Skip to contentSkip to conversation

Linux 2.6.19-rc2

PATCH In review Linus Torvalds 32 participants 70 messages Oct 13, 2006 – Nov 3, 2006 3 messages not in the archive

Subsystems (9)
lore
Share & tools
lore
#0 lore raw

Ok, it's a week since -rc1, so -rc2 is out there.

The diff itself is pretty huge, mainly because of two things:

- we did the big interrupt handler argument passing cleanup, which affects a _lot_ of drivers in very trivial ways.

As a result, the diffstat shows a lot of files with a single line changed (mostly just removing the unused "pt_regs" argument), but the fallout also ended up being a lot of architectures cleaning up their irq handling.

- the initial ext4 code-drop (a lot of it is copyign the ext3 files and renaming functions and variables to "ext4_xyzzy" instead of "ext3_xyzzy")

Both of those changes were done _after_ the -rc1, because nobody wanted to handle the combination of large merges and a lot of fairly cosmetic changes. The irq thing turned out to be perhaps a bit more painful than expected (lots of arch fallout), but it all looks a lot better now.

Aside from those two things, we've got some arch updates (PowerPC, MIPS, SH), and various random fixes. Oh, and Al has been busy annotating things.

So if you had issues with -rc1 (which had a ton of merges - even more so than most -rc1 releases, I think), this should hopefully be a lot better.

Linus
---
Adrian Bunk (2):
HT_IRQ must depend on PCI
[DLM] Kconfig: don't show an empty DLM menu

Akinbou Mita (1):
[PKT_SCHED] sch_htb: use rb_first() cleanup

Al Viro (103):
m68k: dma_alloc_coherent() has gfp_t as the last argument m68k pt_regs fixes
minimal alpha pt_regs fixes
m32r pt_regs fixes
sparc32 pt_regs fixes
sparc64 pt_regs fixes
sparc32 rwlock fix
m68k pt_regs fixes, part 2
alpha pt_regs cleanups: device_interrupt
alpha pt_regs cleanups: handle_irq()
alpha pt_regs cleanups: machine_check()
alpha pt_regs cleanups: collapse set_irq_regs() in titan_dispatch_irqs() missed ia64 pt_regs fixes
misc arm pt_regs fixes
misc ppc pt_regs fixes
missing include in pdaudiocf_irq
missing include of scatterlist.h
missing forward declaration of pt_regs (asm-m68k/signal.h) linux/io.h needs types.h
uml pt_regs fixes
arm: it's OK to pass pointer to volatile as iounmap() argument... m68k/kernel/dma.c assumes !MMU_SUN3
sparc64 irq pt_regs fallout
fallout from alpha pt_regs patches
more ia64 irq handlers
extern doesn't make sense on a definition of function... trivial iomem annotations (arch/powerpc/platfroms/parsemi/pci.c) mv64630_pic NULL noise removal
wrong order of arguments in copy_to_user() in ncpfs
dlm gfp_t annotations
hppfs: readdir callback missed in prototype change
s390 traps.c __user annotations
mos7840 annotations
tifm __iomem annotations, NULL noise removal
[POWERPC] ARCH=ppc pt_regs fixes
advansys __iomem annotations
more fs/compat.c __user annotations
drivers/s390 misc sparse annotations
dccp __user annotations
__iomem annotations in sunzilog
NULL noise removal: advansys
fix misannotations in loop.c
misc sata __iomem annotations
hwdep_compat missed __user annotations
devio __user annotations
drivers/dma trivial annotations
tipc __user annotations
__user annotations: futex
ia64/hp NULL noise removal
ia64/sn __iomem annotations
mtd: remove several bogus casts to void * in iounmap() argument fix misannotation in ioc4.h
make kernel/relay.c __user-clean
passing pointer to setup_timer() should be via unsigned long acpi NULL noise removal
trivial iomem annotations: istallion
ptrace32 trivial __user annotations
gfp annotations: scsi_error
gfp annotations: radix_tree_root
trivial iomem annotations: sata_promise
openprom NULL noise removal
__user annotations: loop.c
em28xx NULL noise removal
fs/inode.c NULL noise removal
cpuset ANSI prototype
ptrdiff_t is %t, not %z
strndup() would better take size_t, not int net/sunrpc/auth_gss/svcauth_gss.c endianness regression missed const in prototype
use %zu for size_t
use %p for pointers
befs: remove bogus typedef
befs: prepare to sanitizing headers
befs: introduce on-disk endian types
befs: missing fs32_to_cpu() in debug.c
befs: endianness annotations
fs/fat endianness annotations
hpfs endianness annotations
smbfs endianness annotations
isofs endianness annotations
fs/partitions endianness annotations
ufs endianness annotations
endianness annotations in s2io
arm __user annotations
arm: use unsigned long instead of unsigned int in get_user() arm-versatile iomem annotations
m32r: C99 initializers in setup.c
m32r: signal __user annotations
m32r: NULL noise removal
m32r: more __user annotations
misuse of strstr
m68k uaccess __user annotations
misc m68k __user annotations
sun3 __iomem annotations
clean m68k ksyms
amiga_floppy_init() in non-modular case
z2_init() in non-modular case
remove bogus arch-specific syscall exports
alpha_ksyms.c cleanup
i2Output always takes kernel data now
fixing includes in alpha_ksyms.c
more kernel_execve() fallout (sbus)
uml shouldn't do HEADERS_CHECK

Alan Cox (2):
libata: Don't believe bogus claims in the older PIO mode register
ide-generic: jmicron fix

Alex Tomas (2):
ext3: add extent map support
ext4: 48bit physical block number support in extents

Alexandre Ratchov (2):
ext4: allow larger descriptor size
ext4: move block number hi bits

Alexey Dobriyan (8):
cdrom: add endianness annotations
serpent: fix endian warnings
chelsio: add endian annotations
Finish annotations of struct vlan_ethhdr
fs/*: use BUILD_BUG_ON
DAC960: use memmove for overlapping areas
lockdep: use BUILD_BUG_ON
md: use BUILD_BUG_ON

Amol Lad (1):
[ALSA] sound/pci/au88x0/au88x0.c: ioremap balanced with iounmap

Andi Kleen (6):
x86-64: Update defconfig
i386: Update defconfig
i386: Fix PCI BIOS config space access
x86: Terminate the kernel stacks for the unwinder
x86-64: Fix FPU corruption
x86-64: Annotate interrupt frame backlink in interrupt handlers

Andreas Mohr (1):
fs/bio.c: tweaks

Andrew Morton (15):
i386: irqs build fix
kauditd_thread warning fix
Fix WARN_ON / WARN_ON_ONCE regression
irq_reqs: export __irq_regs
x86_64 irq_regs fix
ibmveth irq fix
revert "nvidiafb: use generic ddc reading" ext4 uninline ext4_get_group_no_and_offset() ext4 64 bit divide fix
ext4: rename logic_sb_block
ext4 whitespace cleanups grow_buffers() infinite loop fix invalidate_inode_pages2_range() debug
dell_rbu: printk() warning fix [CIFS] cifs Kconfig: don't select CONNECTOR

Aneesh Kumar (2):
Fix typos in mm/shmem_acl.c
fix lockdep-design.txt

Aneesh Kumar K.V (1):
use struct irq_chip instead of struct hw_interrupt_type

Anton Blanchard (1):
[POWERPC] Update MTFSF_L() comment

Arnaud Patard (1):

      [ALSA] emu10k1: Fix outl() in snd_emu10k1_resume_regs()

Atsushi Nemoto (5):
[MIPS] ret_from_irq adjustment
[MIPS] Fix build errors related to wbflush.h on tx4927/tx4938. [MIPS] Make sure cpu_has_fpu is used only in atomic context [MIPS] <asm/irq.h> does not need pt_regs anymore. [MIPS] Optimize and cleanup get_saved_sp, set_saved_sp

Badari Pulavarty (1):

      ext4: 48bit i_file_acl

Benjamin Herrenschmidt (8):
[POWERPC] Fix zImage decompress location [POWERPC] Don't get PCI IRQ from OF for devices with no IRQ page fault retry with NOPAGE_REFAULT
[POWERPC] Make U4 PCIe work on maple
[POWERPC] Fix Maple secondary IDE interrupt [POWERPC] Update maple defconfig
[POWERPC] Fix i2c-powermac platform device usage [POWERPC] Fix windfarm platform device usage

Bill Nottingham (1):
Introduce vfs_listxattr

Brian King (1):
[POWERPC] Update pSeries defconfig for SATA

Chad Sellers (1):
SELinux: Bug fix in polidydb_destroy

Chen, Kenneth W (1):
hugetlb: fix linked list corruption in unmap_hugepage_range()

Christian Borntraeger (2):
[S390] ap bus poll thread priority.
[S390] stacktrace bug.

Christoph Lameter (2):
slab: remove wrongly placed BUG_ON
mm: kevent threads: use MPOL_DEFAULT

Cornelia Huck (5):
[S390] cio: 0 is a valid chpid.
[S390] cio: add missing KERN_INFO printk header. [S390] cio: Use ccw_dev_id and subchannel_id in ccw_device_private [S390] cio: Remove grace period for vary off chpid. [S390] cio: remove casts from/to (void *).

Dan Cyr (1):
[ALSA] hda-intel - New pci id for Nvidia MCP61

Dave Jones (2):
[HEADERS] Put linux/config.h out of its misery. move rmap BUG_ON outside DEBUG_VM

Dave Kleikamp (4):
ext4: initial copy of files from ext3
jbd2: initial copy of files from jbd
jbd2: cleanup ext4_jbd.h
Documentation/filesystems/ext4.txt

David Brownell (1):
ohci: don't play with IRQ regs

David Howells (7):
IRQ: Typedef the IRQ flow handler function type
IRQ: Typedef the IRQ handler function type
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers
IRQ: Use the new typedef for interrupt handler function pointers
ReiserFS: Make sure all dentries refs are released before calling kill_block_super()
AUTOFS: Make sure all dentries refs are released before calling kill_anon_super()
VFS: Destroy the dentries contributed by a superblock on unmounting

David S. Miller (5):
[SPARC64]: Update MAINTAINERS entry. [SPARC64]: Fix of_device bus_id settings. [SPARC64]: Update defconfig.
[SPARC32]: pcic.c needs asm/irq_regs.h [NET]: Do not memcmp() over pad bytes of struct flowi.

David Woodhouse (2):
Add CONFIG_HEADERS_CHECK option to automatically run 'make headers_check' Fix headers_check for O= builds; disable automatic check on UML.

Davide Libenzi (1):
epoll_pwait()

Deepak Saxena (1):
Update smc91x driver with ARM Versatile board info

Dmitry Mishin (2):
ext4: errors behaviour fix
ext3: errors behaviour fix

Eran Tromer (1):
libata: return sense data in HDIO_DRIVE_CMD ioctl

Eric Eric Sesterhenn (1):
reiserfs: null pointer dereferencing in reiserfs_read_bitmap_block

Eric Sesterhenn (3):
[ALSA] Fix memory leak in sound/isa/es18xx.c null dereference in fs/jbd/journal.c Remove unnecessary check in fs/fat/inode.c

Eric W. Biederman (5):
i386/x86_64: FIX pci_enable_irq to set dev->irq to the irq number i386/x86_64: Remove global IO_APIC_VECTOR
x86_64 irq: Allocate a vector across all cpus for genapic_flat. x86_64 irq: Scream but don't die if we receive an unexpected irq x86_64 irq: Properly update vector_irq

Florin Malita (2):
[ALSA] Dereference after free in snd_hwdep_release() fix Module taint flags listing in Oops/panic

Franck Bui-Huu (1):
Fix up mmap_kmem

Frederik Deweerdt (2):
fix qla{2,4} build error
ixp4xxdefconfig arm fixes

Geert Uytterhoeven (8):
m68k: syscall updates
m68k: more syscall updates
m68k/HP300: Enable HIL configuration options m68k/Atari: Interrupt updates
m68k/Apollo: Remove obsolete arch/m68k/apollo/dma.c m68k/MVME167: SERIAL167 is no longer broken m68k/MVME167: SERIAL167 tty flip buffer updates
m68knommu: sync syscalls with m68k

Geoff Levand (2):
[POWERPC] Minor fix for bootargs property [POWERPC] cell: fix default zImage build target

Haavard Skinnemoen (1):
IRQ: Fix AVR32 breakage

Heiko Carstens (4):
[S390] irq change build fixes.
sysrq: irq change build fix.
[S390] irq change improvements.
Disable DETECT_SOFTLOCKUP for s390

Helge Deller (1):
Fix section mismatch in de2104x.c

Henne (1):
sched: fix a kerneldoc error on is_init()

Ingo Molnar (2):
i386: fix rwsem build bug on CONFIG_M386=y
lockdep: fix printk recursion logic

Ishai Rabinovitz (2):
IB/srp: Remove redundant memset()
IB/srp: Enable multiple connections to the same target

Jack Morgenstein (1):
IB/mthca: Query port fix

James Bottomley (3):
[VOYAGER] fix genirq mess
[VOYAGER] fix up attribute packed specifiers in voyager.h [VOYAGER] fix up ptregs removal mess

James Morris (1):
IPsec: propagate security module errors up from flow_cache_lookup

Jamie Lenehan (1):
sh: Fix pr_debug statements for sh4

Jan Blunck (1):
Fix typo in "syntax error if percpu macros are incorrectly used" patch

Jan-Bernd Themann (2):
ehea: firmware (hvcall) interface changes
ehea: fix port state notification, default queue sizes

Jaroslav Kysela (1):

      [ALSA] version 1.0.13

Jeff Garzik (17):
[netdrvr] b44: handle excessive multicast groups arch/i386/kernel/time: don't shadow 'irq' function arg drivers/net: eliminate irq handler impossible checks, needless casts Various drivers' irq handlers: kill dead code, needless casts drivers/net/eepro: kill dead code
drivers/isdn/act2000: kill irq2card_map
irda: donauboe fixes, cleanups
firmware/dcdbas: fix bug in error cleanup
[libata] sata_promise: add PCI ID
tpm: fix error handling
x86/microcode: handle sysfs error
firmware/dell_rbu: handle sysfs errors
ipmi: handle sysfs errors
EISA: handle sysfs errors
firmware/efivars: handle error
drivers/mca: handle sysfs errors
ISDN: several minor fixes

Jens Axboe (4):
elevator: elevator_type member not used
splice: fix pipe_to_file() ->prepare_write() error path
ide-cd: fix breakage with internally queued commands
ide-cd: one more missing REQ_TYPE_CMD_ATA check

Jim Cromie (1):
MAINTAINERS: take over scx200-* and pc8736* drivers

Jiri Kosina (1):
make kernels with CONFIG_X86_GENERIC and !CONFIG_SMP compilable

Joerg Roedel (2):
[IPV6]: Seperate sit driver to extra module
[IPV6]: Seperate sit driver to extra module (addrconf.c changes)

Johann Lombardi (1):
jbd2: rename slab

Jon Mason (4):
x86-64: Calgary IOMMU: deobfuscate calgary_init
x86-64: Calgary IOMMU: Fix off by one when calculating register space location
x86-64: Calgary IOMMU: Update Jon's contact info
x86-64: Calgary IOMMU: print PCI bus numbers in hex

Karl-Johan Karlsson (1):
[MIPS] Show actual CPU information in /proc/cpuinfo

Karsten Keil (1):
bonding: fix deadlock on high loads in bond_alb_monitor()

Karsten Wiese (3):

      [ALSA] Fix bug in snd-usb-usx2y's usX2Y_pcms_lock_check()
      [ALSA] Repair snd-usb-usx2y for usb 2.6.18
      [ALSA] Handle file operations during snd_card disconnects using static file->f_op

Keith Owens (1):
Fix do_mbind warning with CONFIG_MIGRATION=n

Kyle McMartin (1):
[PARISC] Make firmware calls irqsafe-ish...

