allowed to report problems with
NVidia graphics support to this list. I just want to make
sure that this hiccup isn't lost:
Jun 8 15:34:08 styx kernel: Badness in pci_find_subsys at drivers/pci/search.c:167
Jun 8 15:34:08 styx kernel: [dump_stack+30/48] dump_stack+0x1e/0x30
Jun 8 15:34:08 styx kernel: [pci_find_subsys+212/224] pci_find_subsys+0xd4/0xe0
Jun 8 15:34:08 styx kernel: [pci_find_device+49/64] pci_find_device+0x31/0x40
Jun 8 15:34:08 styx kernel: [pci_find_slot+41/80] pci_find_slot+0x29/0x50
Jun 8 15:34:08 styx kernel: [pg0+555404950/1069666304] os_pci_init_handle+0x35/0x62 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+553916511/1069666304] _nv001243rm+0x1f/0x24 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+554642013/1069666304] _nv003797rm+0xa9/0x128 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+555087009/1069666304] _nv001490rm+0x55/0xe4 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+555254100/1069666304] _nv000816rm+0x334/0x384 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+555252815/1069666304] _nv000809rm+0x2f/0x34 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+554637136/1069666304] _nv003816rm+0xf0/0x104 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+554631182/1069666304] _nv003795rm+0x6ea/0xaec [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+555072423/1069666304] _nv001476rm+0x277/0x45c [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+553927578/1069666304] _nv000896rm+0x4a/0x64 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+553933748/1069666304] rm_isr_bh+0xc/0x10 [nvidia]
Jun 8 15:34:08 styx kernel: [pg0+555395476/1069666304] nv_kern_isr_bh+0x11/0x15 [nvidia]
Jun 8 15:34:08 styx kernel: [tasklet_action+70/112] tasklet_action+0x46/0x70
Jun 8 15:34:08 styx kernel: [__do_softirq+140/144] __do_softirq+0x8c/0x90
Jun 8 15:34:08 styx kernel: [do_softirq+43/48] do_softirq+0x2b/0x30
Jun 8 15:34:08 styx kernel: [do_IRQ+213/256] do_IRQ+0xd5/0x100
Jun 8 15:34:08 styx kernel: [common_interrupt+24/32] common_interrupt+0x18/0x20
Jun 8 15:34:08 styx kernel: [do_select+359/640] do_select+0x167/0x280
Jun 8 15:34:08 styx kernel: [sys_select+686/1184] sys_select+0x2ae/0x4a0
Jun 8 15:34:08 styx kernel: [syscall_call+7/11] syscall_call+0x7/0xb
Regards
Harri
On Tue, Jun 08, 2004 at 06:47:38PM +0200, Harald Dunkel wrote:
> I am not sure whether it is allowed to report problems with
> NVidia graphics support to this list. I just want to make
> sure that this hiccup isn't lost:
>
> Jun 8 15:34:08 styx kernel: Badness in pci_find_subsys at drivers/pci/search.c:167
Please search the archives, this is a well known nvidia driver problem.
Good luck,
greg k-h
On Mon, Jun 07, 2004 at 11:37:01PM +0200, Thomas Zehetbauer wrote:
> ACPI seems to be broken on Intel D865PERL board, I can only boot with
> acpi=off, otherwise the kernel either hangs after the "ACPI: Subsystem
> revision 20040326" message or shows a seemingly random oops message.
Is there any progress about this bug?
I hit it too, even sent a few mails about it, but no real fix yet :(
--
Balazs Pozsar