[CFT] 2.6.9-rc4-bird1 patchset

[CFT] 2.6.9-rc4-bird1 patchset

[viro]

	Yes, folks, it's YATree.  It is supposed to act as a staging area
for the stuff like sparse cleanups, etc.

	Right now it's mostly about iomem annotations and fixes; it gets
drivers/net more or less back to sanity and takes a swipe at other parts
of drivers/*.  Other stuff that will eventually get in there includes e.g.
endianness annotations and fixes.  It builds with allmodconfig on x86,
amd64, alpha, sparc32, sparc64 and ppc.  *Note*: right now a bunch of
modules are broken on ppc since it doesn't have ioread*/iowrite* implemented.

	Currently patch is on
ftp.linux.org.uk/pub/people/viro/patch-2.6.9-rc4-bird1.bz2.  Splitup is
in ftp.linux.org.uk/pub/people/viro/patchset/.  Patch is against 2.6.9-rc4;
the first chunk in splitup is bk-net snapshot circa Oct 12, the rest is
the series proper.

	Testers are welcome.  If you have patches in that direction, feel
free to send them for review/inclusion into patchset.

	Hopefully, updated versions of patchset will be posted more or less
daily; for now - based on 2.6.4-rc9, once we get 2.6.9 and things start
moving again - on bk snapshots.

	Current contents:
Base patches:
bk-net.patch	jgarzik's tree
bk-net.fixes	compile fixes - missing include in typhoon.c + use of -mm-only
		function in tulip.

Infrastructure patches [not for merge]
kconfig		teaches allmodconfig to pin a set of options in given state
		[NB: rz has another variant; this is just a temporary]
CHECKFLAGS	allows to add stuff to CHECKFLAGS from command line
disable-DI	disables CONFIG_DEBUG_INFO for test builds

Series proper:
alpha-writeq-fix	alpha compile fix (writeq/readq)
		 some drivers do ifdef on writeq; alpha has it as inline
		 function, so they
		 get confused.
sparc-kconfig-fixes	sparc32 kconfig fixes
		 CONFIG_VT should select CONFIG_INPUT, parport, oss and serial
		 are broken.
alpha-io_remap_page_range compile fix for alpha io_remap_page_range
bw2		fixes for drivers/video/bw2.c
sparc-atomic-UP	compile fix for UP sparc32 with DEBUG_SPINLOCK enabled
gadget-ether	sparc32 compile fixes for drivers/usb/gadget/ether.c
amd64-ioremap	amd64 io.h annotations
amd64-get_user	amd64 uaccess.h annotations
qla2xxx		qla2xx iomem annotations
ppc-ioremap	ppc io.h annotations
s2io		s2io iomem annotations
sparc64-io	sparc64 io.h annotations
cycx		cyclom iomem annotations
		__iomem added where needed in cyclom code
		cycx_setup() gets physical address as an explicit argument
		instead of abusing hw->dpmbase
sparc64-ioremap	added typechecking to sparc64 ioremap
		made sparc64 ioremap an inlined function, fixed a bug in
		sbus/char/cpwatchdog.c found by that.
ppc-uaccess	ppc uaccess.h annotations
aacraid		aacraid iomem annotations
aic-trivial	aic7xxx iomem annotations
aic-ioremap	aic7xxx ioremap cleanup
		ioremap does the right thing when it sees a non-aligned
		address; no need duplicate that in driver.
qla1820		qla1820 iomem annotionats
syscall		pci/syscall.c __user annotations
msi		msi iomem annotations
cpq-hotplug	cpqphp iomem annotations
hotplug-misc	the rest of pci/hotplug iomem annotations
fusion		message/fusion iomem annotations
cycx-64bit	64bit fix in cycx_x25.c
		comparing u32 with ~0UL is wrong
memcpy_io	consistent prototypes for memcpy_..io()
		x86_64 and sparc64 switched to usual prototypes for
		memcpy_toio() et.al.
hd6457		hd6457x iomem annotations
dscc4		dscc4 iomem annotations
netdev-trivial	bunch of trivial iomem annotations in drivers/net
rrunner		rrunner iomem annotations
amd8111e	amd8111e iomem annotations
r8169		r8169 iomem annotations
via-velocity	via-velocity iomem annotations
via-rhine	via-rhine iomem annotations, switch to io{read,write}
tulip		tulip iomem annotations, switch to io{read,write}
winbond		winbond840 iomem annotations, switch to io{read,write}
forcedeth	forcedeth iomem annotations
sundance	sundance iomem annotations, switch to io{read,write}
yellowfin	yellowfin iomem annotations, switch to io{read,write}
ac3200		ac3200 iomem annotations and fixes
		annotated, killed isa_... uses by making ioremap()
		unconditional, fixed the use of isa_... on already remapped
		address.
hp100		hp100 iomem annotations
ne3210		ne3210 iomem annotations
starfire	starfire iomem annotations
3c507		killed isa_... in 3c507
		switched to ioremap() + normal read../write..
depca		depca iomem annotations
hamachi		hamachi iomem annotations
3c359		3c359 iomem annotations
e2100		e2100 iomem annotations and fixes
		added missing ioremap(); driver was using readw() et.al. on
		non-remapped addresses.
ibmtr		beginning of ibmtr iomem annotations
		the easy parts of ibmtr annotations, there will be another
		patch dealing with the rest of it [yet to be merged]
lanstreamer	lanstreamer iomem annotations
lanstreamer-fix	lanstreamer fix
		copy from on-stack array is memcpy(), not memcpy_fromio()
arlan-priv	netdev_priv() in arlan
arlan		arlan iomem annotations and cleanups
		iomem annotations + couple of bad implementations of offsetof()
		replaced with the real thing.
netwave-priv	netdev_priv() in netwave_cs
netwave		netwave iomem annotations
pcmcia		net/pcmcia iomem annotations
olympic		olympic_open() cleanup and fixes
		fixed direct dereference of iomem pointers and plugged a leak.

