Linux 2.4.8-ac5

Linux 2.4.8-ac5

[Alan Cox]

	ftp://ftp.kernel.org/pub/linux/kernel/people/alan/2.4/

		 Intermediate diffs are available from
			http://www.bzimage.org


*
*	This is a fairly experimental -ac so please treat it with care
*

2.4.8-ac5
o	Next batch of IDE driver updates		(Andre Hedrick)
	| qd6580 driver becomes a qd65xx driver
	| 80pin cable detect for serverworks on Dell
	| Mode5 on SIS chipsets
o	Handle ARM mmap were FIRST_USER_PGD_NR is not	(Russell King)
	zero
o	Make the sl82c105.c code common between ARM	(Paul Mackerras)
	and PPC
o	Update cisco hdlc handling in the isdn layer	(Bjoern Zeeb)
o	Add called party number to isdn tty emulation	(Jan Oberlaender,
							 Kai Germaschewski)
o	Eicon warning fix				(Kai Germaschewski)
o	Tiny agp cleanup in severworks code		(Mike Harris)
o	Switch to 2.4.8 nr_free_buffer_pages		(Rik van Riel)
o	Change bootmem bitmap setup			(Rik van Riel)
o	Unlazy the queue movement when we touch
	inactive cache pages (VM balance assumed this)	(Rik van Riel)
o	Update the orinoco drivers			(David Gibson)
o	Update natsemi driver (experimentally anyway)	(Tim Hockin)
o	Update hpt366 blacklists			(Kevin Fleming)
o	Reclaim buffer cache into inactive list when	(Rik van Riel)
	it is too large
o	Documentation tidy ups				(Steven Cole)
o	Switch map_user_kiobuf to use down_read		(Ben LaHaise)

2.4.8-ac4
o	ADFS date/time computation fix			(Russell King)
o	Add ALS120 ident to ns558 joystick		(Filip Van Raemdonck)
o	Make Reiserfs endian and alignment safe		(Jeff Mahoney)
	| Fixes IA64 indirect alignment, S390 alignment
	| Big endian
	| Update inode generator
o	Enable input drivers on ARM			(Russell King)
o	Add intermezzo file system kernel side		(Peter Braam and co)
o	First blocks of ppc64 merge	(Paul Mackerras, Anton Blanchard,
					 Tom Gall and the IBM PPC 64 team)
o	Fix return value bug in mac nvram driver	
o	Make oom killer kill all threads of a set	(Eric Lammerts)


2.4.8-ac3
o	Update ISDN cvs idents				(Kai Germaschewski)
o	Update isdn experimental idents			(Kai Germaschewski)
o	Add Compaq Phoenix 4.06 to irqsafe APM		(Arjan van de Ven)
o	Remove rockridge printk generating too much	(Mikael Pettersson)
	debug output
o	Add a needed include of delay.h			(Arjan van de Ven)
o	Revert access DAC change 			(me)
	| Breaks in some cases
o	ARM documentation, credits, maintainer updates	(Russell King)
o	Atyfb build fixups				(Stelian Pop)
o	Configure.help merges				(Steven Cole)
o	Initial merge of dpt_i2o.c		(Deanna Bonds, Bob Pasteur, 
						 Karen White, Mark Salyzyn)
o	Teach i2o_core to skip dpt cards		(me)
o	Sony Pi driver update				(Stelian Pop)

2.4.8-ac2
o	Fix suspend/resume bugs in eepro100		(Arjan van de Ven)
o	Fix missing spec required delays in PCI PM	(me)
o	Disable PM on eepro100, it breaks even with	(me)
	those fixed
o	Resynchronize Configure.help			(Steven Cole)
o	Further superblock handling updates		(Al Viro)
o	Fix various GPL misreferences		(Andrzej Krzysztofowicz)
o	Updated A20 gate switching code to handle	(Peter Anvin)
	odd (Olivetti etc) and no legacy boxes
o	Fix expand_stack race				(Manfred Spraul)
o	Fix a malloc failure path in ipc		(Manfred Spraul)
o	Revert 2.4.8 scsi_lib change (hangs ide-scsi	(Arjan van de Ven)
	on some drives)
o	Fix binfmt elf strlen->strnlen_user		(me)
o	Add additional checking to binfmt_elf		(me)
	| Last two based on Solar Designers 2.2 work
o	Make pnp_bios dock thread exit when it finds 	(me)
	docking isnt supported
o	Ext3 file system updates			(Andrew Morton,
							 Stephen Tweedie)
o	Merge common components of uhci drivers		(Brad Hards)
o	Remove crud from media Makefile			(Keith Owens)
o	Fix build on Alpha				(David Gilbert)
o	DRM warning/oops fix				(Arjan van de Ven)
o	Fix i810 audio return funny			(Laurent Pinchart)
o	Fix ldm partition checking			(Richard Russon)
o	Agpgart typo fix				(Mike Harris)
o	Revert emu10k changes in 2.4.8, wait until the	(me)
	maintainers actually have debugged code and
	want an update
o	Fix a compile fail case for the 53c700 driver (Andrzej Krzysztofowicz)
o	ARM core and video updates			(Russell King)

