]> git.neil.brown.name Git - history.git/log
history.git
22 years ago[PATCH] remove crt0_fixed.S from m68knommu DragonEngine2 target
Greg Ungerer [Mon, 26 May 2003 02:45:47 +0000 (19:45 -0700)]
[PATCH] remove crt0_fixed.S from m68knommu DragonEngine2 target

Remove crt0_fixed.S from m68knommu DragonEngine2 setup.
No longer required.

22 years agoMerge bk://kernel.bkbits.net/davem/net-2.5
Linus Torvalds [Sun, 25 May 2003 15:17:28 +0000 (08:17 -0700)]
Merge bk://kernel.bkbits.net/davem/net-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

22 years agoMerge bk://kernel.bkbits.net/jgarzik/net-drivers-2.5
Linus Torvalds [Sun, 25 May 2003 15:13:42 +0000 (08:13 -0700)]
Merge bk://kernel.bkbits.net/jgarzik/net-drivers-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

22 years ago[netdrvr tulip] fix bogus merges
Jeff Garzik [Sun, 25 May 2003 20:48:47 +0000 (16:48 -0400)]
[netdrvr tulip] fix bogus merges

22 years ago[netdrvr bonding] minor merge/kbuild fixes
Jeff Garzik [Sun, 25 May 2003 20:47:13 +0000 (16:47 -0400)]
[netdrvr bonding] minor merge/kbuild fixes

22 years ago[netdrvr bonding] add 802.3ad support
Jeff Garzik [Sun, 25 May 2003 20:37:45 +0000 (16:37 -0400)]
[netdrvr bonding] add 802.3ad support

Contributed by Intel, with updates by
Jay Vosburgh @ IBM (bonding maintainer)

22 years ago[PATCH] wireless pcmcia updates
Christoph Hellwig [Sun, 25 May 2003 19:24:25 +0000 (15:24 -0400)]
[PATCH] wireless pcmcia updates

the same patch stil applies..

22 years ago[PATCH] au1000 init cleanups.
Dave Jones [Sun, 25 May 2003 19:23:59 +0000 (15:23 -0400)]
[PATCH] au1000 init cleanups.

Not sure if I incorporated all your feedback on this one last time.
Bug me if not...

22 years ago[PATCH] hamachi PCI DMA fix from 2.4
Dave Jones [Sun, 25 May 2003 19:23:27 +0000 (15:23 -0400)]
[PATCH] hamachi PCI DMA fix from 2.4

Maintainer fix that went into 2.4 last August with the comments
"Get hamachi net driver RX working again.
 Apparently the PCI DMA conversion still has a bug or two left in it..."

22 years ago[PATCH] 3c505 printk levels.
Dave Jones [Sun, 25 May 2003 19:23:04 +0000 (15:23 -0400)]
[PATCH] 3c505 printk levels.

22 years ago[PATCH] xircom init cleanups
Dave Jones [Sun, 25 May 2003 19:22:40 +0000 (15:22 -0400)]
[PATCH] xircom init cleanups

22 years ago[PATCH] fix tlan 64bit check
Dave Jones [Sun, 25 May 2003 19:22:14 +0000 (15:22 -0400)]
[PATCH] fix tlan 64bit check

22 years ago[PATCH] Age old cs89x0 register define 'fixes' ?
Dave Jones [Sun, 25 May 2003 19:21:50 +0000 (15:21 -0400)]
[PATCH] Age old cs89x0 register define 'fixes' ?

Remember these? There never was an outcome as to whether or
not their doing the right thing.  Any complaints from this being
in 2.4 for nearly a year ?

22 years ago[PATCH] Nuke stale comment from bmac
Dave Jones [Sun, 25 May 2003 19:21:24 +0000 (15:21 -0400)]
[PATCH] Nuke stale comment from bmac

Leftovers from before we used memset to clear the struct.

22 years ago[netdvr tulip] nuke stale defines
Jeff Garzik [Sun, 25 May 2003 19:16:05 +0000 (15:16 -0400)]
[netdvr tulip] nuke stale defines

Noticed (indirectly) by davej

22 years ago[PATCH] FEC driver updates to support the ColdFire 5282 CPU (header)
Greg Ungerer [Sun, 25 May 2003 15:11:26 +0000 (08:11 -0700)]
[PATCH] FEC driver updates to support the ColdFire 5282 CPU (header)

For reasons only Motorola will ever know they decided to use a
substantially different register layout for the FEC ethernet device on
the 5282 ColdFire silicon.  This defines an appropriate access structure
for it.