[CFT] 2.6.9-rc4-bird2 patchset

[viro]

	Next version of the patchset.

News:
	* sparse noise in drivers/net/* is practically gone (e.g. i386 is
down to 84 lines of sparse output in there, most of them unrelated to iomem;
that's down from ~10000).
	* a bunch of drivers/net patches had been ACKed for bk-net merge,
so hopefully the next revision will shrink.
	* more iomem patches outside of drivers/net; we are down to 469 gcc
warnings on i386 allmodconfig, more than a half of them from sound/*.  Other
platforms are, surprisingly, not much worse.
	* merged ppc iomap patch from benh, so ppc32 breakage is gone.

 	Currently patch is on
ftp.linux.org.uk/pub/people/viro/patch-2.6.9-rc4-bird2.bz2.  Splitup is
in ftp.linux.org.uk/pub/people/viro/patchset/.  Patch is against 2.6.9-rc4;
the first two chunks in splitup are bk-net snapshot circa Oct 12 and ppc
iomap patch from benh, the rest is the series proper.

 	Testers are welcome.  If you have patches in that direction, feel
free to send them for review/inclusion into patchset.
 
	Next revision will hopefully merge bk-scsi (and sanitize scsi-related
patches).

	Added chunks:
ibmtr2		ibmtr annotations - the rest
		the rest of annotations and cleanup: ->sram_virt abuse removed,
		we have separate ->sram_phys now (not remapped) and keep
		->sram_virt an iomem pointer.
skfp		skfp iomem annotations, switch to io{read,write}
hermes		wireless iomem annotations and fixes, switch to io{read,write}
		hermes.c switched to ioread/iowrite from homegrown analogs,
		its users updated.  Fixed direct dereferencing of ioremapped
		memory in orinoco_plx.
fealnx		fealnx iomem annotations, switch to io{read,write}
lne390		lne390 iomem annotations and fixes
		annotated, killed isa_... uses by making ioremap()
		unconditional, fixed the use of isa_... on already remapped
		address.
includes	a couple of missing includes of asm/irq.h
		disable_irq() needs asm/irq.h and not everyone who needs it
		gets it included indirectly.
wavelan_cs	wavelan_cs iomem annotations
airo		airo iomem annotations
miri_sbus	miri_sbus iomem annotations and fixes
		missing sbus_readl()
depca-noise	depca removal of bogus virt_to_bus() uses
mace1		mace iomem annotations - trivial part
disk_on_chip	disk_on_chip annotations and fixes
		iomem annotations and fix for a bug they'd caught in
		doc2001plus.c (wrong order of arguments in WriteDOC() call).
if_ppp		__user annotations in if_ppp
		annotated ioctl structure
moxa		moxa iomem annotations
sx		sx iomem annotations and fixes
		missing readb(), check of 64Kb alignment of physical address
		done to remapped one.
skystar2	skystar2 iomem annotations
kyro		video/kyro iomem annotations
aty		video/aty iomem annotations
riva		video/riva iomem annotations
video		misc drivers/video iomem annotations
teles		teles0 and telespci iomem annotations
isurf		isurf iomem annotations
ipr		ipr iomem annotations
ips		ips iomem annotations
megaraid	megaraid iomem annotations
nsp32		nsp32 iomem annotations
dpt		partial dpt_i2o iomem annotations
qlogicisp	qlogicisp iomem annotations
aic-old		aic7xxxold iomem annotations
3w-9xxx		3w-9xxx iomem annotations

