]> git.hungrycats.org Git - linux/log
linux
23 years agoupdate version v2.5.4-pre3
Linus Torvalds [Thu, 7 Feb 2002 07:26:17 +0000 (23:26 -0800)]
update version

23 years agoMerge athlon:v2.5/linux
Linus Torvalds [Thu, 7 Feb 2002 07:22:30 +0000 (23:22 -0800)]
Merge athlon:v2.5/linux
into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

23 years agoMerge bk://linuxusb.bkbits.net/linus-2.5
Linus Torvalds [Thu, 7 Feb 2002 07:08:11 +0000 (23:08 -0800)]
Merge bk://linuxusb.bkbits.net/linus-2.5
into athlon.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[PATCH] thread information block
David Howells [Thu, 7 Feb 2002 06:56:27 +0000 (22:56 -0800)]
[PATCH] thread information block

syscall latency improvement

 * There's now an asm/thread_info.h header file with the basic structure
   def and asm offsets in it.

 * There's now a linux/thread_info.h header file which includes the asm
   version and wraps some bitops calls to make convenience functions for
   accessing the low-level flags.

 * The task_struct has had some fields removed (and some flags), and has
   acquired a pointer to the thread_info struct.

 * task_struct's are now allocated on slabs in kernel/fork.c, whereas
   thread_info structs are allocated at the bottom of the stack pages.

 * Some more convenience functions are provided at the end of linux/sched.h to
   access flags in other tasks (these are here because they need to access the
   task_struct).

23 years agoMerge soap.kroah.net:/home/greg/linux/BK/gregkh-2.5
Greg Kroah-Hartman [Thu, 7 Feb 2002 06:28:43 +0000 (22:28 -0800)]
Merge soap.kroah.net:/home/greg/linux/BK/gregkh-2.5
into soap.kroah.net:/home/greg/linux/BK/usb-1

23 years agoAdded driverfs support for the USB subsystem.
Greg Kroah-Hartman [Thu, 7 Feb 2002 06:27:46 +0000 (22:27 -0800)]
Added driverfs support for the USB subsystem.

23 years agotypo fix, fixing a compile time bug due to the change over to the new
Greg Kroah-Hartman [Thu, 7 Feb 2002 06:23:40 +0000 (22:23 -0800)]
typo fix, fixing a compile time bug due to the change over to the new
filesystem definition patches in -pre2 (changes pcihpfs_fs_type to
pcihpfs_type).

23 years agofix compile time bug in usbfs caused by a typo in the superblock changes.
Greg Kroah-Hartman [Thu, 7 Feb 2002 06:15:13 +0000 (22:15 -0800)]
fix compile time bug in usbfs caused by a typo in the superblock changes.

23 years ago[PATCH] tkill ia64 syscall table entry
Chris Wright [Thu, 7 Feb 2002 02:19:35 +0000 (18:19 -0800)]
[PATCH] tkill ia64 syscall table entry

minor nit, but it seems the tkill() patch added a sys_tkill entry without
removing an ia64_ni_syscall entry.

23 years ago[PATCH] BLKGETSIZE64 (bytes not sectors)
Matt Domsch [Thu, 7 Feb 2002 02:19:32 +0000 (18:19 -0800)]
[PATCH] BLKGETSIZE64 (bytes not sectors)

Linus, below is a patch to return the size of a device in bytes, not
in units of 512-byte sectors (which is incorrect per function
definition).  Tim Pepper submitted an equivalent patch for 2.4.x, now
included in 2.4.18-pre4.  Please apply.

23 years ago[PATCH] (6/6) more ->get_sb()
Alexander Viro [Thu, 7 Feb 2002 01:57:18 +0000 (17:57 -0800)]
[PATCH] (6/6) more ->get_sb()

minixfs switched.