22 years ago[PATCH] allow FEC driver config to be used with ColdFire 5282
Greg Ungerer [Sun, 25 May 2003 15:11:01 +0000 (08:11 -0700)]
[PATCH] allow FEC driver config to be used with ColdFire 5282

The ColdFire 5282 CPU also uses the Motorla FEC ethernet core, allow if
to be selected for that config.

22 years ago[PATCH] ColdFire serial driver support for 5282 ColdFire
Greg Ungerer [Sun, 25 May 2003 15:10:41 +0000 (08:10 -0700)]
[PATCH] ColdFire serial driver support for 5282 ColdFire

This adds support for the ColdFire 5282 to the ColdFire serial driver.

Only a few small changes, notably the interrupt setup is a little
different.  Also removed some dead code.

22 years ago[PATCH] add support for 5282 ColdFire to the ColdFire serial header
Greg Ungerer [Sun, 25 May 2003 15:10:30 +0000 (08:10 -0700)]
[PATCH] add support for 5282 ColdFire to the ColdFire serial header

Add support for the 5282 ColdFire to the common ColdFire serial
driver header. Also added the additional registers of the 5272,
they are used to setup a more accurate baud rate timers

22 years ago[PATCH] rework timer code used for different m68knommu/ColdFire CPU's
Greg Ungerer [Sun, 25 May 2003 15:10:19 +0000 (08:10 -0700)]
[PATCH] rework timer code used for different m68knommu/ColdFire CPU's

Rework the m68knommu/ColdFire/5307 Makefile to build the correct
timer code for each of the different ColdFire CPU types.

22 years ago[PATCH] OProfile: fix d_path() usage
John Levon [Sun, 25 May 2003 14:43:05 +0000 (07:43 -0700)]
[PATCH] OProfile: fix d_path() usage

d_path() can return -ENAMETOOLONG these days. Pass it upstream.

22 years ago[PATCH] OProfile: fix stale comment
John Levon [Sun, 25 May 2003 14:42:54 +0000 (07:42 -0700)]
[PATCH] OProfile: fix stale comment

22 years ago[PATCH] OProfile: timer usage override
John Levon [Sun, 25 May 2003 14:42:45 +0000 (07:42 -0700)]
[PATCH] OProfile: timer usage override

A patch mostly by Will Cohen, adding a parameter to OProfile to
over-ride use of the perfctr hardware. Useful for testing and
a host of other things.

22 years ago[PATCH] OProfile: minimize sample error
John Levon [Sun, 25 May 2003 14:42:35 +0000 (07:42 -0700)]
[PATCH] OProfile: minimize sample error

The code that attempts to reset last_task and in_kernel has a race
against samples appearing during the handling of the buffers, that
causes a small number of mis-attribution of samples. Closing the
window is non-obvious, and not worth it, so we just make it smaller.
Even without the patch, there seem to be few such "bad" samples
because its effects are mitigated on a switch into userspace or
a task switch.

22 years ago[PATCH] OProfile: flush work queue on shutdown
John Levon [Sun, 25 May 2003 14:42:27 +0000 (07:42 -0700)]
[PATCH] OProfile: flush work queue on shutdown

My previous fix was incomplete, we could get the same thing happening
on the init-failure path. Fix that.

22 years ago[PATCH] add pit timer for m68knommu/5282 CPU support
Greg Ungerer [Sun, 25 May 2003 14:37:05 +0000 (07:37 -0700)]
[PATCH] add pit timer for m68knommu/5282 CPU support

This code supports the PIT timer of the 5282 ColdFire.

This new timer device is completely different to the previous ColdFire
timer device.

22 years ago[PATCH] create header support for the ColdFire 5282 PIT timer
Greg Ungerer [Sun, 25 May 2003 14:36:55 +0000 (07:36 -0700)]
[PATCH] create header support for the ColdFire 5282 PIT timer

Create a header file to support the PIT timer of the ColdFire 5282 CPU.
This timer is completely different to all previous ColdFire timers.

22 years ago[PATCH] add support for the DMA of the ColdFire 5282
Greg Ungerer [Sun, 25 May 2003 14:36:45 +0000 (07:36 -0700)]
[PATCH] add support for the DMA of the ColdFire 5282

Add DMA base addresses for the ColdFire 5282 CPU.