Laurent Vivier (1):
ext4: 64bit metadata

Linas Vepstas (20):
powerpc/cell spidernet ethtool -i version number info. powerpc/cell spidernet burst alignment patch. Spidernet module parm permissions powerpc/cell spidernet force-end fix powerpc/cell spidernet zlen min packet length powerpc/cell spidernet add missing netdev watchdog Spidernet fix register field definitions Spidernet stop queue when queue is full. powerpc/cell spidernet bogus rx interrupt bit powerpc/cell spidernet fix error interrupt print powerpc/cell spidernet stop error printing patch. powerpc/cell spidernet incorrect offset powerpc/cell spidernet low watermark patch. powerpc/cell spidernet NAPI polling info. powerpc/cell spidernet refine locking powerpc/cell spidernet
powerpc/cell spidernet reduce DMA kicking powerpc/cell spidernet variable name change powerpc/cell spidernet DMA direction fix powerpc/cell spidernet release all descrs

Linus Torvalds (7):
Initial blind fixup for arm for irq changes
ARM: fix up nested irq regs usage
Revert "[POWERPC] Don't get PCI IRQ from OF for devices with no IRQ" Fix extraneous '&' in recent NFS client cleanup
ACPI: Allow setting SCI_EN bit in PM1_CONTROL register Include proper header file for PFN_DOWN()
Linux 2.6.19-rc2

Luca Tettamanti (1):
Fix menuconfig build failure due to missing stdbool.h

Luke Zhang (1):

      [ALSA] WM9712 fixes for ac97_patch.c

Maciej W. Rozycki (2):
swarm: Actually initialize the IDE driver
32-bit compatibility HDIO IOCTLs

Mark Mason (1):
[MIPS] Fix compilation warnings in arch/mips/sibyte/bcm1480/smp.c

Martin Habets (4):
[SPARC32]: Fix prom.c build warning [SPARC32]: Mark srmmu_nocache_init as __init. [SPARC32]: Fix sparc32 modpost warnings with sunzilog [SPARC32]: Fix sparc32 modpost warnings.

Martin Schwidefsky (1):
[S390] Use CONFIG_GENERIC_TIME and define TOD clock source.

Matthew Wilcox (7):
Build fixes for struct pt_regs removal
[PARISC] Use set_irq_regs
[PA-RISC] Fix boot breakage
[PARISC] pdc_init no longer exists
[PARISC] More pt_regs removal
Use linux/io.h instead of asm/io.h
Consolidate check_signature

Matthias Urlichs (1):
document the core-dump-to-a-pipe patch

Maxime Bizon (1):
mv643xx_eth: Fix ethtool stats

Mel Gorman (2):
mm: use symbolic names instead of indices for zone initialisation
mm: remove memmap_zone_idx()

Melissa Howland (2):
[S390] monwriter buffer limit.
[S390] monwriter kzalloc size.

Michael Buesch (1):
b44: fix eeprom endianess issue

Michael Ellerman (2):
ibmveth: Harden driver initilisation
[POWERPC] Fix boot wrapper invocation if CROSS_COMPILE contains spaces

Michael S. Tsirkin (1):
IB/mthca: Fix off-by-one in mthca SRQ creation

Mike Frysinger (1):
include linux/types.h in linux/nbd.h

Miklos Szeredi (1):
[NET]: File descriptor loss while receiving SCM_RIGHTS

Mingming Cao (9):
ext4: rename ext4 symbols to avoid duplication of ext3 symbols
ext4: enable building of ext4
jbd2: rename jbd2 symbols to avoid duplication of jbd symbols
jbd2: enable building of jbd2 and have ext4 use it rather than jbd
ext4: switch fsblk to sector_t
jbd2: sector_t conversion
ext4: blk_type from sector_t to unsigned long long
ext4: removesector_t bits check
jbd2: switch blks_type from sector_t to ull

Monakhov Dmitriy (1):
D-cache aliasing issue in __block_prepare_write

Nathan Lynch (1):
[POWERPC] linux,tce-size property is 32 bits

NeilBrown (2):
md: fix bug where new drives added to an md array sometimes don't sync properly
knfsd: tidy up up meaning of 'buffer size' in nfsd/sunrpc

Nick Piggin (5):
[POWERPC] Fix harmless typo
mm: bug in set_page_dirty_buffers
mm: arch_free_page fix
mm: locks_freed fix
sched: likely profiling

Olaf Hering (4):
fix mesh compile errors after irq changes
[POWERPC] Fix up after irq changes
[POWERPC] SPU fixup after irq changes
[POWERPC] PReP fixup after irq changes

Olof Johansson (2):
powerpc: irq change build breaks
[POWERPC] Fix fsl_soc build breaks

Paolo 'Blaisorblade' Giarrusso (13):
uml: revert wrong patch
uml: correct removal of pte_mkexec
uml: readd forgot prototype
uml: make TT mode compile after setjmp-related changes
uml: make UML_SETJMP always safe
uml: fix processor selection to exclude unsupported processors and features
uml: fix uname under setarch i386
uml: declare in Kconfig our partial LOCKDEP support
uml: allow using again x86/x86_64 crypto code
uml: asm offsets duplication removal
uml: remove duplicate export
uml: deprecate CONFIG_MODE_TT
uml: allow finer tuning for host VMSPLIT setting

Patrick Caulfield (1):
[DLM] fix iovec length in recvmsg

Patrick McHardy (2):
[DECNET]: Fix sfuzz hanging on 2.6.18 [RTNETLINK]: Fix use of wrong skb in do_getlink()

Paul Mackerras (3):
[PPC] Fix some irq breakage with ARCH=ppc
[POWERPC] Fix xmon IRQ handler for pt_regs removal [POWERPC] Fix secondary CPU startup on old "powersurge" SMP powermacs

Paul Mundt (9):
sh: First step at generic timeofday support.
sh: Kill off timer_ops get_frequency().
sh: Updates for IRQ handler changes.
sh: Convert r7780rp IRQ handler to IRQ chip.
sh: Convert INTC2 IRQ handler to irq_chip.
sh: Convert IPR-IRQ to IRQ chip.
sh: Zero-out coherent buffer in consistent_alloc().
sh: Default enable R7780RP IRQs.
sh: interrupt exception handling rework

[email protected] (2):
NetLabel: fix a cache race condition
NetLabel: use SECINITSID_UNLABELED for a base SID

Pekka Enberg (2):
slab: reduce numa text size
um: irq changes break build

Peter Korsgaard (1):
pata-qdi: fix le32 in data_xfer

Peter Osterlund (1):
UDF: Fix mounting read-write

Peter Zijlstra (1):
forcedeth: hardirq lockdep warning

Petr Vandrovec (1):
Get core dump code to work...

Pierre Ossman (1):
mmc: multi sector write transfers

Rafael J. Wysocki (2):
swsusp: Make userland suspend work on SMP again
swsusp: Use suspend_console

Ralf Baechle (23):
[MIPS] Update Malta config.
[MIPS] Complete fixes after removal of pt_regs argument to int handlers. handle_sysrq lost its pt_regs * argument
[MIPS] DEC: pt_regs fixes for dec_intr_halt.
[MIPS] Jazz: Fix I/O port resources.
[MIPS] Jazz: Remove warning. After 7 years probably somebody test this ;) [MIPS] Jazz: build fix - include <linux/screen_info.h>
[MIPS] Jazz defconfig file.
[MIPS] Ocelot C: Build fix - ll_mv64340_irq takes no more regs argument. [MIPS] Fix return type of gt64120_irq.
[MIPS] DEC: pt_regs fixes for buserror handlers
[MIPS] Cleanup unnecessary <asm/ptrace.h> inclusions.
[MIPS] Fix RM9000 wait instruction detection.
[MIPS] qtronix: remove driver.
[MIPS] NUMA: Register all nodes before cpus or sysfs will barf. [RTC] Consistently use of tabs for formatting.
[MIPS] Malta: Fix build for non-MIPS32/64 configuration. [MIPS] Alchemy: nuke usbdev; it's useless as is ...
[MIPS] Workaround for bug in gcc -EB / -EL options.
[MIPS] Cleanup definitions of speed_t and tcflag_t.
[MIPS] BigSur: More useful defconfig.
[MIPS] IP27: Make declaration of setup_replication_mask a proper prototype. [MIPS] Pass NULL not 0 for pointer value.

Randy Dunlap (6):
x86-64: Fix compilation without CONFIG_KALLSYMS
ext4: clean up comments in ext4-extents patch
kernel-doc: fix function name in usercopy.c uaccess.h: match kernel-doc and function names
kernel-doc: drop various "inline" qualifiers
kernel-doc: make parameter description indentation uniform

Ravikiran Thirumalai (1):
Fix build breakage with CONFIG_X86_VSMP

Reinette Chatre (1):
bitmap: parse input from kernel and user buffers

Roland Dreier (2):
RDMA/amso1100: Fix build with debugging off
IPoIB: Check for DMA mapping error for TX packets

Roman Zippel (5):
provide tickadj define
m68k: cleanup string functions
m68k: fix typo in __generic_copy_to_user
m68k: small system.h cleanup
m68k: fix NBPG define

Russell Cattelan (2):
[GFS2] Fix a size calculation error [GFS2] Pass the correct value to kunmap_atomic

Ryusuke Sakato (1):
sh: SH-4A UBC support

Santiago Leon (4):
ibmveth: Add netpoll function
ibmveth: kdump interrupt fix
ibmveth: rename proc entry name
ibmveth: fix int rollover panic

Sasha Khapyorsky (1):
[ALSA] hda/patch_si3054: new codec vendor IDs

Scott Ashcroft (1):
[MIPS] Cobalt: Time runs too quickly

Sean Hefty (2):
IB/cm: Fix timewait crash after module unload IB/cm: Send DREP in response to unmatched DREQ

Stefan Richter (1):
ieee1394: nodemgr: fix startup of knodemgrd

Stephane Eranian (1):
Add carta_random32() library routine

Stephen Hemminger (8):
sky2: incorrect length on receive packets
skge: fix stuck irq when fiber down
skge: pause mapping for fiber
skge: better flow control negotiation
skge: version 1.9
sky2: revert pci express extensions
sky2: set lower pause threshold to prevent overrun thermal throttle: sysfs error checking

Stephen Rothwell (3):
[POWERPC] Update iseries_defconfig [POWERPC] Fix viocons for irq breakage [POWERPC] Fix iseries/smp.c for irq breakage

Steve French (26):
CIFS: Use SEEK_END instead of hardcoded value
[CIFS] Legacy time handling for Win9x and OS/2 part 1
[CIFS] Remove static and unused symbols
[CIFS] Fix build break
[CIFS] Remove unused prototypes
[CIFS] More removing of unused functions
[CIFS] Fix build break ifdef in wrong place
[CIFS] CIFS support for /proc/<pid>/mountstats part 1
[CIFS] Handle legacy servers which return undefined time zone [CIFS] Rename server time zone field
[CIFS] Fix typo in name of new cifs_show_stats
[CIFS] Do not send newer QFSInfo to legacy servers which can not support it [CIFS] Make use of newer QFSInfo dependent on capability bit instead of [CIFS] Allow LANMAN21 support even in both POSIX non-POSIX path [CIFS] Fix readdir of large directories for backlevel servers [CIFS] Allow for 15 minute TZs (e.g. Nepal) and be more explicit about [CIFS] Fix typo
[CIFS] Fix compiler warning with previous patch
[CIFS] readdir (ffirst) enablement of accurate timestamps from legacy servers [CIFS] Fix leaps year calculation for years after 2100
[CIFS] Do not need to adjust for Jan/Feb for leap day
[CIFS] Fix old DOS time conversion to handle timezone
[CIFS] fix typo in previous patch
[CIFS] Level 1 QPathInfo needed for proper OS2 support
[CIFS] Workaround incomplete byte length returned by some
[CIFS] Missing flags2 for DFS

Steven Whitehouse (3):
[GFS2] Fix uninitialised variable
[GFS2] Fix bug where lock not held
[GFS2] Update git tree name/location

Suparna Bhattacharya (1):
ext4: uninitialised extent handling

Timur Tabi (1):
[POWERPC] Add DTS for MPC8349E-mITX board

Tobin Davis (1):
[ALSA] Add new subdevice ids for hda-intel

Tom Tucker (1):
RDMA/amso1100: Add spinlocks to serialize ib_post_send/ib_post_recv

Tony Luck (1):
[IA64] Fix breakage from irq change

Trond Myklebust (3):
NFS: Fix typo in nfs_get_client()
NFS: Fix typo in nfs_get_client()
VM: Fix the gfp_mask in invalidate_complete_page2

Vasily Averin (1):
ext2: errors behaviour fix

Vasily Tarasov (3):
block layer: elevator_find function cleanup block layer: elv_iosched_show should get elv_list_lock block layer: ioprio_best function fix

Venkat Yekkirala (2):
IPsec: correct semantics for SELinux policy matching
IPsec: fix handling of errors for socket policies

Vlad Yasevich (2):
[SCTP]: Fix receive buffer accounting.
[SCTP]: Fix the RX queue size shown in /proc/net/sctp/assocs output.

Yoichi Yuasa (2):
[MIPS] Fix DECserial build error by IRQ hander change [MIPS] Fix timer setup for Jazz

YOSHIFUJI Hideaki (4):
[TCP]: Use TCPOLEN_TSTAMP_ALIGNED macro instead of magic number. [NET]: Use hton{l,s}() for non-initializers.
[NET]: Use typesafe inet_twsk() inline function instead of cast. [NET]: Introduce protocol-specific destructor for time-wait sockets.

Zach Brown (1):
64-bit jbd2 core

#1 lore raw
Quoted 1 line · Linus Torvalds

Ok, it's a week since -rc1, so -rc2 is out there.

Does anybody know what's up with the git server? Hopefully it's just my connection...

[alistair] 18:38 [~/linux-git] git pull
fatal: unexpected EOF
Fetch failure:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git

Signature · 5 lines
Cheers,
Alistair.

Final year Computer Science undergraduate.
1F2 55 South Clerk Street, Edinburgh, UK.
#2 lore raw
Quoted 9 lines · Alistair John Strachan
Quoted 1 line · Linus Torvalds

Ok, it's a week since -rc1, so -rc2 is out there.

Does anybody know what's up with the git server? Hopefully it's just my connection...

[alistair] 18:38 [~/linux-git] git pull
fatal: unexpected EOF
Fetch failure: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git

No, this is the result of a serious problem with gitweb.

We run gitweb behind a cache (otherwise it would be unacceptably expensive), but when httpd starts timing out on gitweb, it spawns gitweb over and over and over again, and the load on the machine skyrockets, throttling other services.

This happens every time we're on one server instead of two (one server is down right now for network rewiring.)

I think for now I'm just going to put a loadavg cap on running gitweb...

-hpa

#3 lore raw
Quoted 20 lines · H. Peter Anvin
Quoted 11 lines · Alistair John Strachan
Quoted 1 line · Linus Torvalds

Ok, it's a week since -rc1, so -rc2 is out there.

Does anybody know what's up with the git server? Hopefully it's just my connection...

[alistair] 18:38 [~/linux-git] git pull
fatal: unexpected EOF
Fetch failure:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git

No, this is the result of a serious problem with gitweb.

We run gitweb behind a cache (otherwise it would be unacceptably expensive), but when httpd starts timing out on gitweb, it spawns gitweb over and over and over again, and the load on the machine skyrockets, throttling other services.

Quoted 4 lines

This happens every time we're on one server instead of two (one server is down right now for network rewiring.)

I think for now I'm just going to put a loadavg cap on running gitweb...

I encountered a similar problem (to a far lower scale) when putting gitweb on my poor parisc server behind my ADSL line. The machine used to OOM several times a week. So I've set up an haproxy instance in front of it with a limit on the number of concurrent connections per backend. All excess connections are queued and served as soon as one slot frees up. The machine has never crashed since. Would you be interested in some help in trying to set it up ? Assuming that epoll is available, I have absolutely no worries about the load. As an added benefit, it could also provide HA and LB but I understand it's not the main concern right now.

Willy

#4 lore raw
In reply to Alistair John Strachan
Quoted 7 lines · On Fri, 13 Oct 2006 18:40:28 +0100 Alistair John Strachan…

Does anybody know what's up with the git server? Hopefully it's just my connection...

[alistair] 18:38 [~/linux-git] git pull
fatal: unexpected EOF
Fetch failure:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git

Same problem here.

Signature · 2 lines
	Paolo Ornati
	Linux 2.6.19-rc1-g9eb20074 on x86_64
#5 lore raw
In reply to Alistair John Strachan
Quoted 3 lines · Alistair John Strachan

Does anybody know what's up with the git server? Hopefully it's just my connection...