23 years ago[PATCH] (5/6) more ->get_sb()
Alexander Viro [Thu, 7 Feb 2002 01:57:14 +0000 (17:57 -0800)]
[PATCH] (5/6) more ->get_sb()

sysvfs and v7 switched.
* super.c: switched to ->get_sb()
* ChangeLog: fixed dates ;-)

23 years ago[PATCH] (4/6) more ->get_sb()
Alexander Viro [Thu, 7 Feb 2002 01:57:09 +0000 (17:57 -0800)]
[PATCH] (4/6) more ->get_sb()

we are done with nodev filesystems - removed dead code in super.c and
fs.h

23 years ago[PATCH] (3/6) more ->get_sb()
Alexander Viro [Thu, 7 Feb 2002 01:57:04 +0000 (17:57 -0800)]
[PATCH] (3/6) more ->get_sb()

coda.

23 years ago[PATCH] (2/6) more ->get_sb()
Alexander Viro [Thu, 7 Feb 2002 01:57:00 +0000 (17:57 -0800)]
[PATCH] (2/6) more ->get_sb()

ncpfs.

23 years ago[PATCH] (1/6) more ->get_sb()
Alexander Viro [Thu, 7 Feb 2002 01:56:55 +0000 (17:56 -0800)]
[PATCH] (1/6) more ->get_sb()

smbfs switched to ->get_sb()

23 years ago[PATCH] Input drivers, step #3
Vojtech Pavlik [Thu, 7 Feb 2002 01:46:57 +0000 (17:46 -0800)]
[PATCH] Input drivers, step #3

This patch updates the input core to the current version, fixing a bunch
of bugs, adding hotplug support and a listing of available input devices
in /proc. The later is very useful for troubleshooting.

Vojtech Pavlik
SuSE Labs

23 years agoMerge master.kernel.org:/home/axboe/linus-merge-2.5
Linus Torvalds [Thu, 7 Feb 2002 01:35:22 +0000 (17:35 -0800)]
Merge master.kernel.org:/home/axboe/linus-merge-2.5
into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

23 years agoMerge http://linux.bkbits.net/linux-2.5
Jens Axboe [Thu, 7 Feb 2002 09:37:19 +0000 (10:37 +0100)]
Merge http://linux.bkbits.net/linux-2.5
into burns.home.kernel.dk:/usr/local/kernel/BK/linux-2.5

23 years ago[PATCH] 2.5: bluetooth compile fix
Robert Love [Thu, 7 Feb 2002 01:32:47 +0000 (17:32 -0800)]
[PATCH] 2.5: bluetooth compile fix

The llseek cleanup contained a typo in bluetooth/hci_vhci.c that
prevents compile.  Thankfully it seems to be the only of this type.

Robert Love

23 years agoRemove deprecated SIOCDEVPRIVATE ioctls from several net drivers. The same capabilit...
Jeff Garzik [Thu, 7 Feb 2002 00:14:36 +0000 (19:14 -0500)]
Remove deprecated SIOCDEVPRIVATE ioctls from several net drivers.  The same capabilities have been accessible via SIOCxMIIxxx ioctls
since their introduction in 2.4.x, and it's DaveM's intention to
eventually remove SIOCDEVPRIVATE during the 2.5.x series.

23 years ago8139too net driver updates:
Jeff Garzik [Thu, 7 Feb 2002 00:07:28 +0000 (19:07 -0500)]
8139too net driver updates:
* Support new clone board.
* Count rx-lost-in-ring events.
* Support ethtool-compatible bitmapped printk message enable number.
* Support MII ethtool helper interface, and associated ioctls.
* Replace some MII-related magic numbers with linux/mii.h constants.
* Add in optional (and hopefully preferred... after testing) code
for resetting RX DMA engine on error.

23 years agofealnx net driver updates:
Jeff Garzik [Wed, 6 Feb 2002 23:58:37 +0000 (18:58 -0500)]
fealnx net driver updates:
* Support MII ethtool helper interface, and associated ioctls.
* Replace some MII-related magic numbers with constants from linux/mii.h.
* Remove deprecated SIOCDEVPRIVATE ioctls, SIOCxMIIxxx ioctl replacements
have been in place since 2.4.x.