[CFT] 2.6.9-bk4-bird1 patchset

[viro]

News:
	* based on 2.6.9-bk4 now
	* a bunch of drivers/net/* stuff got merged, patchset based on more
recent bk-net.

	Currently patch is on
ftp.linux.org.uk/pub/people/viro/patch-2.6.9-bk4-bird1.bz2.  Splitup is
in ftp.linux.org.uk/pub/people/viro/patchset/.
 
 	Added chunks:
ncr		ncr.../sym... iomem annotations
alpha-numa	CONFIG_NUMA is broken on alpha right now
ppc-fix		missing externs in ppc irq.c
netdev-fixes2	compile fixes in typhoon.c and ne2k-pci.c

[CFT] 2.6.9-rc2-bk1-bird1 patchset

[Al Viro]

News:
 	* based on 2.6.10-rc2-bk1 now
	* a bunch of patches got merged - some into Linus' tree, some into
bk-net
	* iomem patches in sound/oss
	* eicon patchset added; that one is *not* for merge until maintainer
ACKs the sucker
	* more __user annotations for recently merged stuff (folks, please
at least try to run make C=2 after your patches and look for regressions).
	* misc iomem annotations and fixes all over the place in drivers/*

 	Currently patch is on
ftp.linux.org.uk/pub/people/viro/patch-2.6.10-rc2-bk1-bird1.bz2.  Splitup is
in ftp.linux.org.uk/pub/people/viro/patchset/.

	Added chunks:

X6	alpha-numa	CONFIG_NUMA is broken on alpha right now
X11	emu10k-ioctl	CONFIG_EMU10K1 is broken on a shitload of platforms
U573	pc300		portability fixes for pc300
X15	xircom		fix for botched MODULE_PARM conversion in xircom
U574	i2o		i2o iomem annotation
U575	cpq		cpqphp_nvram iomem annotation
U576	misc_user	trivial __user annotations
U577	init		more C99 initializers
U578	w1		drivers/w1 iomem annotations
U579	isdn		more trivial isdn iomem annotations
U580	pcilynx		pcilynx iomem annotations
U581	media		trivial drivers/media iomem annotations
U582	oss		sound/oss iomem annotations
U583	ad1889		ad1889 iomem annotations and leak fixes
U584	64bit		misc portability fixes
U585	sound_pci	sound/pci iomem annotations
U586	isdn_fix	isdn/pcbit iomem fix
U587	zoran_fix	zoran iomem fix
U588	imsttfb		imsttfb iomem annotations
U589	tgafb		tgafb iomem annotations
U590	i386_uaccess	i386 uaccess.h annotations
X20	config-rio	marks rio broken
EIC1	eicon_getword	distinguish between iomem and normal memory access
# adds new helpers ({GET,PUT}_{WORD,DWORD}) parallel to READ_WORD et.al.,
# but used on normal memory instead of iomem.  Instances of READ_WORD
# and friends that are applied to vmalloc'ed memory and local variables
# switched to new helpers.
EIC2	eicon_iomem	eicon iomem annotations
EIC3	eicon_readb	eicon byte iomem access fixes
# new helpers - READ_BYTE() and WRITE_BYTE().  On Linux they are needed, since
# direct access to iomem pointers is not allowed (and does not work on some
# architectures).
# memcpy() to/from iomem is not allowed on Linux (same story). Proper
# primitives are memcpy_toio() and memcpy_fromio().  Several places
# misused memcpy(); switched to memcpy_{to,from}io().
EIC4	eicon_fix	s_4bri.c bugfix (bogus &)
# s_4bri.c has a nasty typo - bogus & in qBri_cpu_trapped().
# We have
#	void *base;
# followed by
#	regs[0] = READ_DWORD((&base + offset) + 0x70);
# which is *NOT* what is meant there - instead of access to memory pointed to
# by base we get access to _stack_ at some offset from the place where the
# local variable base lives.