It's likely either
- a mirroring issue (the git.kernel.org server is _not_ the main one I push to, so it can take a while, and if some objects haven't made it, the upload side will exit because the repository isn't "valid" until the full mirror has happened)
- too many clients trying at the same time

In this case, I suspect the latter.

Linus

#6 lore raw
In reply to Linus Torvalds

Hi,

Quoted 2 lines · Linus Torvalds

Ok, it's a week since -rc1, so -rc2 is out there.

Please consider applying this patches.

Regards,
Michal

Signature · 9 lines
Michal K. K. Piotrowski
LTG - Linux Testers Group
(http://www.stardust.webpages.pl/ltg/)


----
[PATCH] Fix 'headers_install' with separate output directory

Signed-off-by: David Woodhouse <[email protected]>
Makefile +1 -1
index 80dac02..bdf7c18 100644
@@ -932,7 +932,7 @@ headers_install_all: include/linux/versi
 
 PHONY += headers_install
 headers_install: include/linux/version.h scripts_basic FORCE
-	@if [ ! -r include/asm-$(ARCH)/Kbuild ]; then \
+	@if [ ! -r $(srctree)/include/asm-$(ARCH)/Kbuild ]; then \
 	  echo '*** Error: Headers not exportable for this architecture ($(ARCH))'; \
 	  exit 1 ; fi
 	$(Q)$(MAKE) $(build)=scripts scripts/unifdef

---

scripts/Makefile.headersinst +1 -1
index cac8f21..6a7b740 100644
@@ -168,7 +168,7 @@ ifdef GENASM
 	$(call cmd,gen)
 
 else
-$(objhdr-y) :		$(INSTALL_HDR_PATH)/$(_dst)/%.h: $(srctree)/$(obj)/%.h $(KBUILDFILES)
+$(objhdr-y) :		$(INSTALL_HDR_PATH)/$(_dst)/%.h: $(objtree)/$(obj)/%.h $(KBUILDFILES)
 	$(call cmd,o_hdr_install)
 
 $(header-y) :		$(INSTALL_HDR_PATH)/$(_dst)/%.h: $(srctree)/$(obj)/%.h $(KBUILDFILES)
#8 lore raw
In reply to Linus Torvalds
Quoted 3 lines · Linus Torvalds

So if you had issues with -rc1 (which had a ton of merges - even more so than most -rc1 releases, I think), this should hopefully be a lot better.

airo driver suspend is still broken. i still need to apply the patch posted originally at (with an explanation of why airo suspend is not working anymore)
http://marc.theaimsgroup.com/?l=linux-kernel&m=116025080215854&w=2 to get the driver to suspend. (the patch still applies to 2.6.19-rc2 with minor fuzz).

--alex--

Signature · 4 lines
| I believe the moment is at hand when, by a paranoiac and active |
|  advance of the mind, it will be possible (simultaneously with  |
|  automatism and other passive states) to systematize confusion  |
|  and thus to help to discredit completely the world of reality. |
#9 lore raw

cover 2.6.19-rc2: known regressions

In reply to Linus Torvalds

As usual, we are swamped with bug reports for regressions after -rc1.

For an easier reading (and hoping linux-kernel might not eat the emails), I've splitted the list of known regressions in three emails:
[1/3] known unfixed regressions
[2/3] knwon regressions with workarounds
[3/3] known regressions with patches

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#10 lore raw

PATCH 1/3 2.6.19-rc2: known unfixed regressions

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18.

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

Subject : CONFIG_X86_VOYAGER=y, CONFIG_SMP=n compile error
References : http://lkml.org/lkml/2006/10/7/51
Submitter : Jesper Juhl <[email protected]>

             commit 7d12e780e003f93433d49ce78cfedf4b4c52adc5

Status : unknown

Subject : CONFIG_X86_VISWS=y, CONFIG_SMP=n compile error
References : http://lkml.org/lkml/2006/10/7/51
Submitter : Jesper Juhl <[email protected]>

             commit 7d12e780e003f93433d49ce78cfedf4b4c52adc5

Status : unknown

Subject : unable to rip cd
References : http://lkml.org/lkml/2006/10/13/100
Submitter : Alex Romosan <[email protected]>
Status : unknown

Subject : sata-via doesn't detect anymore disks attached to VIA vt6421
References : http://bugzilla.kernel.org/show_bug.cgi?id=7255
Submitter : Thierry Vignaud <[email protected]>
Status : unknown

Subject : monitor not active after boot
References : http://lkml.org/lkml/2006/10/5/338
Submitter : Olaf Hering <[email protected]>

             commit 346bc21026e7a92e1d7a4a1b3792c5e8b686133d

Status : unknown

Subject : SMP x86_64 boot problem
References : http://lkml.org/lkml/2006/9/28/330

             http://lkml.org/lkml/2006/10/5/289

Submitter : [email protected]
Status : submitter was asked to git bisect
result of bisecting seems to be wrong

Subject : do_IRQ: No irq handler for vector
References : http://lkml.org/lkml/2006/10/11/13
Submitter : Robert Hancock <[email protected]>

Status : Andrew: a few people are seeing this. Eric is working it.

Subject : messed up keyboard events, TSC related
References : http://bugzilla.kernel.org/show_bug.cgi?id=7291
Submitter : David Gerber <[email protected]>

John Stultz <[email protected]>
Status : Dmitry and John are investigating

Subject : ide-generic no longer finds marvell controller
References : http://bugzilla.kernel.org/show_bug.cgi?id=7353
Submitter : Kenny Graunke <[email protected]>

             commit a4bea10eca68152e84ffc4eaeb9d20ec2ac34664

Status : Alan is investigating

Subject : cpufreq not working on AMD K8
References : http://lkml.org/lkml/2006/10/10/114
Submitter : Christian <[email protected]>

Status : Mark is investigating

#11 lore raw

Re: [1/3] 2.6.19-rc2: known unfixed regressions

Quoted 12 lines · Adrian Bunk

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18.

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

Quoted 5 lines

Subject : do_IRQ: No irq handler for vector
References : http://lkml.org/lkml/2006/10/11/13
Submitter : Robert Hancock <[email protected]>

Status : Andrew: a few people are seeing this. Eric is working it.

Please see commit: 994bd4f9f5a065ead4a92435fdd928ac7fd33809 That part should fine in -rc2

YH found a corner case I missed (in my bug fix refactoring) and submitted a patch earlier today. But that only shows up when we resubmit an irq which is almost never.

Eric

#12 lore raw

PATCH 2/3 2.6.19-rc2: knwon regressions with workarounds

In reply to Adrian Bunk

This email lists some known regressions with workarounds in 2.6.19-rc2 compared to 2.6.18.

Unless proper solutions become available, we should implement the workarounds.

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Please trim the Cc when answering.

Subject : snd-hda-intel <-> forcedeth MSI problem
References : http://lkml.org/lkml/2006/10/5/40

             http://lkml.org/lkml/2006/10/7/164

Submitter : Prakash Punnoor <[email protected]>
Status : suggested workaround for 2.6.19:
deactivate MSI in snd-intel-hda by default

Subject : DVB frontend selection causes compile errors
References : http://lkml.org/lkml/2006/10/8/244
Submitter : Adrian Bunk <[email protected]>

             commit 176ac9da4f09820a43fd48f0e74b1486fc3603ba

"Andrew de Quincey" <[email protected]>
Status : possible workaround: don't allow CONFIG_DVB_FE_CUSTOMISE=y

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

#13 lore raw

PATCH 3/3 2.6.19-r2: known regressions with patches

In reply to Adrian Bunk

This email lists some known regressions in 2.6.19-rc2 compared to 2.6.18 with patches available.

Unless there are specific reasons against doing so, we should try to get these patches into -rc3.

(Especially the lad who wrote the patch for the third entry should ush
his patch...)

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

Subject : spufs: fix a memory leak
References : http://lkml.org/lkml/2006/10/10/63
Submitter : "Noguchi, Masato" <Masato.Noguchi at jp.sony.com> Caused-By : Arnd Bergmann <arnd.bergmann at de.ibm.com>

             commit e1dbff2bafa83f839ef15f51904b0cce9fc89387

Handled-By : "Noguchi, Masato" <Masato.Noguchi at jp.sony.com>
Patch : http://lkml.org/lkml/2006/10/10/63
Status : patch available

Subject : w83781d modprobing failure
References : http://bugzilla.kernel.org/show_bug.cgi?idr93
Submitter : Sylvain BERTRAND <sylvain.bertrand at gmail.com> Caused-By : David Hubbard <david.c.hubbard at gmail.com> (?)

             commit 8202632647278eba7223727dc442f49227c040d0 (?)

Handled-By : Jean Delvare <khali at linux-fr.org>
Patch : http://bugzilla.kernel.org/show_bug.cgi?idr93
Status : patch available

Subject : drivers/char/specialix.c: broken baud conversion
References : http://lkml.org/lkml/2006/10/8/214
Submitter : Adrian Bunk <bunk at stusta.de> Caused-By : Alan Cox <alan at lxorguk.ukuu.org.uk>

             commit 67cc0161ecc9ebee6eba4af6cbfdba028090b1b9

Handled-By : Adrian Bunk <bunk at stusta.de>
Patch : http://lkml.org/lkml/2006/10/11/9
Status : patch available

Subject : ueagle-atm Oops
References : http://lkml.org/lkml/2006/10/6/390
Submitter : Ernst Herzberg <list-lkml at net4u.de> Caused-By : Greg Kroah-Hartman <gregkh at suse.de>

             commit e7ccdfec087f02930c5cdc81143d4a045ae8d361

Handled-By : Matthieu Castet <castet.matthieu at free.fr>
Laurent Riffard <laurent.riffard at free.fr>
Patch : http://lkml.org/lkml/2006/10/11/209
Status : patch available

Subject : oops in xfrm_register_mode
References : http://lkml.org/lkml/2006/10/4/170
Submitter : Steve Fox <drfickle at us.ibm.com>
Handled-By : Vivek Goyal <vgoyal at in.ibm.com>
Status : patch available

Subject : airo suspend fails
References : http://lkml.org/lkml/2006/10/6/3
Submitter : Alex Romosan <romosan at sycorax.lbl.gov> Caused-By : Sukadev Bhattiprolu <sukadev at us.ibm.com> (?)

             commit 3b4c7d640376dbccfe80fc4f7b8772ecc7de28c5  (?)

Handled-By : Dave Kleikamp <shaggy at austin.ibm.com>
Patch : http://lkml.org/lkml/2006/10/7/139
Status : patch available

Subject : kexec broken on x86_64
References : http://lkml.org/lkml/2006/10/5/86
Submitter : Magnus Damm <magnus.damm at gmail.com> Handled-By : Vivek Goyal <vgoyal at in.ibm.com>
Patch : http://lkml.org/lkml/2006/10/5/199
Status : patch available

Subject : NFSv4 fails to mount (timeout)
References : http://bugzilla.kernel.org/show_bug.cgi?idr74
Submitter : Torsten Kaiser <kernel at bardioc.dyndns.org> Caused-By : Trond Myklebust <Trond.Myklebust at netapp.com>

             commit 51b6ded4d9a94a61035deba1d8f51a54e3a3dd86

Handled-By : Trond Myklebust <Trond.Myklebust at netapp.com>
Patch : http://bugzilla.kernel.org/show_bug.cgi?idr74
Status : patch available

#14 lore raw

Re: [3/3] 2.6.19-r2: known regressions with patches

Hi Adrian,

Quoted 17 lines · Adrian Bunk

This email lists some known regressions in 2.6.19-rc2 compared to 2.6.18 with patches available.

Unless there are specific reasons against doing so, we should try to get these patches into -rc3.

(Especially the lad who wrote the patch for the third entry should ush
his patch...)

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering. (...)

Quoted 8 lines

Subject : w83781d modprobing failure
References : http://bugzilla.kernel.org/show_bug.cgi?idr93
Submitter : Sylvain BERTRAND <sylvain.bertrand at gmail.com> Caused-By : David Hubbard <david.c.hubbard at gmail.com> (?)

             commit 8202632647278eba7223727dc442f49227c040d0 (?)

Handled-By : Jean Delvare <khali at linux-fr.org>
Patch : http://bugzilla.kernel.org/show_bug.cgi?idr93
Status : patch available

Patch was tested successfully by Sylvain, I sent it to Greg already, who should push it to Linus soon (with 7 other hwmon patches.)

Thanks,

Signature · 1 line
Jean Delvare
#15 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

In reply to Adrian Bunk
Quoted 7 lines · Adrian Bunk

As usual, we are swamped with bug reports for regressions after -rc1.

For an easier reading (and hoping linux-kernel might not eat the emails), I've splitted the list of known regressions in three emails:
[1/3] known unfixed regressions
[2/3] knwon regressions with workarounds
[3/3] known regressions with patches

There's a raft of ARM regressions as well (see
http://armlinux.simtec.co.uk/kautobuild/2.6.19-rc2/index.html), mostly related to the IRQ changes, as well as this error:

sysctl_net.c:(.text+0x64a8c): undefined reference to `highest_possible_node_id'

Signature · 3 lines
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 Serial core
#16 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 9 lines · Russell King
Quoted 7 lines · Adrian Bunk

As usual, we are swamped with bug reports for regressions after -rc1.

For an easier reading (and hoping linux-kernel might not eat the emails), I've splitted the list of known regressions in three emails:
[1/3] known unfixed regressions
[2/3] knwon regressions with workarounds
[3/3] known regressions with patches

Quoted 3 lines

There's a raft of ARM regressions as well (see
http://armlinux.simtec.co.uk/kautobuild/2.6.19-rc2/index.html), mostly related to the IRQ changes, as well as this error:

Thanks, I'll look at them before preparing the next version of my regressions list.

Quoted 1 line

sysctl_net.c:(.text+0x64a8c): undefined reference to `highest_possible_node_id'

This problem already got an entry a few hours ago:

Subject : undefined reference to highest_possible_node_id
References : http://lkml.org/lkml/2006/9/4/233

             http://lkml.org/lkml/2006/10/15/11

Submitter : Olaf Hering <[email protected]>

             commit 0f532f3861d2c4e5aa7dcd33fb18e9975eb28457

Status : unknown

Quoted 1 line

Russell King

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#17 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 21 lines · Adrian Bunk
Quoted 9 lines · Russell King
Quoted 7 lines · Adrian Bunk

As usual, we are swamped with bug reports for regressions after -rc1.

For an easier reading (and hoping linux-kernel might not eat the emails), I've splitted the list of known regressions in three emails:
[1/3] known unfixed regressions
[2/3] knwon regressions with workarounds
[3/3] known regressions with patches

Quoted 3 lines

There's a raft of ARM regressions as well (see
http://armlinux.simtec.co.uk/kautobuild/2.6.19-rc2/index.html), mostly related to the IRQ changes, as well as this error:

Thanks, I'll look at them before preparing the next version of my regressions list.

Quoted 1 line

sysctl_net.c:(.text+0x64a8c): undefined reference to `highest_possible_node_id'

This problem already got an entry a few hours ago:

Quoted 7 lines

Subject : undefined reference to highest_possible_node_id
References : http://lkml.org/lkml/2006/9/4/233

             http://lkml.org/lkml/2006/10/15/11

Submitter : Olaf Hering <[email protected]>

             commit 0f532f3861d2c4e5aa7dcd33fb18e9975eb28457

Status : unknown

Looking at this commit and the mails, it was known on the 4th September that this patch caused build errors while this change was in -mm, yet it still found its way into mainline on 2nd October.

Is anyone going to look at fixing this problem, or should we be asking for the commit to be reverted?

Signature · 3 lines
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 Serial core
#18 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 30 lines · Russell King
Quoted 21 lines · Adrian Bunk
Quoted 9 lines · Russell King
Quoted 7 lines · Adrian Bunk

As usual, we are swamped with bug reports for regressions after -rc1.

For an easier reading (and hoping linux-kernel might not eat the emails), I've splitted the list of known regressions in three emails:
[1/3] known unfixed regressions
[2/3] knwon regressions with workarounds
[3/3] known regressions with patches

Quoted 3 lines

There's a raft of ARM regressions as well (see
http://armlinux.simtec.co.uk/kautobuild/2.6.19-rc2/index.html), mostly related to the IRQ changes, as well as this error:

Thanks, I'll look at them before preparing the next version of my regressions list.