23 years ago8139cp net driver updates:
Jeff Garzik [Wed, 6 Feb 2002 23:53:50 +0000 (18:53 -0500)]
8139cp net driver updates:
* Support MII ethtool helper interface, and associated ioctls.
* Remove unused CP_CHIP_VERSION constant.
* Update to-do list in driver header.

23 years agos/suser/capable/ cleanup for the LMC WAN driver.
Jeff Garzik [Wed, 6 Feb 2002 23:49:57 +0000 (18:49 -0500)]
s/suser/capable/ cleanup for the LMC WAN driver.
Contributed by one of the kernel janitors.

23 years agoFix kernel "taint" issue with mii.o helper module, by
Jeff Garzik [Wed, 6 Feb 2002 23:48:26 +0000 (18:48 -0500)]
Fix kernel "taint" issue with mii.o helper module, by
adding MODULE_{AUTHOR,DESC,LICENSE} tags.

23 years agoAdditional media-related ethtool ioctls for net drivers
Jeff Garzik [Wed, 6 Feb 2002 23:45:09 +0000 (18:45 -0500)]
Additional media-related ethtool ioctls for net drivers
ns82830, starfire.

23 years agovia-rhine net driver updates:
Jeff Garzik [Wed, 6 Feb 2002 23:40:57 +0000 (18:40 -0500)]
via-rhine net driver updates:
* Support MII ethtool helper library, and associated ioctls.
* Replace some MII-related magic numbers with constants from linux/mii.h.
* Remove deprecated SIOCDEVPRIVATE ioctl usage, SIOCxMIIxxx is the
supported (and already present) replacement.

23 years agowinbond-840 (tulip clone) net driver updates:
Jeff Garzik [Wed, 6 Feb 2002 23:35:55 +0000 (18:35 -0500)]
winbond-840 (tulip clone) net driver updates:
* Support MII ethtool helper interface, and related ioctls.
* Replace some MII-related magic numbers with constants.

23 years agosundance net driver updates:
Jeff Garzik [Wed, 6 Feb 2002 23:31:49 +0000 (18:31 -0500)]
sundance net driver updates:
* Properly identify 4-port server NIC.
* Add new NIC to PCI ids list.
* Fix multicast bug.
Contributed by the vendor, D-Link.

23 years agoAdd basic ethtool ioctl support to pcmcia net drivers
Jeff Garzik [Wed, 6 Feb 2002 23:28:06 +0000 (18:28 -0500)]
Add basic ethtool ioctl support to pcmcia net drivers
3c589_cs, aironet4500_cs, and fmvj18x_cs.

23 years agoUpdate pci-skeleton net driver:
Jeff Garzik [Wed, 6 Feb 2002 23:24:15 +0000 (18:24 -0500)]
Update pci-skeleton net driver:
Replace MII-related magic numbers with meaningful constants
listed in linux/mii.h.

23 years agoJiffies wrap cleanup for the slip net driver.
Jeff Garzik [Wed, 6 Feb 2002 23:20:57 +0000 (18:20 -0500)]
Jiffies wrap cleanup for the slip net driver.
Contributor: somebody from lkml, Tim ?.

23 years agoepic100 now uses the MII ethtool helper lib, so link against it.
Jeff Garzik [Wed, 6 Feb 2002 23:12:28 +0000 (18:12 -0500)]
epic100 now uses the MII ethtool helper lib, so link against it.

23 years agoepic100 net driver updates:
Jeff Garzik [Wed, 6 Feb 2002 23:03:34 +0000 (18:03 -0500)]
epic100 net driver updates:
* Fix wake-up sequence for some clone boards.
* Support media-related ethtool ioctls, using MII ethtool helper lib.

23 years agoFix several bugs in DL2K Gigabit Ethernet driver, updating to
Jeff Garzik [Wed, 6 Feb 2002 22:57:38 +0000 (17:57 -0500)]
Fix several bugs in DL2K Gigabit Ethernet driver, updating to
vendor revision 1.08.
Contributed by the vendor, D-Link.

23 years agode4x5 net driver endian-related fixes.
Jeff Garzik [Wed, 6 Feb 2002 22:51:21 +0000 (17:51 -0500)]
de4x5 net driver endian-related fixes.
Caught by Ralf and the mips crowd.