22 years ago[PATCH] include SIM header for ColdFire 5282
Greg Ungerer [Sun, 25 May 2003 14:36:35 +0000 (07:36 -0700)]
[PATCH] include SIM header for ColdFire 5282

Modify the common SIM header for all ColdFire CPU's to incldue the
new 5282 ColdFire defines.

22 years ago[PATCH] create SIM header definitions for ColdFire 5282
Greg Ungerer [Sun, 25 May 2003 14:36:24 +0000 (07:36 -0700)]
[PATCH] create SIM header definitions for ColdFire 5282

Create header definitions file to support the ColdFire 5282 CPU.
Unfortunately its register layout and setup is quite different
to previous ColdFire CPU family members.

22 years ago[PATCH] add m68knommu/5282 start up code
Greg Ungerer [Sun, 25 May 2003 14:10:06 +0000 (07:10 -0700)]
[PATCH] add m68knommu/5282 start up code

Start up code for the M5282C3 Motorola ColdFire 5282 evaluation board.

New architecture and board for the m68knommu architecture branch, though
it is not very different to most existing ColdFire eval boards.

22 years ago[PATCH] add m68knommu/5282 config init code
Greg Ungerer [Sun, 25 May 2003 14:09:55 +0000 (07:09 -0700)]
[PATCH] add m68knommu/5282 config init code

This creates specific ColdFire/5282 configuration initialization.

22 years ago[PATCH] add m68knommu/5282 specific Makefile
Greg Ungerer [Sun, 25 May 2003 14:09:45 +0000 (07:09 -0700)]
[PATCH] add m68knommu/5282 specific Makefile

22 years ago[PATCH] moew ColdFire 5282 support
Greg Ungerer [Sun, 25 May 2003 14:09:36 +0000 (07:09 -0700)]
[PATCH] moew ColdFire 5282 support

Add ColdFire 5282 support to common ColdFire header file.

22 years ago[PATCH] m68knommu: add ColdFire 5282 support setup
Greg Ungerer [Sun, 25 May 2003 14:03:23 +0000 (07:03 -0700)]
[PATCH] m68knommu: add ColdFire 5282 support setup

22 years agoMerge davem@nuts.ninka.net:/home/davem/src/BK/net-2.5
David S. Miller [Sun, 25 May 2003 13:39:33 +0000 (06:39 -0700)]
Merge davem@nuts.ninka.net:/home/davem/src/BK/net-2.5
into kernel.bkbits.net:/home/davem/net-2.5

22 years ago[PATCH] m68knommu: ColdFire 5282 support Makefile changes
Greg Ungerer [Sun, 25 May 2003 13:35:15 +0000 (06:35 -0700)]
[PATCH] m68knommu: ColdFire 5282 support Makefile changes

22 years ago[PATCH] m68knommu: add configuration options for ColdFire 5282 support
Greg Ungerer [Sun, 25 May 2003 13:35:05 +0000 (06:35 -0700)]
[PATCH] m68knommu: add configuration options for ColdFire 5282 support

22 years ago[PATCH] m68knommu/5206 check timer irq pending
Greg Ungerer [Sun, 25 May 2003 13:34:55 +0000 (06:34 -0700)]
[PATCH] m68knommu/5206 check timer irq pending

Add a function to allow checking if the timer interrupt is pending for
the m68knommu/5206 CPU.  This is used by the architecture timer code for
microsecond accurate time calculations.

22 years ago[PATCH] m68knommu/5249 check timer irq pending
Greg Ungerer [Sun, 25 May 2003 13:34:46 +0000 (06:34 -0700)]
[PATCH] m68knommu/5249 check timer irq pending

Add a function to allow checking if the timer interrupt is pending for
the m68knommu/5249 CPU.  This is used by the architecture timer code for
microsecond accurate time calculations.

22 years ago[PATCH] m68knommu/5206e check timer irq pending
Greg Ungerer [Sun, 25 May 2003 13:34:37 +0000 (06:34 -0700)]
[PATCH] m68knommu/5206e check timer irq pending

Add a function to allow checking if the timer interrupt is pending for
the m68knommu/5206e CPU.  This is used by the architecture timer code
for microsecond accurate time calculations.