Quoted 1 line

sysctl_net.c:(.text+0x64a8c): undefined reference to `highest_possible_node_id'

This problem already got an entry a few hours ago:

Quoted 7 lines

Subject : undefined reference to highest_possible_node_id
References : http://lkml.org/lkml/2006/9/4/233

             http://lkml.org/lkml/2006/10/15/11

Submitter : Olaf Hering <[email protected]>

             commit 0f532f3861d2c4e5aa7dcd33fb18e9975eb28457

Status : unknown

Quoted 6 lines

Looking at this commit and the mails, it was known on the 4th September that this patch caused build errors while this change was in -mm, yet it still found its way into mainline on 2nd October.

Is anyone going to look at fixing this problem, or should we be asking for the commit to be reverted?

Since everyone seems intent at ignoring this issue, here's a patch to try to solve it.

Probably will suffer from occasionally not being included in the kernel, and therefore the function not exported for modules, but at least it's a _lot_ better than the current utterly broken situation.

lib/Makefile +1 -1
index cf98fab..8845514 100644
@@ -5,7 +5,7 @@ #
 lib-y := ctype.o string.o vsprintf.o cmdline.o \
 	 bust_spinlocks.o rbtree.o radix-tree.o dump_stack.o \
 	 idr.o div64.o int_sqrt.o bitmap.o extable.o prio_tree.o \
-	 sha1.o irq_regs.o
+	 sha1.o irq_regs.o nodemask.o
 
 lib-$(CONFIG_MMU) += ioremap.o
 lib-$(CONFIG_SMP) += cpumask.o
lib/cpumask.c +0 -16
index 7a2a73f..3a67dc5 100644
@@ -43,19 +43,3 @@ int __any_online_cpu(const cpumask_t *ma
 	return cpu;
 }
 EXPORT_SYMBOL(__any_online_cpu);
-
-#if MAX_NUMNODES > 1
-/*
- * Find the highest possible node id.
- */
-int highest_possible_node_id(void)
-{
-	unsigned int node;
-	unsigned int highest = 0;
-
-	for_each_node_mask(node, node_possible_map)
-		highest = node;
-	return highest;
-}
-EXPORT_SYMBOL(highest_possible_node_id);
-#endif
lib/nodemask.c new file +19 -0
new file mode 100644
index 0000000..6c49eb5
@@ -0,0 +1,19 @@
+#include <linux/kernel.h>
+#include <linux/cpumask.h>
+#include <linux/module.h>
+
+#if MAX_NUMNODES > 1
+/*
+ * Find the highest possible node id.
+ */
+int highest_possible_node_id(void)
+{
+	unsigned int node;
+	unsigned int highest = 0;
+
+	for_each_node_mask(node, node_possible_map)
+		highest = node;
+	return highest;
+}
+EXPORT_SYMBOL(highest_possible_node_id);
+#endif
Signature · 3 lines
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 Serial core
#19 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 12 lines · On Fri, 20 Oct 2006 19:07:22 +0100 Russell King <[email protected]>…
Quoted 12 lines
Quoted 7 lines

Subject : undefined reference to highest_possible_node_id
References : http://lkml.org/lkml/2006/9/4/233

             http://lkml.org/lkml/2006/10/15/11

Submitter : Olaf Hering <[email protected]>

             commit 0f532f3861d2c4e5aa7dcd33fb18e9975eb28457

Status : unknown

Looking at this commit and the mails, it was known on the 4th September that this patch caused build errors while this change was in -mm, yet it still found its way into mainline on 2nd October.

Quoted 5 lines
Quoted 2 lines

Is anyone going to look at fixing this problem, or should we be asking for the commit to be reverted?

Since everyone seems intent at ignoring this issue, here's a patch to try to solve it.

I sent the below to Linus yesterday...

From: Andrew Morton <[email protected]>

Qooting Adrian:

- net/sunrpc/svc.c uses highest_possible_node_id()

- include/linux/nodemask.h says highest_possible_node_id() is out-of-line #if MAX_NUMNODES > 1

- the out-of-line highest_possible_node_id() is in lib/cpumask.c

- lib/Makefile: lib-$(CONFIG_SMP) += cpumask.o CONFIG_ARCH_DISCONTIGMEM_ENABLE=y, CONFIG_SMP=n, CONFIG_SUNRPC=y

-> highest_possible_node_id() is used in net/sunrpc/svc.c
CONFIG_NODES_SHIFT defined and > 0

-> include/linux/numa.h: MAX_NUMNODES > 1

-> compile error

The bug is not present on architectures where ARCH_DISCONTIGMEM_ENABLE depends on NUMA (but m32r isn't the only affected architecture).

So move the function into page_alloc.c

Cc Adrian Bunk <[email protected]>
Cc Paul Jackson <[email protected]>

---

 lib/cpumask.c   |   16 ----------------
 mm/page_alloc.c |   16 ++++++++++++++++
 2 files changed, 16 insertions(+), 16 deletions(-)
lib/cpumask.c renamed from lib/cpumask.c~highest_possible_node_id-linkage-fix +0 -16
diff -puN lib/cpumask.c~highest_possible_node_id-linkage-fix lib/cpumask.c
@@ -43,19 +43,3 @@ int __any_online_cpu(const cpumask_t *ma
 	return cpu;
 }
 EXPORT_SYMBOL(__any_online_cpu);
-
-#if MAX_NUMNODES > 1
-/*
- * Find the highest possible node id.
- */
-int highest_possible_node_id(void)
-{
-	unsigned int node;
-	unsigned int highest = 0;
-
-	for_each_node_mask(node, node_possible_map)
-		highest = node;
-	return highest;
-}
-EXPORT_SYMBOL(highest_possible_node_id);
-#endif
mm/page_alloc.c renamed from mm/page_alloc.c~highest_possible_node_id-linkage-fix +16 -0
diff -puN mm/page_alloc.c~highest_possible_node_id-linkage-fix mm/page_alloc.c
@@ -3120,3 +3120,19 @@ unsigned long page_to_pfn(struct page *p
 EXPORT_SYMBOL(pfn_to_page);
 EXPORT_SYMBOL(page_to_pfn);
 #endif /* CONFIG_OUT_OF_LINE_PFN_TO_PAGE */
+
+#if MAX_NUMNODES > 1
+/*
+ * Find the highest possible node id.
+ */
+int highest_possible_node_id(void)
+{
+	unsigned int node;
+	unsigned int highest = 0;
+
+	for_each_node_mask(node, node_possible_map)
+		highest = node;
+	return highest;
+}
+EXPORT_SYMBOL(highest_possible_node_id);
+#endif

_

#20 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 15 lines · Andrew Morton
Quoted 12 lines · On Fri, 20 Oct 2006 19:07:22 +0100 Russell King <[email protected]>…
Quoted 12 lines
Quoted 7 lines

Subject : undefined reference to highest_possible_node_id
References : http://lkml.org/lkml/2006/9/4/233

             http://lkml.org/lkml/2006/10/15/11

Submitter : Olaf Hering <[email protected]>

             commit 0f532f3861d2c4e5aa7dcd33fb18e9975eb28457

Status : unknown

Looking at this commit and the mails, it was known on the 4th September that this patch caused build errors while this change was in -mm, yet it still found its way into mainline on 2nd October.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

Is anyone going to look at fixing this problem, or should we be asking for the commit to be reverted?

Since everyone seems intent at ignoring this issue, here's a patch to try to solve it.

I sent the below to Linus yesterday...

Ah, okay. Must not have poped out of the other side of Linus by 6am GMT then. (We also seem to have non-working git snapshots again, so when I looked at the ARM kautobuild it showed the same old errors.)

Signature · 3 lines
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 Serial core
#21 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 3 lines · Russell King

Ah, okay. Must not have poped out of the other side of Linus by 6am GMT then.

Yeah, I applied Andrew's patch-bomb just an hour or two ago.

Quoted 2 lines

(We also seem to have non-working git snapshots again, so when I looked at the ARM kautobuild it showed the same old errors.)

Gaah. Remind me where the autobuild is again..

Linus

#22 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 5 lines · Linus Torvalds
Quoted 2 lines · Russell King

Ah, okay. Must not have poped out of the other side of Linus by 6am GMT then.

Yeah, I applied Andrew's patch-bomb just an hour or two ago.

That explains why I hadn't seen it, and probably explains why there hadn't been any git snapshots since Thursday morning (my time) - so nothings actually broken. Ignore me! 8)

Quoted 4 lines
Quoted 2 lines

(We also seem to have non-working git snapshots again, so when I looked at the ARM kautobuild it showed the same old errors.)

Gaah. Remind me where the autobuild is again..

The main status page is at:
http://armlinux.simtec.co.uk/kautobuild/

You could argue that it should be able to run off the git tree itself - I'll probably even agree with you, but kautobuild isn't my system.

Signature · 3 lines
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 Serial core
#23 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 5 lines · Russell King
Quoted 2 lines

Gaah. Remind me where the autobuild is again..

The main status page is at:
http://armlinux.simtec.co.uk/kautobuild/

Ahh. At least _most_ builds seem ok. I only looked at the first failing one, and it _seems_ to be due to "drivers/usb/gadget/pxa2xx_udc.c" which still includes <asm/irq.h> rather than <linux/irq.h>.

Maybe. I could do the trivial fix myself, but there have been people who have ARM build environments, so maybe somebody who can actually verify whether that one-liner fixes things (or whether there is something else hiding) can do that and send me the patch.

[ Although I've gotten so much email lately that maybe I already missed
that patch. Ahh, the joys of SCM discussions ;) ]

Linus

#24 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

In reply to Russell King
Quoted 3 lines · Russell King

Since everyone seems intent at ignoring this issue, here's a patch to try to solve it.

Heh. I just merged another one (through Andrew), so it _should_ be fixed in my tree already. But thanks,

Linus

#25 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

In reply to Adrian Bunk

Hi

I did search the archives, but it does seem to be the new one. r8169 network driver introduced in 2.6.19-rcX a set_mac_address function, which doesn't work for me. It should resolve the bugreport
http://bugzilla.kernel.org/show_bug.cgi?id=6032 but, as you see from the last comment from the original reporter and from my following comment, it doesn't seem to. I think, it should either be fixed or reverted. My test-system, was a ppc NAS (KuroboxHG):

0000:00:0b.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8169 Gigabit Ethernet (rev 10)
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr- Stepping- SERR- FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 128 (8000ns min, 16000ns max), Cache Line Size: 0x08 (32 bytes) Interrupt: pin A routed to IRQ 16
Region 0: I/O ports at bfff00 [size=256]
Region 1: Memory at bfffff00 (32-bit, non-prefetchable) [size=256]
Capabilities: [dc] Power Management version 0
Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-

Thanks
Guennadi
---
Guennadi Liakhovetski

#26 lore raw

Re: [0/3] 2.6.19-rc2: known regressions

Quoted 8 lines · Guennadi Liakhovetski

I did search the archives, but it does seem to be the new one. r8169 network driver introduced in 2.6.19-rcX a set_mac_address function, which doesn't work for me. It should resolve the bugreport
http://bugzilla.kernel.org/show_bug.cgi?id=6032 but, as you see from the last comment from the original reporter and from my following comment, it doesn't seem to. I think, it should either be fixed or reverted. My test-system, was a ppc NAS (KuroboxHG):

Can you please test the things that Francois asks you to test in the last comment?

That said, it does appear that the patch breaks things for some people, and the upsides seem very limited - only relevant when somebody tries to change the MAC address, which is not a very normal thing to do anyway.

So maybe reverting it is the right thing to do. Guennadi, can you confirm that it is commit a2b98a69 ("r8169: mac address change support") that breaks it, and that reverting just that one commit fixes things for you?

But please check the things that are suggested in the bugzilla entry first.

Francois? Jeff?

Linus

#27 lore raw

r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Linus Torvalds <[email protected]> :
[regression related to r8169 MAC address change]

Quoted 1 line

Francois ? Jeff ?

Go revert it.

Despite what I claimed, I can not find a third-party confirmation by email that it works elsewhere.

It would probably be enough to remove the call to __rtl8169_set_mac_addr() in rtl8169_hw_start() though.

In place of the test suggested in bugzilla, I'd rather see Guennadi test the thing below (acked on netdev by the initial victim if someone wonders why it has not changed the status of bugzilla so far):

Quoted 1 line

From f092d907f78e81846dfaf1008a6409c3c4b58f27 Mon Sep 17 00:00:00 2001

From: Francois Romieu <[email protected]>
Date: Sat, 21 Oct 2006 21:07:36 +0200
Subject: [PATCH] r8169: perform a PHY reset before any other operation at boot time

Realtek's 8139/810x (0x8136) PCI-E comes with a touchy PHY. A big heavy reset seems to calm it down.

Fix for http://bugzilla.kernel.org/show_bug.cgi?id=7378.

---
 drivers/net/r8169.c |   18 ++++++++++++++++++
 1 files changed, 18 insertions(+), 0 deletions(-)
drivers/net/r8169.c +18 -0
index f1c7575..927fc17 100644
@@ -1440,6 +1440,22 @@ static void rtl8169_release_board(struct
 	free_netdev(dev);
 }
 
+static void rtl8169_phy_reset(struct net_device *dev,
+			      struct rtl8169_private *tp)
+{
+	void __iomem *ioaddr = tp->mmio_addr;
+	int i;
+
+	tp->phy_reset_enable(ioaddr);
+	for (i = 0; i < 100; i++) {
+		if (!tp->phy_reset_pending(ioaddr))
+			return;
+		msleep(1);
+	}
+	if (netif_msg_link(tp))
+		printk(KERN_ERR "%s: PHY reset failed.\n", dev->name);
+}
+
 static void rtl8169_init_phy(struct net_device *dev, struct rtl8169_private *tp)
 {
 	void __iomem *ioaddr = tp->mmio_addr;
@@ -1468,6 +1484,8 @@ static void rtl8169_init_phy(struct net_
 
 	rtl8169_link_option(board_idx, &autoneg, &speed, &duplex);
 
+	rtl8169_phy_reset(dev, tp);
+
 	rtl8169_set_speed(dev, autoneg, speed, duplex);
 
 	if ((RTL_R8(PHYstatus) & TBI_Enable) && netif_msg_link(tp))
Signature · 1 line
1.4.2.4
#28 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Quoted 9 lines · Francois Romieu

Linus Torvalds <[email protected]> :
[regression related to r8169 MAC address change]

Quoted 1 line

Francois ? Jeff ?

Go revert it.

Despite what I claimed, I can not find a third-party confirmation by email that it works elsewhere.

Quoted 6 lines

It would probably be enough to remove the call to __rtl8169_set_mac_addr() in rtl8169_hw_start() though.

In place of the test suggested in bugzilla, I'd rather see Guennadi test the thing below (acked on netdev by the initial victim if someone wonders why it has not changed the status of bugzilla so far):

AFAIU, you wanted it applied on the top of the "non-working" kernel (2.6.19-rc2-ish)? No, it didn't work. And, worse yet, I think, it is after testing that patch that the interface got into a state, when netconsole worked, ping worked, but ssh didn't. A poweroff was needed to recover. In case you still need it, here's the info you requested:

00:0b.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8169 Gigabit Ethernet (rev 10)
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr- Stepping- SERR- FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 128 (8000ns min, 16000ns max), Cache Line Size: 32 bytes
Interrupt: pin A routed to IRQ 16
Region 0: I/O ports at febfff00 [size=256]
Region 1: Memory at bffffc00 (32-bit, non-prefetchable) [size=256]
Capabilities: [dc] Power Management version 0
Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-
00: ec 10 69 81 17 00 b0 02 10 00 00 02 08 80 00 00
10: 01 ff bf 00 00 fc ff bf 00 00 00 00 00 00 00 00
20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
30: 00 00 00 00 dc 00 00 00 00 00 00 00 10 01 20 40

dmesg when it didn't work (I do use netconsole, don't think it matters?):

r8169 Gigabit Ethernet driver 2.2LK loaded
eth0: RTL8169s/8110s at 0xc9004c00, 00:0d:0b:99:44:70, IRQ 16
netconsole: device eth0 not up yet, forcing it
r8169: eth0: link down
r8169: eth0: link up

The same when it's working.