23 years agoObvious updates to the alpha port, for recent 2.5.[123] API changes.
Jeff Garzik [Wed, 6 Feb 2002 22:43:39 +0000 (17:43 -0500)]
Obvious updates to the alpha port, for recent 2.5.[123] API changes.
More changes are needed to the task struct/scheduler area before
the alpha port will boot again; this flushes the easy stuff.

23 years agoSet up 'bk ignore' for standard files created in an ia32 build.
Jeff Garzik [Wed, 6 Feb 2002 22:39:29 +0000 (17:39 -0500)]
Set up 'bk ignore' for standard files created in an ia32 build.

23 years agoMerge http://linux.bkbits.net/gkernel-2.5
Jeff Garzik [Wed, 6 Feb 2002 22:04:02 +0000 (17:04 -0500)]
Merge http://linux.bkbits.net/gkernel-2.5
into rum.normnet.org:/spare/work/gkernel-2.5

23 years agoAdd support for basic ethtool ioctls to 3com net drivers,
Jeff Garzik [Wed, 6 Feb 2002 21:01:35 +0000 (16:01 -0500)]
Add support for basic ethtool ioctls to 3com net drivers,
3c501, 3c503, 3c505, 3c507, 3c509, 3c515, 3c523, 3c527.

23 years agokill struct scatterlist address member
Jens Axboe [Wed, 6 Feb 2002 15:32:53 +0000 (16:32 +0100)]
kill struct scatterlist address member

23 years agoFix up tree oddities that Al pointed out.
Linus Torvalds [Wed, 6 Feb 2002 15:17:34 +0000 (07:17 -0800)]
Fix up tree oddities that Al pointed out.

23 years ago[PATCH] (9/9) more ->get_sb() stuff v2.5.4-pre2
Alexander Viro [Wed, 6 Feb 2002 14:34:55 +0000 (06:34 -0800)]
[PATCH] (9/9) more ->get_sb() stuff

More ->get_sb() patches: nfs.  It still acts as nodev - what we want to
(and can) do now is to give it enough smarts to recognize that same tree
might be already mounted.

We are almost done with nodev filesystems by now.

23 years ago[PATCH] (8/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:53 +0000 (06:34 -0800)]
[PATCH] (8/9) more ->get_sb() stuff

More ->get_sb() patches: autofs4

23 years ago[PATCH] (7/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:51 +0000 (06:34 -0800)]
[PATCH] (7/9) more ->get_sb() stuff

More ->get_sb() patches: autofs

23 years ago[PATCH] (6/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:49 +0000 (06:34 -0800)]
[PATCH] (6/9) more ->get_sb() stuff

More ->get_sb() patches: capifs

23 years ago[PATCH] (5/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:47 +0000 (06:34 -0800)]
[PATCH] (5/9) more ->get_sb() stuff

More ->get_sb() patches: openpromfs

23 years ago[PATCH] (4/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:45 +0000 (06:34 -0800)]
[PATCH] (4/9) more ->get_sb() stuff

More ->get_sb() patches: sockfs

23 years ago[PATCH] (3/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:43 +0000 (06:34 -0800)]
[PATCH] (3/9) more ->get_sb() stuff

More ->get_sb() patches: pipefs

23 years ago[PATCH] (2/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:41 +0000 (06:34 -0800)]
[PATCH] (2/9) more ->get_sb() stuff

More ->get_sb() patches: fix for usbdevfs - forgot to remove
DECLARE_FSTYPE() after adding explicit initializer.

23 years ago[PATCH] (1/9) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:34:39 +0000 (06:34 -0800)]
[PATCH] (1/9) more ->get_sb() stuff

More ->get_sb() patches.

shmem/tmpfs switched

23 years agoupdate version number
Linus Torvalds [Wed, 6 Feb 2002 14:24:15 +0000 (06:24 -0800)]
update version number

23 years agoMerge penguin:v2.5/linux
Linus Torvalds [Wed, 6 Feb 2002 14:16:18 +0000 (06:16 -0800)]
Merge penguin:v2.5/linux
into athlon.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[PATCH] Fix mount hash table
Andi Kleen [Wed, 6 Feb 2002 14:13:54 +0000 (06:13 -0800)]
[PATCH] Fix mount hash table

On my 512MB machine with 6 mount points the mount hash table uses 64K.
This patch brings it to a more reasonable size by limiting it to one
page.