2.4.8-ac1
o	Merge Linus 2.4.8
	- Skip VM changes for now
o	Fix sblive build problems			(Rui Sousa)
o	Add Fernando Fuganti to credits			(Fernando Fuganti)
O	Revert printk return change			(Andrew Morton)
o	Add drm-4.0 to mod_subdirs			(Brian Dushaw)
o	Bluez bluetooth updates				(Maksim Krasnyanskiy)
o	Fix serverworks AGP memory leak			(Hugh Dickins)
o	Update DRM 4.1 for Alpha AGPGart support	(Jay Estabrook,
							 Jeff Hartmann)
o	Fix depca crash on unload			(Peter Denison)

2.4.7-ac11
o	Fix dumb bug in the bootflag handling code	(Randy Dunlap)
o	Compaq FC update (makefile clean too)		(Charles White)
o	Add Matt Domsch to the credits			(Matt Domsch)
o	Update Randy Dunlap's contact info		(Randy Dunlap)
o	SA1100 updates					(Russell King)
o	Add PnP support to sf16i			(Ladislav Michl)
o	Acorn drivers update				(Russell King)
o	Fix clashing 'cams' symbol			(Keith Owens)
o	Fix AGP memory leak on serverworks		(Jeff Hartmann)
o	Fix an off by one in shmem.c			(Hugh Dickins)
o	Revert incorrect ATM change			(Mike Westall)
o	DRM makefile fixes				(Keith Owens)
o	Next batch of superblock work			(Al Viro)
o	Fix in2k oopses					(Francois Romieu)
o	Add reboot by SMP reset				(Matt Domsch)
o	Add pm support for laptops that resume the	(Matt Domsch)
	mouse port even if was disabled before..
o	Add DMI quirks for all modern dell systems	(Matt Domsch)
o	Inspiron 4000 needs irqs on for APM		(Arjan van de Ven)
o	YMFPCI fixes, big endian fixes			(Pete Zaitcev)
o	Add winbond W83971D ac_97 codec to tables	(Andrey Panin)
o	pcnet32 oops fix and leak fix			(Paul Gortmaker)
o	Use static inline in sound drivers		(me)
o	Fix DRM build/makefile bugs from merge	(Andrzej Krzysztofowicz)
o	Avoid lp module being pinned down if console	(Tim Waugh)
	enabled

2.4.7-ac10
o	Fix up USB merge mess				(Pete Zaitcev)
	| Fixes a possible USB deadlock
o	IRDA update					(Dag Brattli)
o	Merge DRM for XFree 4.1.x			(XFree86 and others)
o	Update freevxfs idents				(Christoph Hellwig)
o	Fix a bug in access() checks on X_OK with	(Christoph Hellwig)
	DAC ovveride
o	Add another intel bios ident with bad $PIR	(Arjan van de Ven)

2.4.7-ac9
o	Print warnings about buggy 440GX $PIR tables	(Arjan van de Ven)
o	Update ARM softirq code				(Russell King)
o	Compaq FC controller update			(Charles White)
o	Update ARM integrator platform			(Russell King)
o	Miscellaneous ARM fixes				(Russell King)
o	ARM io function updates				(Russell King)
o	Remove duplicate Configure.help items		(Steven Cole)
o	Update ARM shark platform			(Russell King)
o	ARM anakin platform				(Russell King)
o	Allow swap < 2*ram				(Rik van Riel)
o	Set page format bit for scsi-2 tape		(Kai Makisara)
o	Fix compile with shmfs disabled			(Christoph Rohland)
o	RME Hamerfall audio driver			(Guenter Geiger)
o	Further UML fixes				(Jeff Dike)
o	Syncppp fix					(Bob Dunlop)
o	Farsync update					(Bob Dunlop)
o	UML network driver update			(Jeff Dike)
o	Revert aic7xxx makefile changes			(Keith Owens)
o	DaveJ has received enough Rise cpu reports	(Dave Jones)
o	Clean up building without procfs	    (Andrzej Krzysztofowicz)
o	Riscom compile fix			    (Andrzej Krzysztofowicz)