Yes, just commenting out the line

	__rtl8169_set_mac_addr(dev, ioaddr);

fixes it (without the patch from your previous email).

Thanks
Guennadi
---
Guennadi Liakhovetski

#29 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Guennadi Liakhovetski <[email protected]> :
[...]

Quoted 2 lines

AFAIU, you wanted it applied on the top of the "non-working" kernel (2.6.19-rc2-ish)?

No. Please apply it on top of a 2.6.19-rc3 where the mac address change feature has been reverted (or where __rtl8169_set_mac_addr has been commented out at your option).

[...]

Quoted 1 line

dmesg when it didn't work (I do use netconsole, don't think it matters?):

I'd rather fix everything else without netconsole first. It will make my life simpler.

Signature · 1 line
Ueimor
#30 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Quoted 8 lines · Francois Romieu

Guennadi Liakhovetski <[email protected]> :
[...]

Quoted 2 lines

AFAIU, you wanted it applied on the top of the "non-working" kernel (2.6.19-rc2-ish)?

No. Please apply it on top of a 2.6.19-rc3 where the mac address change feature has been reverted (or where __rtl8169_set_mac_addr has been commented out at your option).

Ok, with just __rtl8169_set_mac_addr disabled it works. With netconsole disabled, and your phy_reset patch applied it seems to still work. The printk

+ printk(KERN_ERR "%s: PHY reset failed.\n", dev->name);

doesn't get printed. If I uncomment __rtl8169_set_mac_addr it stops working again. What does it tell us about the original set_mac_address problem?

I haven't said it's an on-board chip, not a plug-in card. Don't know how setting the mac address worked in your configuration, but if it is storred in a prom, maybe it is just missing on my board?

The kernel is not 2.6.19-rc3 either. It is a clone of the powerpc git some time shortly after 2.6.19-rc2.

Thanks
Guennadi
---
Guennadi Liakhovetski

#31 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Quoted 2 lines · Guennadi Liakhovetski

Ok, with just __rtl8169_set_mac_addr disabled it works. With netconsole disabled, and your phy_reset patch applied it seems to still work. The

The "seems" above was the key word. Once again I had a case, when after re-compiling the kernel again with the disabled call to __rtl8169_set_mac_addr only ping worked. And a power-off was required to recover. So, that phy_reset doesn't seem to be very safe either.

Thanks
Guennadi
---
Guennadi Liakhovetski

#32 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Guennadi Liakhovetski <[email protected]> :
[...]

Quoted 4 lines

The "seems" above was the key word. Once again I had a case, when after re-compiling the kernel again with the disabled call to __rtl8169_set_mac_addr only ping worked. And a power-off was required to recover. So, that phy_reset doesn't seem to be very safe either.

Can you replace phy_reset by the patch below and try it twice ?

It's interesting to know if it does not always behave the same.

drivers/net/r8169.c +20 -2
index f1c7575..4b05dea 100644
@@ -570,8 +570,8 @@ static void rtl8169_xmii_reset_enable(vo
 {
 	unsigned int val;
 
-	val = (mdio_read(ioaddr, MII_BMCR) | BMCR_RESET) & 0xffff;
-	mdio_write(ioaddr, MII_BMCR, val);
+	mdio_write(ioaddr, MII_BMCR, BMCR_RESET);
+	val = mdio_read(ioaddr, MII_BMCR);
 }
 
 static void rtl8169_check_link_status(struct net_device *dev,
@@ -1440,6 +1440,22 @@ static void rtl8169_release_board(struct
 	free_netdev(dev);
 }
 
+static void rtl8169_phy_reset(struct net_device *dev,
+			      struct rtl8169_private *tp)
+{
+	void __iomem *ioaddr = tp->mmio_addr;
+	int i;
+
+	tp->phy_reset_enable(ioaddr);
+	for (i = 0; i < 100; i++) {
+		if (!tp->phy_reset_pending(ioaddr))
+			return;
+		msleep(1);
+	}
+	if (netif_msg_link(tp))
+		printk(KERN_ERR "%s: PHY reset failed.\n", dev->name);
+}
+
 static void rtl8169_init_phy(struct net_device *dev, struct rtl8169_private *tp)
 {
 	void __iomem *ioaddr = tp->mmio_addr;
@@ -1468,6 +1484,8 @@ static void rtl8169_init_phy(struct net_
 
 	rtl8169_link_option(board_idx, &autoneg, &speed, &duplex);
 
+	rtl8169_phy_reset(dev, tp);
+
 	rtl8169_set_speed(dev, autoneg, speed, duplex);
 
 	if ((RTL_R8(PHYstatus) & TBI_Enable) && netif_msg_link(tp))
#33 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Quoted 10 lines · Francois Romieu

Guennadi Liakhovetski <[email protected]> :
[...]

Quoted 4 lines

The "seems" above was the key word. Once again I had a case, when after re-compiling the kernel again with the disabled call to __rtl8169_set_mac_addr only ping worked. And a power-off was required to recover. So, that phy_reset doesn't seem to be very safe either.

Can you replace phy_reset by the patch below and try it twice ?

It's interesting to know if it does not always behave the same.

Well, with that one I booted 3 times, all 3 times it worked. I'll leave it in to see if it ever fails. So, what does it tell us about the set_mac_address thing?

Thanks
Guennadi
---
Guennadi Liakhovetski

#34 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Guennadi Liakhovetski <[email protected]> :
[...]

Quoted 1 line

Well, with that one I booted 3 times, all 3 times it worked. I'll leave it

Thanks.

Let's cross fingers.

Quoted 2 lines

in to see if it ever fails. So, what does it tell us about the set_mac_address thing?

It tells nothing more about the set_mac_address thing. If people need MAC address change support, I can surely hack something and keep a patch for future reference. Imho it is anything but 2.6.19 material though.

The patch that I sent to you on 2006/10/29 was enough to fix the link detection issues experienced with the 0x8136 chipset (1. Darren Salt on netdev {25/26/31}/08/2006 and {21/22}/10/2006, 2. Syed Azam on BZ, see http://bugzilla.kernel.org/show_bug.cgi?id=7378).

Your computer was good at spotting issues with the MAC address stuff, so it was the perfect candidate to test pending fixes for different problems. As you noticed, it was not exactly safe to feed the MII control register with some potentially uninitialized stuff, whence the patch from yesterday.

It remains to be seen if:
- it still does the job for the 0x8136
- it does not induce new regressions in existing 8169

o Darren and Syed, are your 0x8136 still happy with the patch 0001-r8169-perform-a-PHY-reset-before-any-other-operation-at-boot-time.txt at http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.19-rc4/r8169
on top of 2.6.19-rc4 ?

o Darren, still ok to keep your S-o-b in it ?

o Could the r8169 users out there check that the same patch does not add new regressions to their favorite 2.6.19-rc4 ?

o Lennert, can you apply the two patches
- 0001-r8169-perform-a-PHY-reset-before-any-other-operation-at-boot-time.txt
- 0002-r8169-more-magic.txt
at http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.19-rc4/r8169 against 2.6.19-rc4 (2.6.19-rc4 reverted the MAC address changes) and see if the n2100 board still needs to remove the SYSErr handler ?

Signature · 1 line
Ueimor
#35 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Quoted 9 lines · Francois Romieu

Guennadi Liakhovetski <[email protected]> :

Quoted 2 lines

in to see if it ever fails. So, what does it tell us about the set_mac_address thing?

It tells nothing more about the set_mac_address thing. If people need MAC address change support, I can surely hack something and keep a patch for future reference. Imho it is anything but 2.6.19 material though.

Aha, ok, thanks. Just noticed that the set_mac_address has been reverted in -rc4, so, that's resolved. Good.

Quoted 5 lines

Your computer was good at spotting issues with the MAC address stuff, so it was the perfect candidate to test pending fixes for different problems. As you noticed, it was not exactly safe to feed the MII control register with some potentially uninitialized stuff, whence the patch from yesterday.

Glad it was useful. I have to warn you though, that that "computer" is not very actively used ATM and doesn't stay on for too long. However, if you can suggest a way to stress test that phy reset thingie, I could run some overnight test.

Thanks
Guennadi
---
Guennadi Liakhovetski

#36 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

In reply to Francois Romieu
Quoted 6 lines · Francois Romieu

o Lennert, can you apply the two patches
- 0001-r8169-perform-a-PHY-reset-before-any-other-operation-at-boot-time.txt
- 0002-r8169-more-magic.txt
at http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.19-rc4/r8169 against 2.6.19-rc4 (2.6.19-rc4 reverted the MAC address changes) and see if the n2100 board still needs to remove the SYSErr handler ?

2.6.19-rc4 + these two patches => doesn't work

2.6.19-rc4 + these two patches + SYSErr removal => works

For reference:
* 2.6.19-rc4 + SYSErr removal => works

So, while these two patches don't fix the problem, they don't seem to be making things worse, something the MAC address change did do.

cheers,
Lennert

#37 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

In reply to Francois Romieu

This message is in MIME format which your mailer apparently does not support. You either require a newer version of your software which supports MIME, or a separate MIME decoding utility. Alternatively, ask the sender of this message to resend it in a different format.

--163681386--1739281754--1718250983
Content-Type: text/plain; charset=us-ascii

I demand that Francois Romieu may or may not have written...

[snip]

Quoted 4 lines

o Darren and Syed, are your 0x8136 still happy with the patch 0001-r8169-perform-a-PHY-reset-before-any-other-operation-at-boot-time.txt at http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.19-rc4/r8169
on top of 2.6.19-rc4 ?

It is.

I then tried 0002-r8169-more-magic.txt, mainly to see that it doesn't cause any problems. Unfortunately, it did... however, a little reading showed that you've stopped enabling transmit and receive for all but four of the chip types supported by the driver.

An incremental fix is attached.

Quoted 1 line

o Darren, still ok to keep your S-o-b in it ?

Yes.

[snip]

Signature · 35 lines
| Darren Salt    | linux or ds at              | nr. Ashington, | Toon
| RISC OS, Linux | youmustbejoking,demon,co,uk | Northumberland | Army
| + Buy less and make it last longer.         INDUSTRY CAUSES GLOBAL WARMING.

Never have a drink when you are feeling sorry for yourself.

--163681386--1739281754--1718250983
Content-Type: text/plain; charset=iso-8859-1; name="0003-r8169-fix-more-magic.patch"
Content-Disposition: attachment; filename="0003-r8169-fix-more-magic.patch"

r8169: fix more magic so that 8101e etc. work again

r8169-more-magic killed transmit & receive on my laptop's RTL8101e. Turns out
to be that the enabling of these functions had been unitnentionally removed.

(This is one of two possible fixes, the other being the removal of the if()
guarding the other tx/rx-enable call. Both work here.)

Signed-off-by: Darren Salt <[email protected]>

--- drivers/net/r8169.c~	2006-11-01 19:53:02.000000000 +0000
+++ drivers/net/r8169.c	2006-11-01 19:54:16.000000000 +0000
@@ -1921,7 +1921,10 @@
 	    (tp->mac_version != RTL_GIGA_MAC_VER_02) &&
 	    (tp->mac_version != RTL_GIGA_MAC_VER_03) &&
 	    (tp->mac_version != RTL_GIGA_MAC_VER_04))
+	{
 		rtl8169_set_rx_tx_config_registers(tp);
+		RTL_W8(ChipCmd, CmdTxEnb | CmdRxEnb);
+	}
 
 	RTL_W8(Cfg9346, Cfg9346_Lock);
 

--163681386--1739281754--1718250983--
#38 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

Darren Salt <[email protected]> :
[...]

Quoted 2 lines

(This is one of two possible fixes, the other being the removal of the if() guarding the other tx/rx-enable call. Both work here.)

I'll update the patch with your change but the removal of the if() would not match Realtek's init sequence.

Lennert, I have compared 2.6.19-rc4 + 0001-r8169-perform-a-PHY-reset-etc with the serie of patches against 2.6.18-rc4 which was reported to work on your n2100 (thread on netdev around 05/09/2006). Can you:

- apply the patch below on top of 2.6.19-rc4 + 0001 and see if it works ? Don't apply 0002, it is not required.

- if it works (it should if I have not messed it again), can you check that it still works if you do not apply the first hunk ? It should as well.

If everything went fine this far, it would be nice to know if both the move of the write to ChipCmd and the mdio_write are required to fix your issue (I'd expect the move alone to not be enough as 0002 got this part right for the 8110sb).

drivers/net/r8169.c +4 -7
index 50b753d..b2fdbb8 100644
@@ -491,7 +491,7 @@ static int rtl8169_poll(struct net_devic
 #endif
 
 static const u16 rtl8169_intr_mask =
-	SYSErr | LinkChg | RxOverflow | RxFIFOOver | TxErr | TxOK | RxErr | RxOK;
+	LinkChg | RxOverflow | RxFIFOOver | TxErr | TxOK | RxErr | RxOK;
 static const u16 rtl8169_napi_event =
 	RxOK | RxOverflow | RxFIFOOver | TxOK | TxErr;
 static const unsigned int rtl8169_rx_config =
@@ -1283,11 +1283,6 @@ static void rtl8169_hw_phy_config(struct
 	/* Shazam ! */
 
 	if (tp->mac_version == RTL_GIGA_MAC_VER_04) {
-		mdio_write(ioaddr, 31, 0x0001);
-		mdio_write(ioaddr,  9, 0x273a);
-		mdio_write(ioaddr, 14, 0x7bfb);
-		mdio_write(ioaddr, 27, 0x841e);
-
 		mdio_write(ioaddr, 31, 0x0002);
 		mdio_write(ioaddr,  1, 0x90d0);
 		mdio_write(ioaddr, 31, 0x0000);
@@ -1855,6 +1850,8 @@ rtl8169_hw_start(struct net_device *dev)
 
 
 	RTL_W8(Cfg9346, Cfg9346_Unlock);
+
+	RTL_W8(ChipCmd, CmdTxEnb | CmdRxEnb);
 	RTL_W8(EarlyTxThres, EarlyTxThld);
 
 	/* Low hurts. Let's disable the filtering. */
@@ -1895,7 +1892,7 @@ rtl8169_hw_start(struct net_device *dev)
 	RTL_W32(TxDescStartAddrLow, ((u64) tp->TxPhyAddr & DMA_32BIT_MASK));
 	RTL_W32(RxDescAddrHigh, ((u64) tp->RxPhyAddr >> 32));
 	RTL_W32(RxDescAddrLow, ((u64) tp->RxPhyAddr & DMA_32BIT_MASK));
-	RTL_W8(ChipCmd, CmdTxEnb | CmdRxEnb);
+
 	RTL_W8(Cfg9346, Cfg9346_Lock);
 
 	/* Initially a 10 us delay. Turned it into a PCI commit. - FR */
#39 lore raw

RE: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

In reply to Darren Salt

Yes, It is still working.

Syed Azam
System Software Engineer
Hewlett-Packard Company

-----Original Message-----
From: Darren Salt [mailto:[email protected]]
Sent: Wednesday, November 01, 2006 1:01 PM
To: [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Azam, Syed S; [email protected]
Subject: Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

This message is in MIME format which your mailer apparently does not support.
You either require a newer version of your software which supports MIME, or
a separate MIME decoding utility. Alternatively, ask the sender of this message to resend it in a different format.

--163681386--1739281754--1718250983
Content-Type: text/plain; charset=us-ascii

I demand that Francois Romieu may or may not have written...

[snip]

Quoted 2 lines

o Darren and Syed, are your 0x8136 still happy with the patch

0001-r8169-perform-a-PHY-reset-before-any-other-operation-at-boot-time.t xt

Quoted 2 lines

at http://www.fr.zoreil.com/linux/kernel/2.6.x/2.6.19-rc4/r8169
on top of 2.6.19-rc4 ?

It is.

I then tried 0002-r8169-more-magic.txt, mainly to see that it doesn't cause
any problems. Unfortunately, it did... however, a little reading showed that
you've stopped enabling transmit and receive for all but four of the chip
types supported by the driver.

An incremental fix is attached.

Quoted 1 line

o Darren, still ok to keep your S-o-b in it ?

Yes.

[snip]

Signature · 41 lines
| Darren Salt    | linux or ds at              | nr. Ashington, | Toon
| RISC OS, Linux | youmustbejoking,demon,co,uk | Northumberland | Army
| + Buy less and make it last longer.         INDUSTRY CAUSES GLOBAL
WARMING.