23 years ago[PATCH] Fix page cache limit wrapping in filesystems
Andi Kleen [Wed, 6 Feb 2002 14:13:48 +0000 (06:13 -0800)]
[PATCH] Fix page cache limit wrapping in filesystems

Several file systems in tree that nominally support files >2GB set their
s_maxbytes value to ~0ULL. This has the nasty side effect on 32bit machines
that when a file write reaches the page cache limit (e.g. 2^43) it'll silently
wrap and destroy data at the beginning of the file.

This patch changes the file systems in question to fill in a proper limit.

I also have an alternate patch that adds a check for this generically
in super.c, but preliminary comments from Al suggested that he prefered
to do it in the file systems, so it is done this way way.

23 years ago[PATCH] Automatic file-max sizing
Andi Kleen [Wed, 6 Feb 2002 14:13:41 +0000 (06:13 -0800)]
[PATCH] Automatic file-max sizing

The default for NR_FILES of 8192 is far too low for many workloads. This
patch does dynamic sizing for it instead. It assumes file+inode+dentry
are roughly 1K and will use upto 10% of the memory for it.

Also removes two obsolete prototypes.

23 years ago[PATCH] Bootmem speedup
Andi Kleen [Wed, 6 Feb 2002 14:13:35 +0000 (06:13 -0800)]
[PATCH] Bootmem speedup

This patch speeds the bootmem freeing up a bit. Not particularly important,
but helps on some slow simulators where the loop can chew up significant
CPU time.

23 years ago[PATCH] Slab name cleanup - last try
Andi Kleen [Wed, 6 Feb 2002 14:13:29 +0000 (06:13 -0800)]
[PATCH] Slab name cleanup - last try

Hallo Linus,

This is the slab name cleanup you originally requested when the reiserfs
slab name too long boot bug was introduced. It replaces the static
name limit for slab cache names with the requirement for the caller
to supply a stable name. This version addresses all concerns that were given
over earlier versions of the patch.

As I do not feel very strongly about the slab cache name limit and you
also do not seem to be interested in it and the original bug of reiserfs
not booting anymore has been long fixed in your tree with the patch
you originally rejected from me I will not resend this patch after this
try.

Patch against 2.5.4pre1.

-Andi

23 years ago[PATCH] (6/6) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:08:26 +0000 (06:08 -0800)]
[PATCH] (6/6) more ->get_sb() stuff

Now that all FS_SINGLE filesystems are using ->get_sb(), FS_SINGLE
can die.

23 years ago[PATCH] (5/6) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:08:24 +0000 (06:08 -0800)]
[PATCH] (5/6) more ->get_sb() stuff

devfs converted.

23 years ago[PATCH] (4/6) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:08:22 +0000 (06:08 -0800)]
[PATCH] (4/6) more ->get_sb() stuff

driverfs converted.

23 years ago[PATCH] (3/6) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:08:20 +0000 (06:08 -0800)]
[PATCH] (3/6) more ->get_sb() stuff

pcihpfs converted.

23 years ago[PATCH] (2/6) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:08:18 +0000 (06:08 -0800)]
[PATCH] (2/6) more ->get_sb() stuff

devpts converted.

23 years ago[PATCH] (1/6) more ->get_sb() stuff
Alexander Viro [Wed, 6 Feb 2002 14:08:16 +0000 (06:08 -0800)]
[PATCH] (1/6) more ->get_sb() stuff

Fix for idiotic typo in __get_sb_single().

23 years ago[PATCH] (8/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:58:11 +0000 (05:58 -0800)]
[PATCH] (8/8) ->get_sb() switchover

converted fs/block_dev.c ("bdev") to ->get_sb.  use get_sb_nodev() as
helper.