2.4.7-ac8
o	Kill accidental bit of S/390 merge I meant	(Bill Nottingham)
	to skip (hotplug should be working again now)
o	Fix host_info_lock namespace on ieee1394	(Keith Owens)
o	Fix duplicate rio serial init			(Keith Owens)
o	Fix dead init_zoran_cards symbol		(Keith Owens)
o	Don't define EXPORT_SYMTAB in cmpci		(Keith Owens)
o	Don't define EXPORT_SYMTABL in sisfb		(Keith Owens)
o	Fix warnings in ess_solo1 			(me)
o	Fix deadlock in moxa mxser driver		(Christophe Barbé)
o	First of many needed devfs race fixes		(Al Viro)
o	Windows 2000 vfat name mapping fix		(Wolfram Pienkoss)
o	Clean up ubd, CONFIG_IOMEM->CONFIG_MAPPER	(Jeff Dike)
o	Fix a UML crash, make uml devices pluggable	(Jeff Dike)
o	Use page cache in hostfs, fix UML stat64 bits	(Jeff Dike)
o	Complete UML ppc support merge			(Chris Emerson)
o	Complete UML Configure.help			(Bill Stearns)
o	Add rep nop so the poor old Pentium IV doesnt
	go thermal slowdown every long mdelay		(Arjan van de Ven)
o	Fix various invalid Config script items		(Christoph Hellwig)
o	SYS5fs BSD style symlink support (SCO etc)	(Christoph Hellwig)
o	Fix PnPBIOS reporting on io v mem		(Andrey Panin)
o	Fix atyfb compilation problems with vaio bits	(Keith Owens)
o	Make HP support in AMI Megaraid run time 	(Michael Johnson)
o	Fix an ext3 buffer credit accounting bug	(Andrew Morton)
o	Add hardware volume control support to ALi	(Matt Wu)
o	Clean up the above a little for non ALi, fix	(me)
	rmmod crash
o	Adaptec scsi update (6.2.1)			(Justin Gibbs)
	| + gcc 3.0 fixes
o	Handle broken PIV SMP tables			(Maciej Rozycki)
o	Switch to static inline on ARM subtree		(Russell King)
o	SHMfs updates, race fix				(Christoph Rohland)

2.4.7-ac7
o	Import safer bits of 2.4.8pre
	- skipped vm hackery, buffer hacks
	- skipped S/390 (my tree should be newer anyway)
o	Fix duplicates in ldm.h				(Matthew Gardiner)
o	ARM updates					(Russell King)
o	Remove qlogicfc_asm (license clash question + we(Jes Sorensen, me)
	| are shipping qlogic release candidate not final
	| firmware.  Set card to use firmware in flash
o	Example iomem driver + fixes for UML		(Greg Lonnon)
o	Clean up unique UML machine id code		(Henrik Nordstrom)
o	Small UML bug fixes				(Jeff Dike)
o	Fix UML uaccess macro bug			(Jeff Dike)
o	Don't refuse unneccessarily on remount with	(Petr Vandrovec)
	data= option with ext3
o	tdfx driver cleanups				(Paul Mundt)
o	Replace masq crash fixes with corrected and	(Rusty Russell)
	"official" versions.
o	Fix umsdos symlink bug				(Delbert Matlock)
	
2.4.7-ac6
o	Switch JFFS to use completions			(me)
o	Fix wrong unlock in try_to_sync_unused_inodes	(Petr Vandrovec)
o	Update UML to compile again			(Jeff Dike)
o	Clean up all drivers with clashing names for	(me)
	complete, wait_for_completion
o	Some gcc 3.0 warning cleanups
o	Add SCO AFS awareness (read only) to sysv	(Christoph Hellwig)
o	Netfilter crash fixes				(Mark Boucher)
o	Second batch of superblock race/cleanup work	(Al Viro)

2.4.7-ac5
o	Resolve ext3 and superblock change incompat	(me, Al Viro)
	| Fixes hang on journal recovery
o	Fix freevxfs leak				(Andries Brouwer)
o	Add another eepro100 ident			(Matt Wilson)
o	Further tweakes to make rpm based on 
	suggestions by Keith Owens			(Keith Owens, me)
o	Fix ISA dma range check bug in cs89x0		(Paul)
o	Fix ISA dma range check on 3c505		(Paul)
o	Switch md driver to use completions		(Neil Brown)
o	Add more sanity checks to the dmi scanner	(me)
o	Fix hash sign assumptions in reiserfs		(Jeff Mahoney)
	| IMPORTANT: this makes things consistent, it also means
	| that if you have chars > 127 in file names _and_ you are
	| running unsigned char little endian default (probably a tiny tiny
	| number of mips users only) you will need to archive and recreate
	| those files. X86 users are _NOT_ affected.
o	Add the infrastructure for ac97_ops ready for	(me)
	digital audio etc