Never have a drink when you are feeling sorry for yourself.

--163681386--1739281754--1718250983
Content-Type: text/plain; charset=iso-8859-1;
name="0003-r8169-fix-more-magic.patch"
Content-Disposition: attachment;
filename="0003-r8169-fix-more-magic.patch"

r8169: fix more magic so that 8101e etc. work again

r8169-more-magic killed transmit & receive on my laptop's RTL8101e.
Turns out
to be that the enabling of these functions had been unitnentionally
removed.

(This is one of two possible fixes, the other being the removal of the
if()
guarding the other tx/rx-enable call. Both work here.)

Signed-off-by: Darren Salt <[email protected]>

--- drivers/net/r8169.c~	2006-11-01 19:53:02.000000000 +0000
+++ drivers/net/r8169.c	2006-11-01 19:54:16.000000000 +0000
@@ -1921,7 +1921,10 @@
 	    (tp->mac_version != RTL_GIGA_MAC_VER_02) &&
 	    (tp->mac_version != RTL_GIGA_MAC_VER_03) &&
 	    (tp->mac_version != RTL_GIGA_MAC_VER_04))
+	{
 		rtl8169_set_rx_tx_config_registers(tp);
+		RTL_W8(ChipCmd, CmdTxEnb | CmdRxEnb);
+	}
 
 	RTL_W8(Cfg9346, Cfg9346_Lock);
 

--163681386--1739281754--1718250983--
#40 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

In reply to Guennadi Liakhovetski

Guennadi Liakhovetski <[email protected]> :
[...]

Quoted 3 lines

doesn't get printed. If I uncomment __rtl8169_set_mac_addr it stops working again. What does it tell us about the original set_mac_address problem?

Probably that it is issued too early/bluntly. I'll redo it later.

[...]

Quoted 2 lines

The kernel is not 2.6.19-rc3 either. It is a clone of the powerpc git some time shortly after 2.6.19-rc2.

You miss 73f5e28b336772c4b08ee82e5bf28ab872898ee1 and 733b736c91dd2c556f35dffdcf77e667cf10cefc. It should not matter.

Signature · 1 line
Ueimor
#41 lore raw

Re: r8169 mac address change (was Re: [0/3] 2.6.19-rc2: known regressions)

In reply to Francois Romieu
Quoted 9 lines · Francois Romieu

Linus Torvalds <[email protected]> :
[regression related to r8169 MAC address change]

Quoted 1 line

Francois ? Jeff ?

Go revert it.

Despite what I claimed, I can not find a third-party confirmation by email that it works elsewhere.

It works for me:
,--

|root@flower:/home/olecom# ip l set eth0 addr 00:00:00:00:00:02
|root@flower:/home/olecom# ip l set eth0 addr 00:00:00:00:00:03
|root@flower:/home/olecom# ip l set eth0 addr 00:00:00:00:00:04
|root@flower:/home/olecom# ip l set eth0 addr 04:44:44:44:44:04
|root@flower:/home/olecom# ip l show
|1: lo: <LOOPBACK,UP,10000> mtu 16436 qdisc noqueue
|    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
|    2: eth0: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen 1000
|    link/ether 04:44:44:44:44:04 brd ff:ff:ff:ff:ff:ff
|root@flower:/home/olecom# lspci -v | grep Realtek
|    02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
|RTL8111/8168B PCI Express Gigabit Ethernet controller (rev 01)
|root@flower:/home/olecom#
|root@flower:/home/olecom# uname -a
|Linux flower 2.6.19-rc2-vanilla-ai #4 SMP PREEMPT Tue Oct 17 02:19:16
|UTC 2006 x86_64 GNU/Linux
|root@flower:/home/olecom#
`--
____
#42 lore raw
In reply to Linus Torvalds
Quoted 2 lines · Linus Torvalds

Ok, it's a week since -rc1, so -rc2 is out there.

The tar.gz is broken:
tar tvfz /mounts/mirror/kernel/v2.6/testing/linux-2.6.19-rc2.tar.gz| head
-rw-r--r-- git/git 542 2006-10-13 18:25:04 linux-2.6.19-rc2.gitignore
-rw-r--r-- git/git 18693 2006-10-13 18:25:04 linux-2.6.19-rc2COPYING
-rw-r--r-- git/git 90289 2006-10-13 18:25:04 linux-2.6.19-rc2CREDITS

#43 lore raw

2.6.19-rc2: known unfixed regressions (v2)

In reply to Linus Torvalds

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

Subject : ppc prep boot hang
References : http://lkml.org/lkml/2006/10/14/58
Submitter : Meelis Roos <[email protected]>
Status : unknown

Subject : X60s: BUG()s, lose ACPI events after suspend/resume
References : http://lkml.org/lkml/2006/10/10/39
Submitter : Martin Lorenz <[email protected]>
Status : unknown

Subject : T60 stops triggering any ACPI events
References : http://lkml.org/lkml/2006/10/4/425

             http://lkml.org/lkml/2006/10/16/262

Submitter : "Michael S. Tsirkin" <[email protected]>
Status : unknown

Subject : x86_64: build error: unknown pseudo-op: `.cfi_signal_frame'
References : http://lkml.org/lkml/2006/10/16/8
Submitter : Ingo Molnar <[email protected]>

Andi Kleen <[email protected]>
commit adf1423698f00d00b267f7dca8231340ce7d65ef
Status : Jan: That is nothing I added, I suppose Andi did.
But the fix is obvious and trivial.

Subject : undefined reference to highest_possible_node_id
References : http://lkml.org/lkml/2006/9/4/233

             http://lkml.org/lkml/2006/10/15/11

Submitter : Olaf Hering <[email protected]>

             commit 0f532f3861d2c4e5aa7dcd33fb18e9975eb28457

Status : unknown

Subject : many ARM compile failures after the post -rc1 IRQ patches
References : http://lkml.org/lkml/2006/10/15/29

             http://armlinux.simtec.co.uk/kautobuild/2.6.19-rc2/index.html
Submitter  : Russell King <[email protected]>
Caused-By  : David Howells <[email protected]>
Status     : unknown

Subject : CONFIG_X86_VOYAGER=y, CONFIG_SMP=n compile error
References : http://lkml.org/lkml/2006/10/7/51
Submitter : Jesper Juhl <[email protected]>

             commit 7d12e780e003f93433d49ce78cfedf4b4c52adc5

Status : unknown

Subject : CONFIG_X86_VISWS=y, CONFIG_SMP=n compile error
References : http://lkml.org/lkml/2006/10/7/51
Submitter : Jesper Juhl <[email protected]>

             commit 7d12e780e003f93433d49ce78cfedf4b4c52adc5

Status : unknown

Subject : sata-via doesn't detect anymore disks attached to VIA vt6421
References : http://bugzilla.kernel.org/show_bug.cgi?id=7255
Submitter : Thierry Vignaud <[email protected]>
Status : unknown

Subject : SMP x86_64 boot problem
References : http://lkml.org/lkml/2006/9/28/330

             http://lkml.org/lkml/2006/10/5/289

Submitter : [email protected]
Status : submitter was asked to git bisect
result of bisecting seems to be wrong

Subject : monitor not active after boot
References : http://lkml.org/lkml/2006/10/5/338
Submitter : Olaf Hering <[email protected]>

             commit 346bc21026e7a92e1d7a4a1b3792c5e8b686133d

Status : unknown

Subject : unable to rip cd
References : http://lkml.org/lkml/2006/10/13/100
Submitter : Alex Romosan <[email protected]>
Status : unknown

Subject : ide-generic no longer finds marvell controller
References : http://bugzilla.kernel.org/show_bug.cgi?id=7353
Submitter : Kenny Graunke <[email protected]>

             commit a4bea10eca68152e84ffc4eaeb9d20ec2ac34664

Status : Alan is investigating

Subject : cpufreq not working on AMD K8
References : http://lkml.org/lkml/2006/10/10/114
Submitter : Christian <[email protected]>

Status : Mark is investigating

#44 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v2)

Quoted 6 lines · Adrian Bunk

Subject : monitor not active after boot
References : http://lkml.org/lkml/2006/10/5/338
Submitter : Olaf Hering <[email protected]>

             commit 346bc21026e7a92e1d7a4a1b3792c5e8b686133d

Status : unknown

The nvidiafb change was removed again. I will see if I can figure out why the EDID is lost.

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

#45 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v2)

Quoted 11 lines · Olaf Hering
Quoted 6 lines · Adrian Bunk

Subject : monitor not active after boot
References : http://lkml.org/lkml/2006/10/5/338
Submitter : Olaf Hering <[email protected]>

             commit 346bc21026e7a92e1d7a4a1b3792c5e8b686133d

Status : unknown

The nvidiafb change was removed again. I will see if I can figure out why the EDID is lost.

Thanks for the information, I missed this patch.

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

#49 lore raw

PATCH 1/1 drivers/ide/pci/generic.c: re-add the __setup("all-generic-ide",...)

Quoted 5 lines · Alan Cox
Quoted 4 lines · Adrian Bunk

IOW, your patch does break existing setups since the change to module_param() requires prefixing with the module name (the ata_generic option with the same name is irrelevant)?

Quoted 7 lines
Quoted 4 lines

Considering that drivers/ide/ is slowly approaching a RIP status, is such an incompatible change really required?

I'd be more inclined to revert your patch.

We shouldn't revert it - there is a real problem for some users whose distro has it modular this fixed. We might want to honour both

Agreed, patch below.

cu
Adrian

<-- snip -->

The change from __setup() to module_param_named() requires users to prefix the option with "generic.".

This patch re-adds the __setup() additionally to the module_param_named().

Usually it would make sense getting rid of such an obsolete __setup() at some time, but considering that drivers/ide/ is slowly approaching a RIP status it's already implicitely scheduled for removal.

This patch fixes kernel Bugzilla #7353.

drivers/ide/pci/generic.c renamed from drivers/ide/pci/generic.c.old +13 -0
@@ -40,6 +40,19 @@
 
 static int ide_generic_all;		/* Set to claim all devices */
 
+/*
+ * the module_param_named() was added for the modular case
+ * the __setup() is left as compatibility for existing setups
+ */
+#ifndef MODULE
+static int __init ide_generic_all_on(char *unused)
+{
+	ide_generic_all = 1;
+	printk(KERN_INFO "IDE generic will claim all unknown PCI IDE storage controllers.");
+	return 1;
+}
+__setup("all-generic-ide", ide_generic_all_on);
+#endif
 module_param_named(all_generic_ide, ide_generic_all, bool, 0444);
 MODULE_PARM_DESC(all_generic_ide, "IDE generic will claim all unknown PCI IDE storage controllers.");
 
#50 lore raw

Re: [2.6.19 patch] drivers/ide/pci/generic.c: re-add the __setup("all-generic-ide",...)

Quoted 48 lines · On Thu, 19 Oct 2006 17:26:51 +0200 Adrian Bunk wrote:
Quoted 5 lines · Alan Cox
Quoted 4 lines · Adrian Bunk

IOW, your patch does break existing setups since the change to module_param() requires prefixing with the module name (the ata_generic option with the same name is irrelevant)?

Quoted 7 lines
Quoted 4 lines

Considering that drivers/ide/ is slowly approaching a RIP status, is such an incompatible change really required?

I'd be more inclined to revert your patch.

We shouldn't revert it - there is a real problem for some users whose distro has it modular this fixed. We might want to honour both

Agreed, patch below.

cu
Adrian

<-- snip -->

The change from __setup() to module_param_named() requires users to prefix the option with "generic.".

This patch re-adds the __setup() additionally to the module_param_named().

Usually it would make sense getting rid of such an obsolete __setup() at some time, but considering that drivers/ide/ is slowly approaching a RIP status it's already implicitely scheduled for removal.

This patch fixes kernel Bugzilla #7353.

drivers/ide/pci/generic.c renamed from drivers/ide/pci/generic.c.old +5 -0
@@ -40,6 +40,19 @@
 
 static int ide_generic_all;		/* Set to claim all devices */
 
+/*
+ * the module_param_named() was added for the modular case
+ * the __setup() is left as compatibility for existing setups
+ */
+#ifndef MODULE

drivers/ide/pci/generic.c line 57

Quoted 10 lines
drivers/ide/pci/generic.c renamed from drivers/ide/pci/generic.c.old +8 -0
+static int __init ide_generic_all_on(char *unused)
+{
+	ide_generic_all = 1;
+	printk(KERN_INFO "IDE generic will claim all unknown PCI IDE storage controllers.");
+	return 1;
+}
+__setup("all-generic-ide", ide_generic_all_on);
+#endif
 module_param_named(all_generic_ide, ide_generic_all, bool, 0444);
 MODULE_PARM_DESC(all_generic_ide, "IDE generic will claim all unknown PCI IDE storage controllers.");

Missing update to Documentation/kernel-parameters.txt ? (maybe it's been missing forever?)

---
~Randy

#51 lore raw

Re: [2.6.19 patch] drivers/ide/pci/generic.c: re-add the __setup("all-generic-ide",...)

Quoted 18 lines · Randy Dunlap

On Thu, 19 Oct 2006 17:26:51 +0200 Adrian Bunk wrote:
...

Quoted 16 lines
drivers/ide/pci/generic.c renamed from drivers/ide/pci/generic.c.old +10 -0
@@ -40,6 +40,19 @@
 
 static int ide_generic_all;		/* Set to claim all devices */
 
+/*
+ * the module_param_named() was added for the modular case
+ * the __setup() is left as compatibility for existing setups
+ */
+#ifndef MODULE
+static int __init ide_generic_all_on(char *unused)
+{
+	ide_generic_all = 1;
+	printk(KERN_INFO "IDE generic will claim all unknown PCI IDE storage controllers.");
+	return 1;
Quoted 8 lines

drivers/ide/pci/generic.c line 57

Quoted 5 lines
drivers/ide/pci/generic.c renamed from drivers/ide/pci/generic.c.old +3 -0
+}
+__setup("all-generic-ide", ide_generic_all_on);
+#endif
 module_param_named(all_generic_ide, ide_generic_all, bool, 0444);
 MODULE_PARM_DESC(all_generic_ide, "IDE generic will claim all unknown PCI IDE storage controllers.");

Missing update to Documentation/kernel-parameters.txt ? (maybe it's been missing forever?)

It's been missing forever.

I'm not sure whether documenting it now where it's deprecated and nearly dead makes sense..

Quoted 1 line

~Randy

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#52 lore raw

Re: [2.6.19 patch] drivers/ide/pci/generic.c: re-add the __setup("all-generic-ide",...)

Ar Iau, 2006-10-19 am 18:13 +0200, ysgrifennodd Adrian Bunk:

Quoted 7 lines
Quoted 2 lines

Missing update to Documentation/kernel-parameters.txt ? (maybe it's been missing forever?)

It's been missing forever.

I'm not sure whether documenting it now where it's deprecated and nearly dead makes sense..

Its not dead, its so useful that drivers/ata also supports it

#53 lore raw

Re: [2.6.19 patch] drivers/ide/pci/generic.c: re-add the __setup("all-generic-ide",...)

Quoted 10 lines · Alan Cox

Ar Iau, 2006-10-19 am 18:13 +0200, ysgrifennodd Adrian Bunk:

Quoted 7 lines
Quoted 2 lines

Missing update to Documentation/kernel-parameters.txt ? (maybe it's been missing forever?)

It's been missing forever.

I'm not sure whether documenting it now where it's deprecated and nearly dead makes sense..

Its not dead, its so useful that drivers/ata also supports it

But in the drivers/ata case it's a module parameter, not a __setup kernel parameter.

And I don't think it makes sense to manually add module parameters to kernel-parameters.txt

If a documentation of all module parameters is considered useful, someone should write a script to automatically generate such a list.

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#54 lore raw

Re: [2.6.19 patch] drivers/ide/pci/generic.c: re-add the __setup("all-generic-ide",...)

Quoted 7 lines · On Fri, 20 Oct 2006 23:05:33 +0200 Adrian Bunk wrote:
Quoted 6 lines · Alan Cox

Ar Iau, 2006-10-19 am 18:13 +0200, ysgrifennodd Adrian Bunk:

Quoted 5 lines
Quoted 2 lines

Missing update to Documentation/kernel-parameters.txt ? (maybe it's been missing forever?)