23 years ago[PATCH] (7/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:58:09 +0000 (05:58 -0800)]
[PATCH] (7/8) ->get_sb() switchover

converted ramfs/rootfs to ->get_sb().  Use get_sb_nodev() as helper.

23 years ago[PATCH] (6/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:58:07 +0000 (05:58 -0800)]
[PATCH] (6/8) ->get_sb() switchover

converte usb to ->get_sb()

23 years ago[PATCH] (5/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:58:04 +0000 (05:58 -0800)]
[PATCH] (5/8) ->get_sb() switchover

converte binfmt_misc to ->get_sb()

23 years ago[PATCH] (4/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:58:02 +0000 (05:58 -0800)]
[PATCH] (4/8) ->get_sb() switchover

convert procfs to ->get_sb().  Similar to ext2, except that we use
get_sb_single() instead of get_sb_bdev().  Notice the we _don't_ keep
FS_SINGLE in flags - as the matter of fact, with ->get_sb() FS_SINGLE is
useless.

23 years ago[PATCH] (3/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:58:00 +0000 (05:58 -0800)]
[PATCH] (3/8) ->get_sb() switchover

ext2 switched to ->get_sb().  For that we had
        a) turned ext2_read_super() into ext2_fill_super() (instead of
"return NULL if failed and sb if succeeded" it's "return error value or
0").
        b) added ext2_get_sb(type, flags, dev, data) - all it does is
return get_sb_bdev(type, flags, dev, data, ext2_fill_super);
        c) replaced DECLARE_...() with explicit initializer.
That's it.

23 years ago[PATCH] (2/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:57:58 +0000 (05:57 -0800)]
[PATCH] (2/8) ->get_sb() switchover

get_sb_bdev() et.al. get a new argument (fill_super) and become
exported.

23 years ago[PATCH] (1/8) ->get_sb() switchover
Alexander Viro [Wed, 6 Feb 2002 13:57:56 +0000 (05:57 -0800)]
[PATCH] (1/8) ->get_sb() switchover

OK, here comes the long-promised switchover to ->get_sb().

New method added - ->get_sb(type, flags, dev, data).  At this point we
still keep ->read_super(), but it will go away in the end of series.
->get_sb() is a generalization of get_sb_{bdev,nodev,single}() - as the
matter of fact, these will become common helpers used by filesystems in
their ->get_sb().

23 years agoMerge bk://linuxusb.bkbits.net/linus-2.5
Linus Torvalds [Wed, 6 Feb 2002 13:35:28 +0000 (05:35 -0800)]
Merge bk://linuxusb.bkbits.net/linus-2.5
into athlon.transmeta.com:/home/torvalds/v2.5/linux

23 years agofix multi write and pio issues
Jens Axboe [Wed, 6 Feb 2002 12:51:34 +0000 (13:51 +0100)]
fix multi write and pio issues

23 years agoMerge soap.kroah.net:/home/greg/linux/bk_main/gregkh-2.5
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:54:35 +0000 (23:54 -0800)]
Merge soap.kroah.net:/home/greg/linux/bk_main/gregkh-2.5
into soap.kroah.net:/home/greg/linux/bk_main/usb-3

23 years agoMerge soap.kroah.net:/home/greg/linux/bk_main/gregkh-2.5
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:53:21 +0000 (23:53 -0800)]
Merge soap.kroah.net:/home/greg/linux/bk_main/gregkh-2.5
into soap.kroah.net:/home/greg/linux/bk_main/usb-2

23 years agousb_submit_urb() changes for USB drivers that live outside of the drivers/usb subdire...
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:51:22 +0000 (23:51 -0800)]
usb_submit_urb() changes for USB drivers that live outside of the drivers/usb subdirectory.

23 years agousb_submit_urb() changes.
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:50:45 +0000 (23:50 -0800)]
usb_submit_urb() changes.
Patch done by Oliver Neukum.