2.4.7-ac4
o	Fix inode cache shrinkage problems in 2.4.7	(Al Viro)
	| This should cure the problem where it gets 
	| really slow over time. Its not the final fix
o	Make aironet compile again			(Arjan van de Ven)
o	Fix an incredibly stupid i2o_scsi bug causing	(me)
	crashes with the adaptec 2100 and other stuff
o	Fix memory corruption if using gcc 3 and serial	(Thomas Hood)
	probing fails
o	Fix out of memory handling with raid		(Neil Brown)
o	Fix a raid mishandling bug on errors		(Neil Brown)
o	Add promise 20268 software raid card idents	(me)
o	Fix leaktek winview601 problems			(Leandro Lucarella)
o	Merge ntfs 1.1.16				(Anton Altaparmakov)
o	Avoid panic when reiserfs attempts to mount	(Nikita Danilov)
	invalid superblock
o	Error rather than panic on journal replay I/O	(Chris Mason)
	error in reiserfs
o	Update atp870u driver				(Wittman Lee)
o	First batch of superblock handling cleanup	(Al Viro)
o	Restore module oops dumping			(Kai Germaschewski)
o	NFSD update					(Neil Brown)
o	Fix ieee1394 sleep with spinlock held		(Andi Kleen)
o	Add AMD 760-MP to the Agp table			(me)
o	Rip out zillions of duplicated -ESPIPE		(Christoph Hellwig)
	llseek methods for a common one
o	Fix qlogic direction flag handling in 2.4	(Jeff Andre)
o	Clear inode->i-blocks on deletion in reisefs	(Nikita Danilov)
o	Merge rest of S/390 tty driver fixes		(Ulrich Weigand)
o	Finish adapting S/390 to new softirq code	(Ulrich Weigand)
o	Remove accidental duplicate block in Makefile	(Ulrich Weigand)
o	Update DASD drivers				(Ulrich Weigand)
o	Further pnpbios fixes				(Andrey Panin)
o	Switch pnp to use slab not malloc.h	(Arnaldo Carvalho de Melo)
o	Switch parport_cs to slab not malloc.h	(Arnaldo Carvalho de Melo)
o	Further config cleanups				(Steven Cole)
o	Fix make spec with no .version			(Keith Owens)
o	Further ipchains fixes				(Rusty Russell)
o	Add procfs info option for reiserfs stats	(Nikita Danilov)
o	Remove dead code from the reiserfs tree		(Jeff Mahoney)
o	Quota updates					(Jan Kara)
o	FreeVxFS leak fixes, allow block sizes != 1024	(Christoph Hellwig)
o	Improve serial_cs reporting			(Jonathan Corbet)
o	Add v7 fs sanity checks to the sys5 fs code	(Linus Torvalds,
				Christoph Hellwig, Al Viro, Andries Brouwer)

2.4.7-ac3
o	Add "make rpm" target				(me)
	| Remember to install not upgrade kernel rpms
	| as you will want to keep old ones around too
o	Fix FPU emulation breakage			(Brian Gerst)
o	Fix minix subpartition handling			(Andries Brouwer)
o	Add another odd vaio bios to the apm list	(Robert Dunlop)
o	Further Configure.help updates			(Steven Cole)
o	Update usb configure and help texts		(Brad Hards)
o	Fix kmem read loop bug				(Hugh Dickins)
o	Two warning fixes				(Art Haas)
o	Fix missing icmp errors for udp			(Alexey Kuznetsov)
o	Fix 3c59x module load problem			(Hugh Dickins)
o	Update pwc driver				(Nemosoft Unv)
o	Remove old non kernel code from reiserfs tree	(Nikita Danilov)
o	Remove unneeded reiserfsck code from tree	(Nikita Danilov)
o	Replace checks in CONFIG_REISERFS_CHECK with	(Nikita Danilov)
	cleaner macros
o	Add const's correct formatting/typechecking	(Nikita Danilov)
o	Clean up do_reiserfs_warning macro		(Jeff Mahoney)
o	Add Randolph Chung to CREDITS			(Matthew Wilcox)
o	Merge some pa risc tree Configure.help		(Matthew Wilcox)
o	Update the pa risc tree docs 			(Matthew Wilcox)
o	Add pa-risc keyboard drivers	(Debacker Xavier, Marteau Thomas,
					 Djoudi Malek, Philipp Rumpf, 
					 Alex deVries)
o	Update lasi ethernet drivers for pa-risc
o	Correct acenic check for parisc to hppa
o	HPPA port requires pci-setup
o	Add headers for som binary loader		(Matthew Wilcox)