22 years ago[PATCH] arch/* strlcpy conversion
Ben Collins [Sun, 25 May 2003 13:32:08 +0000 (06:32 -0700)]
[PATCH] arch/* strlcpy conversion

Obvious strlcpy conversions in arch/*. In fact, mips and mips64 had an
actual bug in sys_sysmips(). Confirmed with Keith Wesolowski.

22 years ago[FUTEX]: Fix kernel/futex.c warning on 64-bit.
David S. Miller [Sun, 25 May 2003 13:31:21 +0000 (06:31 -0700)]
[FUTEX]: Fix kernel/futex.c warning on 64-bit.

22 years ago[FUTEX]: Fix kernel/compat.c after requeueing futex changes.
David S. Miller [Sun, 25 May 2003 13:30:35 +0000 (06:30 -0700)]
[FUTEX]: Fix kernel/compat.c after requeueing futex changes.

22 years ago[BRIDGE]: Deal with non-linear SKBs in ebtables.
Bart De Schuymer [Sun, 25 May 2003 13:11:11 +0000 (06:11 -0700)]
[BRIDGE]: Deal with non-linear SKBs in ebtables.

22 years ago[XFRM_USER]: Add XFRM_MSG_UPDPOLICY, analogue of SADB_X_SPDUPDATE.
Herbert Xu [Sun, 25 May 2003 13:06:41 +0000 (06:06 -0700)]
[XFRM_USER]: Add XFRM_MSG_UPDPOLICY, analogue of SADB_X_SPDUPDATE.

22 years ago[XFRM_USER]: Wrong use of RTM_BASE.
Krishna Kumar [Sun, 25 May 2003 12:46:42 +0000 (05:46 -0700)]
[XFRM_USER]: Wrong use of RTM_BASE.

22 years ago[IPV6]: Convert /proc/net/tcp6 to seq_file.
Hideaki Yoshifuji [Sun, 25 May 2003 12:44:01 +0000 (05:44 -0700)]
[IPV6]: Convert /proc/net/tcp6 to seq_file.

22 years ago[IPV4/6]: Common TCP procfs infrastructure.
Hideaki Yoshifuji [Sun, 25 May 2003 12:43:02 +0000 (05:43 -0700)]
[IPV4/6]: Common TCP procfs infrastructure.

22 years ago[IPV6]: Convert /proc/net/udp6 to seq_file.
Hideaki Yoshifuji [Sun, 25 May 2003 12:41:37 +0000 (05:41 -0700)]
[IPV6]: Convert /proc/net/udp6 to seq_file.

22 years ago[IPV4/6]: Common UDP procfs infrastructure.
Hideaki Yoshifuji [Sun, 25 May 2003 12:40:29 +0000 (05:40 -0700)]
[IPV4/6]: Common UDP procfs infrastructure.

22 years ago[NET]: strncpy -> strlcpy conversions.
Ben Collins [Sun, 25 May 2003 12:36:42 +0000 (05:36 -0700)]
[NET]: strncpy -> strlcpy conversions.

22 years ago[NETLINK]: Use module_init() in netlink_dev.c
Linus Torvalds [Sun, 25 May 2003 12:34:24 +0000 (05:34 -0700)]
[NETLINK]: Use module_init() in netlink_dev.c

22 years agoMerge bk://kernel.bkbits.net/acme/net-2.5
David S. Miller [Sun, 25 May 2003 12:32:04 +0000 (05:32 -0700)]
Merge bk://kernel.bkbits.net/acme/net-2.5
into nuts.ninka.net:/home/davem/src/BK/net-2.5

22 years ago[PATCH] bio splitting
Jens Axboe [Sun, 25 May 2003 12:28:57 +0000 (05:28 -0700)]
[PATCH] bio splitting

So here it is, easy split support for md and dm. Neil, the changes over
your version are merely:

- Make a global bio split pool instead of requring device setup of one.
  Will waste 8 * sizeof(struct bio_pair) of RAM, but... For 2.6 at least
  it has to be a core functionality.

- Various style changes to follow the kernel guide lines.

22 years ago[NET]: post-sysfs netdev cleanup.
Jeff Muizelaar [Sun, 25 May 2003 12:22:47 +0000 (05:22 -0700)]
[NET]: post-sysfs netdev cleanup.