23 years agousb_submit_urb() changes for all drivers/usb/serial/*.c drivers
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:50:05 +0000 (23:50 -0800)]
usb_submit_urb() changes for all drivers/usb/serial/*.c drivers

23 years agousb_submit_urb() changes for the drivers/usb/*.c drivers.
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:49:08 +0000 (23:49 -0800)]
usb_submit_urb() changes for the drivers/usb/*.c drivers.
Some driver changes were done by Oliver Neukum and David Brownell.

23 years ago"mem_flags" support for the cris usb host controller.
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:46:41 +0000 (23:46 -0800)]
"mem_flags" support for the cris usb host controller.
These changes have been approved by the CRIS maintainer.

23 years agousb host controllers:
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:45:50 +0000 (23:45 -0800)]
usb host controllers:
- add mem_flags support
- portions of this patch by Oliver Neukum and David Brownell.

23 years agousb core:
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:43:18 +0000 (23:43 -0800)]
usb core:
- initial add of the "int mem_flags" call to the usb_submit_urb() call

23 years agopatch from dimitry for the usb ibmcam driver that does:
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:32:06 +0000 (23:32 -0800)]
patch from dimitry for the usb ibmcam driver that does:
   - Updates the documentation
   - Adds Veo Stingray support
   - Fixes hotplug table dependency upon now-defunct symbol
   - deletes drivers/usb/ibmcam.h

23 years agopatch from Wolfgang Mües <wmues@nexgo.de> for the usb auerswald.c driver:
Greg Kroah-Hartman [Wed, 6 Feb 2002 07:25:26 +0000 (23:25 -0800)]
patch from Wolfgang Mües <wmues@nexgo.de> for the usb auerswald.c driver:
   - Code-Review from Oliver Neukum: remove SMP races.
   - Added some wake_up calls after auerbuf_releasebuf to wake up tasks waiting
     for cp buffers.
   - Change the module count handling to automatic (owner: THIS_MODULE).

23 years agofatfs_syms.c:
Linus Torvalds [Wed, 6 Feb 2002 06:13:11 +0000 (22:13 -0800)]
fatfs_syms.c:
  Al Viro: fix a typo (destory->destroy)

23 years ago[PATCH] reiserfs fix for inodes with wrong item versions (2.5)
Hans Reiser [Wed, 6 Feb 2002 02:57:18 +0000 (18:57 -0800)]
[PATCH] reiserfs fix for inodes with wrong item versions (2.5)

   This is hopefully last bugfix for a bug introduced by struct inode splitting.
   Because of setting i_flags to some value and then cleaning the i_flags
   contents later, on-disk items received wrong item version ob v3.6 filesystems

23 years ago[PATCH] fix for exit_fat_fs()
Alexander Viro [Wed, 6 Feb 2002 02:53:17 +0000 (18:53 -0800)]
[PATCH] fix for exit_fat_fs()

Kudos to OGAWA Hirofumi for finding that one - gcc barfs on the
current code, trying to define an alias to external symbol.

23 years ago[PATCH] missing ->i_zombie in hpux_getdents()
Alexander Viro [Wed, 6 Feb 2002 02:53:15 +0000 (18:53 -0800)]
[PATCH] missing ->i_zombie in hpux_getdents()

hpux_getdents() contains a years-old copy of code from fs/readdir.c.
Switched to use of vfs_readdir(), which does correct locking.

23 years ago[PATCH] more sleep_on() removals
Alexander Viro [Wed, 6 Feb 2002 02:53:13 +0000 (18:53 -0800)]
[PATCH] more sleep_on() removals

Yet another flag/wait_queue pair bites the dust - hfs bitmap_{lock,wait}
was trying to implement a semaphore.

23 years ago[PATCH] death of hfs rename_lock
Alexander Viro [Wed, 6 Feb 2002 02:46:49 +0000 (18:46 -0800)]
[PATCH] death of hfs rename_lock

Useless (lock,waitqueue) pair killed - we have fs-wide
exclusion between renames enforced by VFS, so hfs rename_{lock,wait}
can be removed.

23 years ago[PATCH] hpfs cleanup (switch from sleep_on() to semaphore)
Alexander Viro [Wed, 6 Feb 2002 02:46:47 +0000 (18:46 -0800)]
[PATCH] hpfs cleanup (switch from sleep_on() to semaphore)

Pair (flag, wait_queue) in hpfs replaced with semaphore.

23 years ago[PATCH] 2.5.4-pre1: further llseek cleanup (3/3)
Robert Love [Wed, 6 Feb 2002 02:46:45 +0000 (18:46 -0800)]
[PATCH] 2.5.4-pre1: further llseek cleanup (3/3)

The previous patch did not provide protection for device lseek methods
(drivers/* stuff).  This patch pushes the BKL into each of the remaining
lseek methods -- without them we have a race.

I'd much prefer to have a a better lock to push down than the BKL, but
that will have to wait.

Before you balk at the size, remember patch #2 in this series which
removed much code ;-)

Thanks to Al for assistance, especially a listing of affected files.

Robert Love

23 years ago[PATCH] 2.5.4-pre1: further llseek cleanup (2/3)
Robert Love [Wed, 6 Feb 2002 02:46:42 +0000 (18:46 -0800)]
[PATCH] 2.5.4-pre1: further llseek cleanup (2/3)

This is the second patch of three implementing further llseek cleanups,
against 2.5.4-pre1.  It does not depend on the other patches.

This patch cleans up various code and quite nicely removes much more
code than it adds.  Specifically:

- remove static lseek method which merely reimplements
  the standard no_llseek in the following seven files:
  hci_vhci.c, ite8172.c, nec_vrc5477.c, auerswald.c,
  pipe.c, netlink_dev.c, and socket.c

- remove fs/ufs/file.c::ufs_file_lseek -- Al says it is
  reimplementing generic_file_llseek, so let's use that
  instead (the comment about 32-bit sizes shouldn't be
  an issue, the generic method checks size)

- include smp_lock.h in 3 files missed from previous
  'remove bkl' patch

- Documentation/filesystem/Locking update

Please, apply.

Robert Love

23 years ago[PATCH] 2.5.4-pre1: further llseek cleanup (1/3)
Robert Love [Wed, 6 Feb 2002 02:46:40 +0000 (18:46 -0800)]
[PATCH] 2.5.4-pre1: further llseek cleanup (1/3)

This is the first of three patches implementing further llseek cleanup,
against 2.5.4-pre1.

The 'push locking into llseek methods' patch was integrated into 2.5.3.
The networking filesystems, however, do not protect i_size and can not
rely on the inode semaphore used in generic_file_llseek.

This patch implements a remote_llseek method, which is basically the
pre-2.5.3 version of generic_file_llseek.  Locking is done via the BKL.
When we have a saner locking system in place, we can push it into this
function in lieu.

Ncpfs, nfs, and smbfs have been converted to use this new llseek.

Note this is updated over the previous posted patch.

Robert Love

23 years ago[PATCH] (5/5) beginning of getattr series.
Alexander Viro [Wed, 6 Feb 2002 02:46:38 +0000 (18:46 -0800)]
[PATCH] (5/5) beginning of getattr series.

ia64: switched from sys_new*stat() to vfs_*stat() and cleaned up.

23 years ago[PATCH] (4/5) beginning of getattr series.
Alexander Viro [Wed, 6 Feb 2002 02:46:36 +0000 (18:46 -0800)]
[PATCH] (4/5) beginning of getattr series.

solaris: switched from sys_new*stat() to vfs_*stat() and cleaned up.
Missing LFS check added.

23 years ago[PATCH] (3/5) beginning of getattr series.
Alexander Viro [Wed, 6 Feb 2002 02:46:34 +0000 (18:46 -0800)]
[PATCH] (3/5) beginning of getattr series.

irix_...() switched from sys_new*stat() to vfs_*stat() and cleaned up.
Missing LFS check added

23 years ago[PATCH] (2/5) beginning of getattr series.
Alexander Viro [Wed, 6 Feb 2002 02:46:32 +0000 (18:46 -0800)]
[PATCH] (2/5) beginning of getattr series.

stat(2) variants in arch/* that used to copy inode fields manually
switched to vfs_*stat() and partially cleaned up