2.4.7-ac2
o	Update motion eye driver, fix stack and hang	(Hugh Dickins)
	problems
o	Fix sigsuspend bug on Alpha			(Richard Henderson)
o	CRIS port updates				(Bjorn Wesen)
o	Add support for Win2K dynamic disk partitions	(Richard Russon,
							 Anton Altaparmakov)
o	ISDN multilink fix				(Karsten Keil)
o	HFC PCI support is now ITU approved		(Werner Cornelius)
o	Update sonypi driver				(Stelian Pop)
o	Fix sleep with irq disabled bug in act2000 	(Kai Germaschewski)
o	Farsync driver cleanups				(Bob Dunlop)
o	Remove surplus tty==NULL checks in isdn tty	(Rob Radez,
							 Kai Germaschewski)
o	Remove sleep with irq disabled in icn driver	(Kai Germaschewski)
o	Remove sleep with irq disabled bug in isdnloop	(Kai Germaschewski)
o	Remove bcopy and duplicate snprintf from eicon	(me)
	driver
o	Remove bzero from eicon driver, use skb_purge	(Armin Schindler)
o	Updated NinjaScsi driver			(YOKOTA Hiroshi)
o	Fix a jffs2 memory leak				(David Woodhouse)
o	Updated bttv driver. Corruption on close fixes	(Gerd Knorr)
o	Audio driver for bt878 tv chip onboard audio	(Gerd Knorr)
o	Correct nvram byte range limit			(Aaron Rendahl)
o	Include linux/smp_lock in shmem			(Clemens)
o	Add DMI/apm support for broken bioses reporting (Marc Boucher)
	battery life wrong endian
o	Fix dumb bug in the sbf code			(me)
o	Fix pnp shutdown oops				(me)

2.4.7-ac1
o	Alpha fixes and updates				(Jay Estabrook)
	- ioremap fixes for AGP on Alpha
	- pci_iommu support for AGP on Alpha
	- AGP arch support code for Alpha UP1x00
	- Fix instruction fault bug with old Jensen/UDB
	  PALcode
o	Further alpha pci iommu cleanups/fixes		(Ivan Kokshaysky)
o	Fix DAC960 for completion change		(Jens Axboe)
o	Fix data corruption on ide tape while reading	(Pete Zaitcev)
	near EOF
o	Switch up_and_exit to complete_and_exit		(David Woodhouse)
o	Fix crash on scsi request alloc failure in sd	(Rasmus Andersen)
o	Handle scsi register failure in ultrastor	(Rasmus Andersen)
o	Clean up ioremap as u32 stuff in ibmtr		(me)
o	Fix max_sector cleanup in paride		(Andrea Arcangeli)
o	
o	Bring IDE floppy up to date with maintainer	(Gunther Mayer)
o	Remove escaped junk from Makefiles		(Christoph Hellwig)
o	Fix cs46xx checks on ioctl calls		(Simon Horman)
o	Update cpqfc driver				(Charles White)
o	NTFS fixes			(Anton Altaparmakov, Rasmus Andersen)
o	Further FATfs updates				(OGAWA Hirofumi)
o	S/390 network driver updates			(Ulrich Weigand + )
o	Core S/390 changes to get it building again	(Martin Schwidefsky)
o	Fix S/390 tree asm blocks to compile with their	(       and
	newest gcc set
o	Update S/390 documentation			(other IBM folks)
o	Update S/390 tape driver			(       "	)
o	Update S/390 console driver			(	"	)
o	Don't do net hotplug during booting		(	"	)
o	Remove tools stuff from S/390 tree		(	"	)
o	Update S/390 irq/softirq handling		(	"	)
o	Add shared kerne support for S/390 VM		(	"	)
o	Update cpqarray driver				(Charles White)
o	Add missing barrier() calls in serial drivers	(me)
o	Fix menuconfig return code on small screen	(Herbert Xu)
o	Small UML fixups				(Jeff Dike)
o	Add COW support to UML block driver		(Jeff Dike)
o	USB network driver updates/fixes		(David Brownell)
o	PnP parsing bug fix				(Andrey Panin)
o	Fix UFS checking of NULL error cases		(Andreas Dilger)
o	Further shmem bits				(Christoph Rohland)
o	Update the mmap changes to handle OSF emulation	(Maciej Rozycki)
o	Fix failed register handling in g_NCR5380	(Rasmus Andersen)
o	Correct errors in devices.txt examples		(Andreas Dilger)
o	Synclink driver update				(Paul Fulghum)
o	Fix scc region requests for latches		(Rob Turk)
o	Fix FB_ACTIVATE_NOW/VBL handling on aty128fb	(Andreas Hundt)
o	Correct CRTC_OFFSET_CNTL on aty128fb		(Andreas Hundt)
o	Add Devfs support to rio500			(Gregory Norris
							 Greg Kroah-Hartmann)
o	3c59x driver updates				(Andrew Morton,
							 Donald Becker)
o	Fix ircomm handling with some mobile phones	(Andrea Arcangeli,
							 Dag Brattli)
o	Clean up resource handling in ibmtr driver	(Rasmus Andersen)
o	Handle out of memory for device alloc in gdth	(Rasmus Andersen)
o	Fix mixer leak in sb driver			(Mike Galbraith)
o	Make printk use vsnprintf			(Andrew Morton)
o	Fix bug in atm_do_connect_dev (bogus EINVALs)	(Germán González)
o	Add Sony DSC-575 to unusual devices usb list	(Denis Benoit)
o	Add gemtek pci radio card driver		(Vladimir Shebordaev)