22 years agoMerge nuts.ninka.net:/home/davem/src/BK/network-2.5
David S. Miller [Sun, 25 May 2003 11:58:35 +0000 (04:58 -0700)]
Merge nuts.ninka.net:/home/davem/src/BK/network-2.5
into nuts.ninka.net:/home/davem/src/BK/net-2.5

22 years ago[PATCH] m68knommu/5407 check timer irq pending
Greg Ungerer [Sun, 25 May 2003 11:41:45 +0000 (04:41 -0700)]
[PATCH] m68knommu/5407 check timer irq pending

Add a function to allow checking if the timer interrupt is pending for
the m68knommu/5407 CPU.  This is used by the architecture timer code for
microsecond accurate time calculations.

22 years ago[PATCH] m68knommu/5272 check timer irq pending
Greg Ungerer [Sun, 25 May 2003 11:41:38 +0000 (04:41 -0700)]
[PATCH] m68knommu/5272 check timer irq pending

Add a function to allow checking if the timer interrupt is pending for
the m68knommu/5272 CPU.  This is used by the architecture timer code for
microsecond accurate time calculations.

22 years ago[PATCH] m68knommu/5307 check timer irq pending
Greg Ungerer [Sun, 25 May 2003 11:41:29 +0000 (04:41 -0700)]
[PATCH] m68knommu/5307 check timer irq pending

Add a function to allow checking if the timer interrupt is pending for
the m68knommu/5307 CPU.  This is used by the architecture timer code for
microsecond accurate time calculations.

22 years ago[PATCH] calculate microsecond offsets for m68knommu/ColdFire timers
Greg Ungerer [Sun, 25 May 2003 11:41:21 +0000 (04:41 -0700)]
[PATCH] calculate microsecond offsets for m68knommu/ColdFire timers

This adds offset calculcation to the m68knommu/ColdFire timers.  Allows
microsecond accuracy for gettimeofday and friends.

22 years ago[PATCH] lock xtime struct in m68knommu/ColdFire timers
Greg Ungerer [Sun, 25 May 2003 11:21:14 +0000 (04:21 -0700)]
[PATCH] lock xtime struct in m68knommu/ColdFire timers

This patch adds appropriate locking around updates of the xtime
structure for the m68knommu ColdFire architecture timers.

22 years ago[PATCH] update m68knommu defconfig
Greg Ungerer [Sun, 25 May 2003 11:21:05 +0000 (04:21 -0700)]
[PATCH] update m68knommu defconfig

This patch updates the defconfig for the m68knommu architecture.

22 years ago[PATCH] update m68knommu link script with 5282 support
Greg Ungerer [Sun, 25 May 2003 11:20:56 +0000 (04:20 -0700)]
[PATCH] update m68knommu link script with 5282 support

This patch does a couple of things to the m68knommu common linker
script:
 - adds support for the 5282 ColdFire CPU
 - fixes broken setup for the Dragon Engine board

22 years ago[PATCH] add dummy VMALLOC_ defines to m68knommu
Greg Ungerer [Sun, 25 May 2003 11:20:46 +0000 (04:20 -0700)]
[PATCH] add dummy VMALLOC_ defines to m68knommu

Add dummy VMALLOC_START and VMALLOC_END defines to m68knommu architectures.
Some code in the proc filesystem breaks without them now.

22 years ago[PATCH] kernel/* strlcpy conversion
Ben Collins [Sun, 25 May 2003 11:19:40 +0000 (04:19 -0700)]
[PATCH] kernel/* strlcpy conversion

22 years ago[PATCH] change get_sb prototype
Andries E. Brouwer [Sun, 25 May 2003 11:04:50 +0000 (04:04 -0700)]
[PATCH] change get_sb prototype

(i) The prototypes for free_vfsmnt(), alloc_vfsmnt(), do_kern_mount()
    so far occurred in several individual c files.  Now they are in
    <linux/mount.h>.

(ii) do_kern_mount() has a third argument name that is typically a
     constant.  It is called with "rootfs", "nfsd", type->name,
     "capifs", "usbdevfs", "binfmt_misc" etc.  So, it should have a
     prototype that expresses this:

do_kern_mount(const char *fstype, int flags, const char *name, void *data);

This makes the ugly cast

-       return do_kern_mount(type->name, 0, (char *)type->name, NULL);
+       return do_kern_mount(type->name, 0, type->name, NULL);

go away. Now do_kern_mount() calls type->get_sb(), so also get_sb()
must have a const third argument. That is what the patch below does.

If I am not mistaken, precisely two filesystems do not treat this
argument as a constant, namely afs and cifs. A separate patch
gives some cleanup there.

22 years ago[PATCH] fix ColdFire 5407 cache flushing
Greg Ungerer [Sun, 25 May 2003 10:33:22 +0000 (03:33 -0700)]
[PATCH] fix ColdFire 5407 cache flushing

This fixes some ColdFire 5407 cache bogosity.  Previous code was pushing
all cache lines and the invalidating all of the cache.  The push should
be enough, and now with underlying fixes the the cache setup registers
it is.  Removed the whole invalidate cycle.

Original patch cam from Allessandro Rubini.

22 years ago[PATCH] fix cache settings for m68knommu 5407 MOTOROLA target
Greg Ungerer [Sun, 25 May 2003 10:33:10 +0000 (03:33 -0700)]
[PATCH] fix cache settings for m68knommu 5407 MOTOROLA target

Correct the wrong cache setup used on the MOTOROLA 5407 based board.
Specifically it enables precise exception mode and write buffering.

Original patch from Alessendra Rubini.

22 years ago[PATCH] fix cache settings for m68knommu 5407 CLEOPATRA target
Greg Ungerer [Sun, 25 May 2003 10:32:59 +0000 (03:32 -0700)]
[PATCH] fix cache settings for m68knommu 5407 CLEOPATRA target

Correct the wrong cache setup used on the CLEOPATRA 5407 based board.
Specifically it enables precise exception mode and write buffering.

Original patch from Alessendra Rubini.

22 years agoMerge bk://kernel.bkbits.net/jgarzik/net-drivers-2.5
Linus Torvalds [Sun, 25 May 2003 10:31:31 +0000 (03:31 -0700)]
Merge bk://kernel.bkbits.net/jgarzik/net-drivers-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

22 years ago[PATCH] module refcounts for airport driver
Paul Mackerras [Sun, 25 May 2003 16:00:29 +0000 (12:00 -0400)]
[PATCH] module refcounts for airport driver

This patch takes out the MOD_INC/DEC_USE_COUNT in the airport (Apple
wireless ethernet) driver.  The driver already does SET_MODULE_OWNER
on the netdevice, so the MOD_INC/DEC_USE_COUNT are unnecessary and
just cause warnings.

22 years ago[PATCH] module owner for ppp_synctty.c
Paul Mackerras [Sun, 25 May 2003 16:00:05 +0000 (12:00 -0400)]
[PATCH] module owner for ppp_synctty.c

This patch fixes ppp_synctty.c (used for doing PPP over some synchronous
serial HDLC links) so that it sets the owner field of the tty line
discipline it exports, rather than using MOD_INC/DEC_USE_COUNT.  This
is more or less from Stephen Hemminger.

Please apply.

Thanks,
Paul.

22 years ago[PATCH] full stop/start on ethtool set speed/duplex/autoneg
Scott Feldman [Sun, 25 May 2003 15:59:39 +0000 (11:59 -0400)]
[PATCH] full stop/start on ethtool set speed/duplex/autoneg

* Cleanup ethtool/mii_ioctl sets of speed/duplex/autoneg by
  stop/set/start driver to ensure sets stick.  Must hold
  xmit_lock around stop/start.

22 years agoMerge redhat.com:/garz/repo/linus-2.5
Jeff Garzik [Sun, 25 May 2003 15:49:53 +0000 (11:49 -0400)]
Merge redhat.com:/garz/repo/linus-2.5
into redhat.com:/garz/repo/net-drivers-2.5

22 years agoMerge bk://kernel.bkbits.net/jgarzik/irda-2.5
Linus Torvalds [Sun, 25 May 2003 10:28:43 +0000 (03:28 -0700)]
Merge bk://kernel.bkbits.net/jgarzik/irda-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

22 years ago[irda] module refcounts in irlan
Rusty Russell [Sun, 25 May 2003 15:38:11 +0000 (11:38 -0400)]
[irda] module refcounts in irlan

22 years agoMerge redhat.com:/garz/repo/linus-2.5
Jeff Garzik [Sun, 25 May 2003 15:25:32 +0000 (11:25 -0400)]
Merge redhat.com:/garz/repo/linus-2.5
into redhat.com:/garz/repo/irda-2.5

22 years ago[PATCH] drivers/ide/ppc/pmac.c compile fix
Benjamin Herrenschmidt [Sun, 25 May 2003 09:30:37 +0000 (02:30 -0700)]
[PATCH] drivers/ide/ppc/pmac.c compile fix

This patch fixes the build of drivers/ide/ppc/pmac.c.

22 years ago[PATCH] Fix mac adbhid driver
Paul Mackerras [Sun, 25 May 2003 09:30:13 +0000 (02:30 -0700)]
[PATCH] Fix mac adbhid driver

22 years ago[PATCH] Fix snd_seq_queue_find_name()
Ben Collins [Sun, 25 May 2003 09:23:35 +0000 (02:23 -0700)]
[PATCH] Fix snd_seq_queue_find_name()

While going through sound/ for strncpy replacing, I came across this
routine:

/* return the (first) queue matching with the specified name */
queue_t *snd_seq_queue_find_name(char *name)
{
        int i;
        queue_t *q;

        for (i = 0; i < SNDRV_SEQ_MAX_QUEUES; i++) {
                if ((q = queueptr(i)) != NULL) {
                        if (strncpy(q->name, name, sizeof(q->name)) == 0)
                                return q;
                        queuefree(q);
                }
        }
        return NULL;
}

I'm _really_ sure that they meant to use strncmp() here instead.

Like this.

22 years ago[PATCH] return valid vma from get_user_pages for non-MMU systems
Greg Ungerer [Sun, 25 May 2003 09:20:57 +0000 (02:20 -0700)]
[PATCH] return valid vma from get_user_pages for non-MMU systems

A couple of updates for the non-MMU memory managment code:

 - declare the overcommit variables
 - return a real vma pointer from get_user_pages (it is used by some code)
 - correct args to vmap()

22 years ago[PATCH] do_mounts.c strlcpy
Ben Collins [Sun, 25 May 2003 09:20:07 +0000 (02:20 -0700)]
[PATCH] do_mounts.c strlcpy

Use strlcpy in root_dev_setup()

22 years ago[PATCH] fs/* conversions for strlcpy
Ben Collins [Sun, 25 May 2003 09:19:58 +0000 (02:19 -0700)]
[PATCH] fs/* conversions for strlcpy

I only converted the cases where it was obvious that the intent was to
truncate on overflow. Lots of places for maxpath/readlink type stuff I
left alone.

22 years ago[PATCH] sound/* strncpy conversion
Ben Collins [Sun, 25 May 2003 09:19:48 +0000 (02:19 -0700)]
[PATCH] sound/* strncpy conversion

This does a lot of cleanup for strncpy->strlcpy, replaces some
sprintf/snprintf's aswell.  There were only two places where things
weren't straight forward.  All-in-all a good cleanup though.

22 years ago[PATCH] A few more strlcpy's for drivers/base/
Ben Collins [Sun, 25 May 2003 09:19:37 +0000 (02:19 -0700)]
[PATCH] A few more strlcpy's for drivers/base/

22 years agoDo a strlcat() to go with the strlcpy().
Linus Torvalds [Sun, 25 May 2003 09:17:12 +0000 (02:17 -0700)]
Do a strlcat() to go with the strlcpy().

22 years ago[PATCH] credits update
Manfred Spraul [Sun, 25 May 2003 08:23:24 +0000 (01:23 -0700)]
[PATCH] credits update

22 years ago[PATCH] kbuild: Get more focus on warnings
Sam Ravnborg [Sun, 25 May 2003 08:20:52 +0000 (01:20 -0700)]
[PATCH] kbuild: Get more focus on warnings

Make the default kernel build less verbose, to make warnings show up
more clearly.

22 years ago[PATCH] namespace.c fix
Andries E. Brouwer [Sun, 25 May 2003 08:20:42 +0000 (01:20 -0700)]
[PATCH] namespace.c fix

Yet another one in the namespace.c series.

The code in graft_tree() used to be correct, but the code

        err = -ENOENT;
        down(&nd->dentry->d_inode->i_sem);
        if (IS_DEADDIR(nd->dentry->d_inode))
                goto out_unlock;

        spin_lock(&dcache_lock);
        if (IS_ROOT(nd->dentry) || !d_unhashed(nd->dentry)) {
...
}
spin_unlock(&dcache_lock);
 out_unlock:

was made incorrect in 2.5.29 when

        err = security_sb_check_sb(mnt, nd);
        if (err)
                goto out_unlock;

was inserted.  It has happened more often that people overlooked
a preexisting setting of err.