It's been missing forever.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

I'm not sure whether documenting it now where it's deprecated and nearly dead makes sense..

Its not dead, its so useful that drivers/ata also supports it

But in the drivers/ata case it's a module parameter, not a __setup kernel parameter.

That's just an implementation nit/detail. Users don't care which way it's implemented, they just need to see some reasonable documentation.

Quoted 2 lines

And I don't think it makes sense to manually add module parameters to kernel-parameters.txt

There are module parameters there already...
so we are being inconsistent.

Quoted 2 lines

If a documentation of all module parameters is considered useful, someone should write a script to automatically generate such a list.

I think that sounds great -- in theory. Really, I do. I even wrote a (simple) script for it last night.[1] (but someone else is free to redo it, and probably not in shell script :)

And maybe one "development community" answer is that this is a distro problem, let them handle it. (I don't like that answer, but possibly the distros are OK with it. I don't know.)

Ideally, users would be able to see/read documentation (like kernel or module parameters) (a) without reading the source code and
(b) without building the module binary files. Maybe that's too much to ask of the development community, so the users can just build all 1500 or so (and growing) loadable kernel modules and run 'modinfo' on them to see what the possible module parameters are. Of course, if they need this information to be able to install their (only) Linux system, then they are out of luck, or they can use their other (or working) OS to search the internet for such documenation.

Anyway, regarding your suggestion: Yes, I think that it would be good to generate such documentation instead of maintaining it (and sometimes not doing that). Maybe someone can & will make that happen.

---
~Randy
[1] http://www.xenotime.net/linux/scripts/module-params

#55 lore raw
In reply to Linus Torvalds
Quoted 1 line · Linus Torvalds

Ok, it's a week since -rc1, so -rc2 is out there.

A bit behind, but booting still takes ages on my laptop as libata/ata_piix tries to probe a device that isn't there (I reported this previously against -rc1, but got no response):

Linux version 2.6.19-rc2 (root@farstrider) (gcc version 4.1.2 20061007 (prerelease) (Debian 4.1.1-16)) #1 PREEMPT Tue Oct 17 11:40:02 EDT 2006
BIOS-provided physical RAM map:
BIOS-e820: 0000000000000000 - 000000000009f800 (usable)
BIOS-e820: 000000000009f800 - 00000000000a0000 (reserved)
BIOS-e820: 00000000000ce000 - 00000000000d0000 (reserved)
BIOS-e820: 00000000000dc000 - 0000000000100000 (reserved)
BIOS-e820: 0000000000100000 - 000000003f6f0000 (usable)
BIOS-e820: 000000003f6f0000 - 000000003f6fb000 (ACPI data)
BIOS-e820: 000000003f6fb000 - 000000003f700000 (ACPI NVS)
BIOS-e820: 000000003f700000 - 0000000040000000 (reserved)
BIOS-e820: 00000000fec10000 - 00000000fec20000 (reserved)
BIOS-e820: 00000000ffb00000 - 00000000ffc00000 (reserved)
BIOS-e820: 00000000fff00000 - 0000000100000000 (reserved) 1014MB LOWMEM available.
Entering add_active_range(0, 0, 259824) 0 entries of 256 used Zone PFN ranges:
DMA 0 -> 4096
Normal 4096 -> 259824
early_node_map[1] active PFN ranges
0: 0 -> 259824
On node 0 totalpages: 259824
DMA zone: 32 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 4064 pages, LIFO batch:0
Normal zone: 1997 pages used for memmap
Normal zone: 253731 pages, LIFO batch:31
DMI 2.3 present.
ACPI: RSDP (v000 FUJ ) @ 0x000f5d80
ACPI: RSDT (v001 FUJ FJNB189 0x01070000 FUJ 0x00000100) @ 0x3f6f4c92
ACPI: FADT (v001 FUJ FJNB189 0x01070000 FUJ 0x00000100) @ 0x3f6faf8c
ACPI: SSDT (v001 FUJ FJNB189 0x01070000 INTL 0x20030522) @ 0x3f6fab42
ACPI: BOOT (v001 FUJ FJNB189 0x01070000 FUJ 0x00000100) @ 0x3f6faf64
ACPI: DSDT (v001 FUJ FJNB189 0x01070000 MSFT 0x0100000e) @ 0x00000000
ACPI: PM-Timer IO Port: 0xfc08
Allocating PCI resources starting at 50000000 (gap: 40000000:bec10000) Detected 1100.104 MHz processor.
Built 1 zonelists. Total pages: 257795
Kernel command line: root=/dev/sda3 ro acpi_sleep=s3_bios resume2=file:/dev/sda3:0x1f2df40
Enabling fast FPU save and restore... done.
Enabling unmasked SIMD FPU exception support... done. Initializing CPU#0
CPU 0 irqstacks, hard=b0348000 soft=b0347000
PID hash table entries: 4096 (order: 12, 16384 bytes)
Console: colour VGA+ 80x25
Dentry cache hash table entries: 131072 (order: 7, 524288 bytes) Inode-cache hash table entries: 65536 (order: 6, 262144 bytes)
Memory: 1027448k/1039296k available (1660k kernel code, 11372k reserved, 525k data, 120k init, 0k highmem)
virtual kernel memory layout:

   fixmap  : 0xffff8000 - 0xfffff000   (  28 kB)
   vmalloc : 0xf0000000 - 0xffff6000   ( 255 MB)
   lowmem  : 0xb0000000 - 0xef6f0000   (1014 MB)
     .init : 0xb0324000 - 0xb0342000   ( 120 kB)
     .data : 0xb029f263 - 0xb0322a2c   ( 525 kB)
     .text : 0xb0100000 - 0xb029f263   (1660 kB)

Checking if this processor honours the WP bit even in supervisor mode... Ok. Calibrating delay using timer specific routine.. 2200.79 BogoMIPS (lpj=1100399) Mount-cache hash table entries: 512
CPU: After generic identify, caps: a7e9f9bf 00000000 00000000 00000000 00000180 00000000 00000000
CPU: L1 I cache: 32K, L1 D cache: 32K
CPU: L2 cache: 1024K
CPU: After all inits, caps: a7e9f9bf 00000000 00000000 00000040 00000180 00000000 00000000
Intel machine check architecture supported.
Intel machine check reporting enabled on CPU#0.
Compat vDSO mapped to ffffe000.
CPU: Intel(R) Pentium(R) M processor 1100MHz stepping 05
Checking 'hlt' instruction... OK.
ACPI: Core revision 20060707
ACPI: setting ELCR to 0200 (from 0800)
NET: Registered protocol family 16
ACPI: bus type pci registered
PCI: PCI BIOS revision 2.10 entry at 0xfd982, last bus=3
PCI: Using configuration type 1
Setting up standard PCI resources
ACPI: Interpreter enabled
ACPI: Using PIC for interrupt routing
ACPI: PCI Root Bridge [PCI0] (0000:00)
PCI: Probing PCI hardware (bus 00)
Boot video device is 0000:00:02.0
PCI quirk: region fc00-fc7f claimed by ICH4 ACPI/GPIO/TCO
PCI quirk: region fc80-fcbf claimed by ICH4 GPIO
PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
PCI: Transparent bridge - 0000:00:1e.0
PCI: Bus #03 (-#06) is hidden behind transparent bridge #01 (-#03) (try 'pci=assign-busses')
Please report the result to linux-kernel to fix this permanently
ACPI: PCI Interrupt Routing Table [\_SB_.PCI0._PRT]
ACPI: PCI Interrupt Link [LNKA] (IRQs 3 4 5 7 9 10 *11)
ACPI: PCI Interrupt Link [LNKB] (IRQs 3 4 5 7 9 10 *11)
ACPI: PCI Interrupt Link [LNKC] (IRQs 3 4 5 7 9 10 *11)
ACPI: PCI Interrupt Link [LNKD] (IRQs 3 4 5 7 9 10 *11)
ACPI: PCI Interrupt Link [LNKE] (IRQs 3 4 5 7 9 10 *11)
ACPI: PCI Interrupt Link [LNKF] (IRQs 3 4 5 7 9 10 11) *0, disabled.
ACPI: PCI Interrupt Link [LNKG] (IRQs 3 4 5 7 9 10 11) *0, disabled.
ACPI: PCI Interrupt Link [LNKH] (IRQs 3 4 5 7 9 10 *11)
ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.HUB_._PRT]
Linux Plug and Play Support v0.97 (c) Adam Belay
pnp: PnP ACPI init
pnp: PnP ACPI: found 8 devices
SCSI subsystem initialized
PCI: Using ACPI for IRQ routing
PCI: If a device doesn't work, try "pci=routeirq". If it helps, post a report
PCI: Ignore bogus resource 6 [0:0] of 0000:00:02.0
PCI: Bus 2, cardbus bridge: 0000:01:0a.0
IO window: 00003400-000034ff
IO window: 00003800-000038ff
PREFETCH window: 50000000-51ffffff
MEM window: 56000000-57ffffff
PCI: Bus 3, cardbus bridge: 0000:01:0a.1
IO window: 00003c00-00003cff
IO window: 00001000-000010ff
PREFETCH window: 52000000-53ffffff
MEM window: 58000000-59ffffff
PCI: Bridge: 0000:00:1e.0
IO window: 3000-3fff
MEM window: d0200000-d02fffff
PREFETCH window: 50000000-53ffffff
PCI: Setting latency timer of device 0000:00:1e.0 to 64
ACPI: PCI Interrupt Link [LNKA] enabled at IRQ 11
PCI: setting IRQ 11 as level-triggered
ACPI: PCI Interrupt 0000:01:0a.0[A] -> Link [LNKA] -> GSI 11 (level, low) -> IRQ 11
ACPI: PCI Interrupt Link [LNKB] enabled at IRQ 11
ACPI: PCI Interrupt 0000:01:0a.1[B] -> Link [LNKB] -> GSI 11 (level, low) -> IRQ 11
NET: Registered protocol family 2
IP route cache hash table entries: 32768 (order: 5, 131072 bytes) TCP established hash table entries: 131072 (order: 7, 524288 bytes) TCP bind hash table entries: 65536 (order: 6, 262144 bytes)
TCP: Hash tables configured (established 131072 bind 65536)
TCP reno registered
Simple Boot Flag at 0x7f set to 0x1
SGI XFS with no debug enabled
io scheduler noop registered
io scheduler cfq registered (default)
ACPI: AC Adapter [AC] (on-line)
ACPI: Battery Slot [CMB1] (battery present)
ACPI: Battery Slot [CMB2] (battery absent)
ACPI: Power Button (FF) [PWRF]
ACPI: Power Button (CM) [PWRB]
ACPI: Lid Switch [LID]
ACPI: CPU0 (power states: C1[C1] C2[C2] C3[C3])
Linux agpgart interface v0.101 (c) Dave Jones
agpgart: Detected an Intel 855 Chipset.
agpgart: Detected 8060K stolen memory.
agpgart: AGP aperture is 128M @ 0xd8000000
libata version 2.00 loaded.
ata_piix 0000:00:1f.1: version 2.00ac6
PCI: Enabling device 0000:00:1f.1 (0005 -> 0007)
ACPI: PCI Interrupt Link [LNKC] enabled at IRQ 11
ACPI: PCI Interrupt 0000:00:1f.1[A] -> Link [LNKC] -> GSI 11 (level, low) -> IRQ 11
PCI: Setting latency timer of device 0000:00:1f.1 to 64
ata1: PATA max UDMA/100 cmd 0x1F0 ctl 0x3F6 bmdma 0x1810 irq 14
ata2: PATA max UDMA/100 cmd 0x170 ctl 0x376 bmdma 0x1818 irq 15 scsi0 : ata_piix
ata1.00: ATA-6, max UDMA/100, 156301488 sectors: LBA
ata1.00: ata1: dev 0 multi count 16
ata1.00: configured for UDMA/100
scsi1 : ata_piix
ata2.00: ATAPI, max UDMA/33
ata2.01: qc timeout (cmd 0xa1)
ata2.01: failed to IDENTIFY (I/O error, err_mask=0x4)
ata2: failed to recover some devices, retrying in 5 secs
ata2.01: qc timeout (cmd 0xa1)
ata2.01: failed to IDENTIFY (I/O error, err_mask=0x4)
ata2: failed to recover some devices, retrying in 5 secs
ata2.01: qc timeout (cmd 0xa1)
ata2.01: failed to IDENTIFY (I/O error, err_mask=0x4)
ata2: failed to recover some devices, retrying in 5 secs
ata2.00: configured for UDMA/33
scsi 0:0:0:0: Direct-Access ATA FUJITSU MHT2080A 0022 PQ: 0 ANSI: 5
SCSI device sda: 156301488 512-byte hdwr sectors (80026 MB)
sda: Write Protect is off
sda: Mode Sense: 00 3a 00 00
SCSI device sda: drive cache: write back
SCSI device sda: 156301488 512-byte hdwr sectors (80026 MB)
sda: Write Protect is off
sda: Mode Sense: 00 3a 00 00
SCSI device sda: drive cache: write back
sda: sda1 sda2 < sda5 > sda3
sd 0:0:0:0: Attached scsi disk sda
scsi 1:0:0:0: CD-ROM MATSHITA UJDA755 DVD/CDRW 1.00 PQ: 0 ANSI: 5

As is actually detected, there's a hard drive on the first channel and a DVD-ROM/CDRW on the second channel.

00:00.0 Host bridge: Intel Corporation 82852/82855 GM/GME/PM/GMV Processor to I/O Controller (rev 02)
00:00.1 System peripheral: Intel Corporation 82852/82855 GM/GME/PM/GMV Processor to I/O Controller (rev 02)
00:00.3 System peripheral: Intel Corporation 82852/82855 GM/GME/PM/GMV Processor to I/O Controller (rev 02)
00:02.0 VGA compatible controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 02)
00:02.1 Display controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 02)
00:1d.0 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #1 (rev 03)
00:1d.1 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #2 (rev 03)
00:1d.7 USB Controller: Intel Corporation 82801DB/DBM (ICH4/ICH4-M) USB2 EHCI Controller (rev 03)
00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev 83) 00:1f.0 ISA bridge: Intel Corporation 82801DBM (ICH4-M) LPC Interface Bridge (rev 03)
00:1f.1 IDE interface: Intel Corporation 82801DBM (ICH4-M) IDE Controller (rev 03)
00:1f.3 SMBus: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) SMBus Controller (rev 03)
00:1f.5 Multimedia audio controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller (rev 03)
00:1f.6 Modem: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Modem Controller (rev 03)
01:0a.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
01:0a.1 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
01:0a.2 FireWire (IEEE 1394): Ricoh Co Ltd R5C552 IEEE 1394 Controller (rev 03) 01:0a.3 System peripheral: Ricoh Co Ltd R5C576 SD Bus Host Adapter (rev 01) 01:0a.4 System peripheral: Ricoh Co Ltd R5C592 Memory Stick Bus Host Adapter 01:0c.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
01:0d.0 Ethernet controller: Atheros Communications, Inc. AR5212 802.11abg NIC (rev 01)

Signature · 3 lines
Kevin 'radsaq' Radloff
[email protected]
http://thesaq.com/
#56 lore raw

Ar Gwe, 2006-10-20 am 14:30 -0400, ysgrifennodd Kevin Radloff:

Quoted 6 lines · Linus Torvalds
Quoted 1 line

Ok, it's a week since -rc1, so -rc2 is out there.

A bit behind, but booting still takes ages on my laptop as libata/ata_piix tries to probe a device that isn't there (I reported this previously against -rc1, but got no response):

Probing is somewhat broken in 2.6.18 - something in the core code changed as its upset quite a few drivers at once. One case causes repeated errors and finally detection of an ATAPI device, the other causes repeated errors and then failure when no device is present but takes a few minutes and keeps IRQs locked off for long periods. Both appear to be fallouts from the new EH code.

Alan

#57 lore raw
Quoted 7 lines · Alan Cox

Ar Gwe, 2006-10-20 am 14:30 -0400, ysgrifennodd Kevin Radloff:

Quoted 5 lines · Linus Torvalds
Quoted 1 line

Ok, it's a week since -rc1, so -rc2 is out there.