 
---
Alan Cox <alan@lxorguk.ukuu.org.uk>
Red Hat Kernel Hacker
& Linux 2.2 Maintainer                        Brainbench MVP for TCP/IP
http://www.linux.org.uk/diary                 http://www.brainbench.com

Re: Linux 2.4.8-ac5

[Udo A. Steinberg]

Alan Cox wrote:
> *
> *       This is a fairly experimental -ac so please treat it with care
> *
> 
> 2.4.8-ac5

Hi Alan,

2.4.8-ac5 makes the kpnpbios kernel thread go zombie here every time right
during boot. I know that 2.4.8-ac1 didn't have this problem, but didn't try
-ac2 to -ac4. If you want me to check which -ac release was the last that
got it right, just say and I'll check.

Regards,
Udo.

bash-2.04# ps uxa
USER       PID %CPU %MEM   VSZ  RSS TTY      STAT START   TIME COMMAND
root         1 13.1  0.0   344  188 ?        S    01:21   0:05 init
root         2  0.0  0.0     0    0 ?        Z    01:21   0:00 [kpnpbios <defunct>]
root         3  0.0  0.0     0    0 ?        SW   01:21   0:00 [keventd]
root         4  3.4  0.0     0    0 ?        SW   01:21   0:01 [kapm-idled]
root         5  0.0  0.0     0    0 ?        SWN  01:21   0:00 [ksoftirqd_CPU0]
root         6  0.0  0.0     0    0 ?        SW   01:21   0:00 [kswapd]
root         7  0.0  0.0     0    0 ?        SW   01:21   0:00 [kreclaimd]
root         8  0.0  0.0     0    0 ?        SW   01:21   0:00 [bdflush]
root         9  0.0  0.0     0    0 ?        SW   01:21   0:00 [kupdated]
root        11  0.0  0.0     0    0 ?        SW   01:21   0:00 [khubd]
...

Re: Linux 2.4.8-ac5

[Andrew Morton]

"Udo A. Steinberg" wrote:
> 
> Alan Cox wrote:
> > *
> > *       This is a fairly experimental -ac so please treat it with care
> > *
> >
> > 2.4.8-ac5
> 
> Hi Alan,
> 
> 2.4.8-ac5 makes the kpnpbios kernel thread go zombie here every time right
> during boot.

Interesting that kpnpbiosd has a ppid of zero, whereas keventd, which
is started a few statements later has a ppid of one.  hmmm..

Signalling, exitting from and waiting upon kernel threads is all
screwed up.  Sometimes it's because a userspace process doesn't
reap children.  Sometimes it's because when the userspace process
_does_ wait on the kernel thread, the wait4() code decides it
doesn't want to deliver the SIGCHLD and everything breaks.  Other
times it's because the wait4 code decides the thread can't deliver
the exit signal to init because it isn't set to SIGCHLD. I used
to understand the details of all this but that grey cell died.

I bet this ancient reparent-kernel-thread-to-init patch fixes it.  It always
does.


--- linux-2.4.4-pre3/kernel/sched.c	Sun Apr 15 15:34:25 2001
+++ linux-akpm/kernel/sched.c	Mon Apr 16 20:40:47 2001
@@ -32,6 +32,7 @@
 extern void timer_bh(void);
 extern void tqueue_bh(void);
 extern void immediate_bh(void);
+extern struct task_struct *child_reaper;
 
 /*
  * scheduler variables
@@ -1260,32 +1261,53 @@
 /*
  *	Put all the gunge required to become a kernel thread without
  *	attached user resources in one place where it belongs.
+ *
+ * 	Kernel 2.4.4-pre3, akpm: reparent the caller
+ *	to init and set the exit signal to SIGCHLD so the thread
+ *	will be properly reaped if it exits.
  */
 
 void daemonize(void)
 {
 	struct fs_struct *fs;
-
+	struct task_struct *this_task = current;
 
 	/*
 	 * If we were started as result of loading a module, close all of the
 	 * user space pages.  We don't need them, and if we didn't close them
 	 * they would be locked into memory.
 	 */
-	exit_mm(current);
+	exit_mm(this_task);
 
-	current->session = 1;
-	current->pgrp = 1;
+	this_task->session = 1;
+	this_task->pgrp = 1;
 
 	/* Become as one with the init task */
 
-	exit_fs(current);	/* current->fs->count--; */
+	exit_fs(this_task);		/* this_task->fs->count--; */
 	fs = init_task.fs;
-	current->fs = fs;
+	this_task->fs = fs;
 	atomic_inc(&fs->count);
- 	exit_files(current);
-	current->files = init_task.files;
-	atomic_inc(&current->files->count);
+ 	exit_files(this_task);		/* this_task->files->count-- */
+	this_task->files = init_task.files;
+	atomic_inc(&this_task->files->count);
+
+	write_lock_irq(&tasklist_lock);
+
+	/* Reparent to init */
+	REMOVE_LINKS(this_task);
+	this_task->p_pptr = child_reaper;
+	this_task->p_opptr = child_reaper;
+	SET_LINKS(this_task);
+
+	/* Set the exit signal to SIGCHLD so we signal init on exit */
+	if (this_task->exit_signal != 0) {
+		printk(KERN_ERR "task `%s' exit_signal %d in daemonize()\n",
+			this_task->comm, this_task->exit_signal);
+	}
+	this_task->exit_signal = SIGCHLD;
+
+	write_unlock_irq(&tasklist_lock);
 }
 
 void __init init_idle(void)

Re: Linux 2.4.8-ac5

[Nathan Walp]

[-- Attachment #1: Type: text/plain, Size: 545 bytes --]

> Hi Alan,
> 
> 2.4.8-ac5 makes the kpnpbios kernel thread go zombie here every time right
> during boot. I know that 2.4.8-ac1 didn't have this problem, but didn't try
> -ac2 to -ac4. If you want me to check which -ac release was the last that
> got it right, just say and I'll check.
> 

ac4 does it too, on an old system I have at home...detail available if needed.


-- 
Nathan Walp             || faceprint@faceprint.com
GPG Fingerprint:        ||   http://faceprint.com/
5509 6EF3 928B 2363 9B2B  DA17 3E46 2CDC 492D DB7E


[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: Linux 2.4.8-ac5

[Udo A. Steinberg]

Andrew Morton wrote:
> 
> Interesting that kpnpbiosd has a ppid of zero, whereas keventd, which
> is started a few statements later has a ppid of one.  hmmm..

[...]

> I bet this ancient reparent-kernel-thread-to-init patch fixes it.  It always
> does.

It does indeed. Nice work, Andrew - works like a charm. Formerly the kpnpbios
thread never exited, which is why the problem never showed, now it exits ok.
I've rediffed your patch against 2.4.8-ac5, because I was getting rejects -
in case Alan wants to apply it to his tree.

Regards,
Udo.

--- linux-vanilla/kernel/sched.c        Wed Aug 15 03:19:24 2001
+++ linux-2.4.8-ac/kernel/sched.c       Wed Aug 15 03:06:31 2001
@@ -35,6 +35,7 @@
 extern void timer_bh(void);
 extern void tqueue_bh(void);
 extern void immediate_bh(void);
+extern struct task_struct *child_reaper;
 
 /*
  * scheduler variables
@@ -1227,32 +1228,53 @@
 /*
  *     Put all the gunge required to become a kernel thread without
  *     attached user resources in one place where it belongs.
+ *
+ *     Kernel 2.4.4-pre3, akpm: reparent the caller
+ *     to init and set the exit signal to SIGCHLD so the thread
+ *     will be properly reaped if it exits.
  */
 
 void daemonize(void)
 {
        struct fs_struct *fs;
-
+       struct task_struct *this_task = current;
 
        /*
         * If we were started as result of loading a module, close all of the
         * user space pages.  We don't need them, and if we didn't close them
         * they would be locked into memory.
         */
-       exit_mm(current);
+       exit_mm(this_task);
 
-       current->session = 1;
-       current->pgrp = 1;
+       this_task->session = 1;
+       this_task->pgrp = 1;
 
        /* Become as one with the init task */

-       exit_fs(current);       /* current->fs->count--; */
+       exit_fs(this_task);             /* this_task->fs->count--; */
        fs = init_task.fs;
-       current->fs = fs;
+       this_task->fs = fs;
        atomic_inc(&fs->count);
-       exit_files(current);
-       current->files = init_task.files;
-       atomic_inc(&current->files->count);
+       exit_files(this_task);          /* this_task->files->count-- */
+       this_task->files = init_task.files;
+       atomic_inc(&this_task->files->count);
+
+       write_lock_irq(&tasklist_lock);
+
+       /* Reparent to init */
+       REMOVE_LINKS(this_task);
+       this_task->p_pptr = child_reaper;
+       this_task->p_opptr = child_reaper;
+       SET_LINKS(this_task);
+
+       /* Set the exit signal to SIGCHLD so we signal init on exit */
+       if (this_task->exit_signal != 0) {
+               printk(KERN_ERR "task %s' exit_signal %d in daemonize()\n",
+                       this_task->comm, this_task->exit_signal);
+       }
+       this_task->exit_signal = SIGCHLD;
+
+       write_unlock_irq(&tasklist_lock);
 }
 
 void __init init_idle(void)

Re: Linux 2.4.8-ac5

[Paul]

	Dear Alan;

	Does this include all the 2.4.8 vm mods except
'use once' ?  Could you make a brief comment on anything
signifigant that exists in Linus' tree but not yours?

Thanks;
Paul
set@pobox.com

Re: Linux 2.4.8-ac5

[Rik van Riel]

On Wed, 15 Aug 2001, Paul wrote:

> 	Does this include all the 2.4.8 vm mods except
> 'use once' ?  Could you make a brief comment on anything
> signifigant that exists in Linus' tree but not yours?

Alan's kernel has some additions which are not in Linus'
tree, Linus has a few things not in Alan's tree.

I'm currently trying to merge some of the more obviously
right things from -linus to -ac.

In my opinion, use-once is something of which all the
consequences are not quite clear yet, so I don't think
I'll be porting that over.

I will be looking at changes to reduce the impact of
streaming IO further, but don't expect anything to be
merged before its impact is well understood and tested.

regards,

Rik
--
IA64: a worthy successor to i860.

http://www.surriel.com/		http://distro.conectiva.com/

Send all your spam to aardvark@nl.linux.org (spam digging piggy)

Re: Linux 2.4.8-ac5

[Eli Carter]

"Udo A. Steinberg" wrote:
> 
> Alan Cox wrote:
> > *
> > *       This is a fairly experimental -ac so please treat it with care
> > *
> >
> > 2.4.8-ac5
> 
> Hi Alan,
> 
> 2.4.8-ac5 makes the kpnpbios kernel thread go zombie here every time right
> during boot. I know that 2.4.8-ac1 didn't have this problem, but didn't try
> -ac2 to -ac4. If you want me to check which -ac release was the last that
> got it right, just say and I'll check.

I noticed a [kpnpbios <defunct>] on the 2.4.8-ac3 I built at home.

Eli 
--------------------.     Real Users find the one combination of bizarre
Eli Carter           \ input values that shuts down the system for days.
eli.carter(a)inet.com `-------------------------------------------------

Re: Linux 2.4.8-ac5

[Andrew Morton]

"Udo A. Steinberg" wrote:
> 
> Andrew Morton wrote:
> >
> > Interesting that kpnpbiosd has a ppid of zero, whereas keventd, which
> > is started a few statements later has a ppid of one.  hmmm..
> 
> [...]
> 
> > I bet this ancient reparent-kernel-thread-to-init patch fixes it.  It always
> > does.
> 
> It does indeed. Nice work, Andrew - works like a charm. Formerly the kpnpbios
> thread never exited, which is why the problem never showed, now it exits ok.
> I've rediffed your patch against 2.4.8-ac5, because I was getting rejects -
> in case Alan wants to apply it to his tree.
> 

I think it's not ready.  One possible problem is if a process
starts a kernel thread, then that thread calls daemonize() and
then someone waits on the thread's exit.  With this patch they'll
hang, because the thread switched parents.  I don't know if that
happens at present, but I suspect a better API is to make 
reparent_to_init() a standalone API function, and call that from
places where it's appropriate.

Also, it should do other things like set the thread's UID, scheduling
parameters, etc to something sane (just copy them in from init_task).

So sigh.  I guess I need to finish this off.  But right now
I'm having too much fun writing
prepopulate-the-pagecache-with-stuff-to-make-it-boot-faster.c :)

-