22 years ago[PATCH] use update_mmu_cache() in install_page
Andrew Morton [Sun, 25 May 2003 08:15:34 +0000 (01:15 -0700)]
[PATCH] use update_mmu_cache() in install_page

Tell the arch that install_page() has just added a page into the pagetables.

22 years ago[PATCH] tulip warning fix
Andrew Morton [Sun, 25 May 2003 08:15:25 +0000 (01:15 -0700)]
[PATCH] tulip warning fix

From: William Lee Irwin III <wli@holomorphy.com>

Silence a printk warning.

22 years ago[PATCH] ACPI constant overflow fixes
Andrew Morton [Sun, 25 May 2003 08:15:16 +0000 (01:15 -0700)]
[PATCH] ACPI constant overflow fixes

From: William Lee Irwin III <wli@holomorphy.com>

64-bit ACPI fixes

22 years ago[PATCH] arch/i386/kernel/srat.c cast warning fix
Andrew Morton [Sun, 25 May 2003 08:15:07 +0000 (01:15 -0700)]
[PATCH] arch/i386/kernel/srat.c cast warning fix

From: William Lee Irwin III <wli@holomorphy.com>

Fix "cast to pointer from integer of a different size".

22 years ago[PATCH] Change mmu_gathers into per-cpu data
Andrew Morton [Sun, 25 May 2003 08:14:58 +0000 (01:14 -0700)]
[PATCH] Change mmu_gathers into per-cpu data

From: Martin Hicks <mort@wildopensource.com>

Here is a patch that changes mmu_gathers into a per-cpu resource.  It
includes the changes for all arches except ia64.  I've sent a separate patch
to David Mosberger for ia64.

22 years ago[PATCH] two PNP memory leaks
Andrew Morton [Sun, 25 May 2003 08:14:47 +0000 (01:14 -0700)]
[PATCH] two PNP memory leaks

From: Hollis Blanchard <hollis@austin.ibm.com>

Fix two pnp error-path memory leaks, caught by Stanford memory leak checker
circa 2.5.48.

22 years ago[PATCH] fix init/do_mounts_rd.c memory leak
Andrew Morton [Sun, 25 May 2003 08:14:37 +0000 (01:14 -0700)]
[PATCH] fix init/do_mounts_rd.c memory leak

From: Hollis Blanchard <hollis@austin.ibm.com>

Another potential memory leak the Stanford checker caught at 2.5.48: while
closing and opening floppy disks, buf could be allocated and never freed.

22 years ago[PATCH] fix drivers/net/ewrk.c memory leak
Andrew Morton [Sun, 25 May 2003 08:14:29 +0000 (01:14 -0700)]
[PATCH] fix drivers/net/ewrk.c memory leak

From: Hollis Blanchard <hollis@austin.ibm.com>

Hi, this was caught by the Stanford memory leak checker a while ago
(2.5.48). If the tmp_stats allocation fails, tmp is not being freed.

22 years ago[PATCH] Better fix for ia32 subarch circular dependencies
Andrew Morton [Sun, 25 May 2003 08:14:18 +0000 (01:14 -0700)]
[PATCH] Better fix for ia32 subarch circular dependencies

From: john stultz <johnstul@us.ibm.com>

This is a rework of John's recent change which resolved a circular include
dependency: a function in mach_apic.h requires hard_smp_processor_id() and
hard_smp_processor_id() requires macros from mach_apic.h

So this patch (against bk-current) reverts the previous, and fixes the same
circular dependency in a much cleaner way, by moving a piece of the circular
chain into its own .h file, rather then removing hard_smp_processor_id() and
accessing the apic by hand.

22 years ago[PATCH] rd.c: separate queue per disk
Andrew Morton [Sun, 25 May 2003 08:14:07 +0000 (01:14 -0700)]
[PATCH] rd.c: separate queue per disk

From: Maneesh Soni <maneesh@in.ibm.com>

Provides a separate request queue for each ramdisk instance.  Without this,
the kernel oopses when the block later tries to unregister the same set of
kobject things multiple times.  This makes rd.c consistent with all other
"disk" devices.