A bit behind, but booting still takes ages on my laptop as libata/ata_piix tries to probe a device that isn't there (I reported this previously against -rc1, but got no response):

Quoted 6 lines

Probing is somewhat broken in 2.6.18 - something in the core code changed as its upset quite a few drivers at once. One case causes repeated errors and finally detection of an ATAPI device, the other causes repeated errors and then failure when no device is present but takes a few minutes and keeps IRQs locked off for long periods. Both appear to be fallouts from the new EH code.

There are definitely warts related to the new EH stuff, but specifically for SATA + ata_piix, it has been a long hard road of trying various probing mechanisms. Tejun has some patches that revert all the PCS work and rewinds back to original SATA ata_piix probing, in -mm for testing.

If testing feedback proves positive, let's go ahead and fast-track that up the line.

With ata_piix, I would worry more about PCS register follies than core libata. Users can try the module option force_pcs=[0|1|2] to experiment and see if any of the three possibilities improves their boot.

Jeff

#58 lore raw

2.6.19-rc2: known unfixed regressions (v3)

In reply to Linus Torvalds

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

Subject : ppc prep boot hang
References : http://lkml.org/lkml/2006/10/14/58
Submitter : Meelis Roos <[email protected]>
Status : unknown

Subject : X60s: BUG()s, lose ACPI events after suspend/resume
References : http://lkml.org/lkml/2006/10/10/39
Submitter : Martin Lorenz <[email protected]>
Status : unknown

Subject : T60 stops triggering any ACPI events
References : http://lkml.org/lkml/2006/10/4/425

             http://lkml.org/lkml/2006/10/16/262

Submitter : "Michael S. Tsirkin" <[email protected]>
Status : unknown

Subject : CONFIG_X86_VOYAGER=y, CONFIG_SMP=n compile error
References : http://lkml.org/lkml/2006/10/7/51
Submitter : Jesper Juhl <[email protected]>

             commit 7d12e780e003f93433d49ce78cfedf4b4c52adc5

Status : unknown

Subject : CONFIG_X86_VISWS=y, CONFIG_SMP=n compile error
References : http://lkml.org/lkml/2006/10/7/51
Submitter : Jesper Juhl <[email protected]>

             commit 7d12e780e003f93433d49ce78cfedf4b4c52adc5

Status : unknown

Subject : sata-via doesn't detect anymore disks attached to VIA vt6421
References : http://bugzilla.kernel.org/show_bug.cgi?id=7255
Submitter : Thierry Vignaud <[email protected]>
Status : unknown

Subject : SMP x86_64 boot problem
References : http://lkml.org/lkml/2006/9/28/330

             http://lkml.org/lkml/2006/10/5/289

Submitter : [email protected]
Status : submitter was asked to git bisect
result of bisecting seems to be wrong

Subject : unable to rip cd
References : http://lkml.org/lkml/2006/10/13/100
Submitter : Alex Romosan <[email protected]>
Status : unknown

Subject : cpufreq not working on AMD K8
References : http://lkml.org/lkml/2006/10/10/114
Submitter : Christian <[email protected]>

Status : Mark is investigating

Subject : MSI errors during boot (CONFIG_PCI_MULTITHREAD_PROBE)
References : http://lkml.org/lkml/2006/10/16/291
Submitter : Stephen Hemminger <[email protected]>

Status : Greg is working on a fix

Subject : snd-hda-intel <-> forcedeth MSI problem
References : http://lkml.org/lkml/2006/10/5/40

             http://lkml.org/lkml/2006/10/7/164

Submitter : Prakash Punnoor <[email protected]>
Status : patches are being discussed

#60 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

In reply to Adrian Bunk
Quoted 3 lines · Adrian Bunk

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

[...]

Quoted 5 lines

Subject : unable to rip cd
References : http://lkml.org/lkml/2006/10/13/100
Submitter : Alex Romosan <[email protected]>
Status : unknown

FWIW Alex, I just ripped track 2 of a Trace Adkins CD using grip and cdparanoia, then listened to the track in mplayer, while running 2.6.19-rc2. No problem at all. This is however, an older FC2 system, so I'd be inclined to point the finger at cdparanoia's latest version. Mine hasn't been updated for quite a while. I have these installed:

cdparanoia-alpha9.8-20.1
cdparanoia-libs-alpha9.8-20.1
cdparanoia-devel-alpha9.8-20.1

Signature · 7 lines
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2006 by Maurice Eugene Heskett, all rights reserved.
#61 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Quoted 17 lines · Gene Heskett
Quoted 3 lines · Adrian Bunk

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

[...]

Quoted 5 lines

Subject : unable to rip cd
References : http://lkml.org/lkml/2006/10/13/100
Submitter : Alex Romosan <[email protected]>
Status : unknown

FWIW Alex, I just ripped track 2 of a Trace Adkins CD using grip and cdparanoia, then listened to the track in mplayer, while running 2.6.19-rc2. No problem at all. This is however, an older FC2 system, so I'd be inclined to point the finger at cdparanoia's latest version. Mine hasn't been updated for quite a while. I have these installed:

Quoted 3 lines

cdparanoia-alpha9.8-20.1
cdparanoia-libs-alpha9.8-20.1
cdparanoia-devel-alpha9.8-20.1

the system doesn't lock up all the time, but every time i start ripping i get this in syslog:

Oct 22 08:08:16 trinculo kernel: hdc: write_intr: wrong transfer direction!

which didn't use to happen before 2.6.19-rc2 (the lockups did). anyway, i just gave it another try, grip wasn't able to rip the cd but i was able to eject the cd from the drive and then abort execution. i am using cdparanoia that came with grip, and this is a very old version (2.96, the last before they switched to gnome). i also tried with the recent version of cdparanoia but the same thing happens.

--alex--

Signature · 4 lines
| I believe the moment is at hand when, by a paranoiac and active |
|  advance of the mind, it will be possible (simultaneously with  |
|  automatism and other passive states) to systematize confusion  |
|  and thus to help to discredit completely the world of reality. |
#62 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Quoted 26 lines · Alex Romosan
Quoted 17 lines · Gene Heskett
Quoted 2 lines · Adrian Bunk

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

[...]

Quoted 4 lines

Subject : unable to rip cd
References : http://lkml.org/lkml/2006/10/13/100
Submitter : Alex Romosan <[email protected]>
Status : unknown

FWIW Alex, I just ripped track 2 of a Trace Adkins CD using grip and cdparanoia, then listened to the track in mplayer, while running 2.6.19-rc2. No problem at all. This is however, an older FC2 system, so I'd be inclined to point the finger at cdparanoia's latest version. Mine hasn't been updated for quite a while. I have these installed:

Quoted 3 lines

cdparanoia-alpha9.8-20.1
cdparanoia-libs-alpha9.8-20.1
cdparanoia-devel-alpha9.8-20.1

the system doesn't lock up all the time, but every time i start ripping i get this in syslog:

Oct 22 08:08:16 trinculo kernel: hdc: write_intr: wrong transfer
Quoted 9 lines

direction!

which didn't use to happen before 2.6.19-rc2 (the lockups did). anyway, i just gave it another try, grip wasn't able to rip the cd but i was able to eject the cd from the drive and then abort execution. i am using cdparanoia that came with grip, and this is a very old version (2.96, the last before they switched to gnome). i also tried with the recent version of cdparanoia but the same thing happens.

My grip is a bit newer than that, 3.20 I believe.

Quoted 1 line

--alex--

Signature · 7 lines
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2006 by Maurice Eugene Heskett, all rights reserved.
#63 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

In reply to Adrian Bunk
Quoted 11 lines · Adrian Bunk

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

Quoted 6 lines

Subject : CONFIG_X86_VISWS=y, CONFIG_SMP=n compile error
References : http://lkml.org/lkml/2006/10/7/51
Submitter : Jesper Juhl <[email protected]>

             commit 7d12e780e003f93433d49ce78cfedf4b4c52adc5

Status : unknown

Attached patch fixes this problem.

Signature · 2 lines
Andrey Panin		| Linux and UNIX system administrator
[email protected]		| PGP key: wwwkeys.pgp.net
 arch/i386/mach-visws/visws_apic.c       |    7 +---
 include/asm-i386/mach-visws/do_timer.h  |   53 --------------------------------
 include/asm-i386/mach-visws/mach_apic.h |    5 +++
 3 files changed, 8 insertions(+), 57 deletions(-)
arch/i386/mach-visws/visws_apic.c +3 -4
diff -urdpNX /usr/share/dontdiff linux-2.6.19-rc2.vanilla/arch/i386/mach-visws/visws_apic.c linux-2.6.19-rc2/arch/i386/mach-visws/visws_apic.c
@@ -122,7 +122,7 @@ static void end_cobalt_irq(unsigned int 
 	spin_unlock_irqrestore(&cobalt_lock, flags);
 }
 
-static struct hw_interrupt_type cobalt_irq_type = {
+static struct irq_chip cobalt_irq_type = {
 	.typename =	"Cobalt-APIC",
 	.startup =	startup_cobalt_irq,
 	.shutdown =	disable_cobalt_irq,
@@ -159,7 +159,7 @@ static void end_piix4_master_irq(unsigne
 	spin_unlock_irqrestore(&cobalt_lock, flags);
 }
 
-static struct hw_interrupt_type piix4_master_irq_type = {
+static struct irq_chip piix4_master_irq_type = {
 	.typename =	"PIIX4-master",
 	.startup =	startup_piix4_master_irq,
 	.ack =		ack_cobalt_irq,
@@ -167,9 +167,8 @@ static struct hw_interrupt_type piix4_ma
 };
 
 
-static struct hw_interrupt_type piix4_virtual_irq_type = {
+static struct irq_chip piix4_virtual_irq_type = {
 	.typename =	"PIIX4-virtual",
-	.startup =	startup_8259A_irq,
 	.shutdown =	disable_8259A_irq,
 	.enable =	enable_8259A_irq,
 	.disable =	disable_8259A_irq,
include/asm-i386/mach-visws/do_timer.h +0 -53

53 lines not on this page · raw message

include/asm-i386/mach-visws/mach_apic.h +5 -0
diff -urdpNX /usr/share/dontdiff linux-2.6.19-rc2.vanilla/include/asm-i386/mach-visws/mach_apic.h linux-2.6.19-rc2/include/asm-i386/mach-visws/mach_apic.h
@@ -51,6 +51,11 @@ static inline void clustered_apic_check(
 {
 }
 
+static inline int apicid_to_node(int logical_apicid)
+{
+	return 0;
+}
+
 /* Mapping from cpu number to logical apicid */
 static inline int cpu_to_logical_apicid(int cpu)
 {
#65 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Quoted 6 lines · Meelis Roos
Quoted 4 lines

Subject : ppc prep boot hang
References : http://lkml.org/lkml/2006/10/14/58
Submitter : Meelis Roos <[email protected]>
Status : unknown

Seems to be fixed in 2.6.19-rc2+git as of 20061022.

Thanks for the information, I've removed it from my list.

Quoted 1 line

Meelis Roos

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#66 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Hi Adrian,

here is another one:

Subject : [ohci1394 on PPC_PMAC] pci_set_power_state() failure and breaking suspend
References : http://lkml.org/lkml/2006/10/24/13
Submitter : Benjamin Herrenschmidt <[email protected]>

             commit ea6104c22468239083857fa07425c312b1ecb424

Status : looking for answer when to ignore return code of pci_set_power_state

Signature · 3 lines
Stefan Richter
-=====-=-==- =-=- ==---
http://arcgraph.de/sr/
#67 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Quoted 10 lines · Stefan Richter

Hi Adrian,

here is another one:

Subject : [ohci1394 on PPC_PMAC] pci_set_power_state() failure and breaking suspend
References : http://lkml.org/lkml/2006/10/24/13
Submitter : Benjamin Herrenschmidt <[email protected]>

             commit ea6104c22468239083857fa07425c312b1ecb424

Status : looking for answer when to ignore return code of pci_set_power_state

Thanks, added.

Quoted 1 line

Stefan Richter

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#68 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

In reply to Adrian Bunk

Quoting r. Adrian Bunk <[email protected]>:

Quoted 5 lines

Subject: 2.6.19-rc2: known unfixed regressions (v3)

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

Quoted 7 lines

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

skip, hope I didn't trim too much.

Quoted 6 lines

Subject : T60 stops triggering any ACPI events
References : http://lkml.org/lkml/2006/10/4/425

             http://lkml.org/lkml/2006/10/16/262

Submitter : "Michael S. Tsirkin" <[email protected]>
Status : unknown

Just retested with 2.6.19-rc3 - it's still there:
e.g. after I do a full kernel compile, my T60 stops triggering any ACPI events:
tail -f /var/log/acpid does not show anything, even on Fn/F4 which is supposed to be always enabled. Restarting the acpid doesn't do anything either - ACPI starts working again, for a while, only after reboot.

Works fine in 2.6.18 ( + this patch http://lkml.org/lkml/2006/7/20/56).

Signature · 1 line
MST
#69 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Quoted 23 lines · Michael S. Tsirkin

Quoting r. Adrian Bunk <[email protected]>:

Quoted 5 lines

Subject: 2.6.19-rc2: known unfixed regressions (v3)

This email lists some known unfixed regressions in 2.6.19-rc2 compared to 2.6.18 that are not yet fixed Linus' tree.

Quoted 7 lines

If you find your name in the Cc header, you are either submitter of one of the bugs, maintainer of an affectected subsystem or driver, a patch of you caused a breakage or I'm considering you in any other way possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.

skip, hope I didn't trim too much.

Quoted 6 lines

Subject : T60 stops triggering any ACPI events
References : http://lkml.org/lkml/2006/10/4/425

             http://lkml.org/lkml/2006/10/16/262

Submitter : "Michael S. Tsirkin" <[email protected]>
Status : unknown

Quoted 7 lines

Just retested with 2.6.19-rc3 - it's still there:
e.g. after I do a full kernel compile, my T60 stops triggering any ACPI events:
tail -f /var/log/acpid does not show anything, even on Fn/F4 which is supposed to be always enabled. Restarting the acpid doesn't do anything either - ACPI starts working again, for a while, only after reboot.

Works fine in 2.6.18 ( + this patch http://lkml.org/lkml/2006/7/20/56).

Bugzilla.kernel.org, assign it to acpi people...
Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#70 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Quoting r. Pavel Machek <[email protected]>:

Quoted 13 lines
Quoted 13 lines
Quoted 6 lines

Subject : T60 stops triggering any ACPI events
References : http://lkml.org/lkml/2006/10/4/425

             http://lkml.org/lkml/2006/10/16/262

Submitter : "Michael S. Tsirkin" <[email protected]>
Status : unknown

Just retested with 2.6.19-rc3 - it's still there:
e.g. after I do a full kernel compile, my T60 stops triggering any ACPI events:
tail -f /var/log/acpid does not show anything, even on Fn/F4 which is supposed to be always enabled. Restarting the acpid doesn't do anything either - ACPI starts working again, for a while, only after reboot.

Quoted 3 lines
Quoted 1 line

Works fine in 2.6.18 ( + this patch http://lkml.org/lkml/2006/7/20/56).

Bugzilla.kernel.org, assign it to acpi people...

Already done, http://bugzilla.kernel.org/show_bug.cgi?id=7408

Signature · 1 line
MST
#71 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

In reply to Adrian Bunk

Am Sonntag 22 Oktober 2006 14:23 schrieben Sie:

Quoted 5 lines

Subject : snd-hda-intel <-> forcedeth MSI problem
References : http://lkml.org/lkml/2006/10/5/40

             http://lkml.org/lkml/2006/10/7/164

Submitter : Prakash Punnoor <[email protected]>
Status : patches are being discussed

2.6.19-rc3 contains a work-around. Case closed. :-)

Signature · 3 lines
(°=                 =°)
//\ Prakash Punnoor /\\
V_/                 \_V
#72 lore raw

Re: 2.6.19-rc2: known unfixed regressions (v3)

Quoted 8 lines · Prakash Punnoor

Am Sonntag 22 Oktober 2006 14:23 schrieben Sie:

Quoted 5 lines

Subject : snd-hda-intel <-> forcedeth MSI problem
References : http://lkml.org/lkml/2006/10/5/40

             http://lkml.org/lkml/2006/10/7/164

Submitter : Prakash Punnoor <[email protected]>
Status : patches are being discussed

2.6.19-rc3 contains a work-around. Case closed. :-)

Thanks for the confirmation.

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed