]> git.hungrycats.org Git - linux/log
linux
23 years ago[PATCH] fix devfs_mk_dir prototype
Christoph Hellwig [Wed, 23 Apr 2003 11:28:03 +0000 (04:28 -0700)]
[PATCH] fix devfs_mk_dir prototype

Return an error code instead of a devfs_handle_t.  The handle isn't
useful for anything and the !CONFIG_DEVFS_FS stub in fact returned
NULL which made it entirely useless.  Thus only one driver is actually
checking the retval in the current tree..

23 years ago[PATCH] don't include devfs_fs_kernel.h in global headers
Christoph Hellwig [Wed, 23 Apr 2003 11:27:54 +0000 (04:27 -0700)]
[PATCH] don't include devfs_fs_kernel.h in global headers

Now that devfs_handle_t is gone from all structs there is no
reason to include it in headers.

Fix the fallout by including previously implicit headers and fixing
the drivers that didn't include devfs_fs_kernel.h explicitly.

23 years ago[PATCH] don't use mem_map_reserve/mem_map_unreserve
Christoph Hellwig [Wed, 23 Apr 2003 11:27:45 +0000 (04:27 -0700)]
[PATCH] don't use mem_map_reserve/mem_map_unreserve

They are obsfucating aliases for SetPageReserved/ClearPageReserved.
And once they're gone we can nuke <linux/wrapper.h>

23 years ago[PATCH] kill LOCAL_END_REQUEST
Christoph Hellwig [Wed, 23 Apr 2003 11:27:34 +0000 (04:27 -0700)]
[PATCH] kill LOCAL_END_REQUEST

And uninline end_request - it's calling to many functions to be useful
inline.

23 years ago[PATCH] split initrd from ramdisk driver
Christoph Hellwig [Wed, 23 Apr 2003 11:27:26 +0000 (04:27 -0700)]
[PATCH] split initrd from ramdisk driver

They don't have any code in common, so the initrd support can
go into a separate file and not require ramdisk support.

Lots of ifdefs gone and smaller kernel images for initrd users.

23 years agoSince "apply_alternatives()" also runs at module load time it must
Linus Torvalds [Wed, 23 Apr 2003 11:21:46 +0000 (04:21 -0700)]
Since "apply_alternatives()" also runs at module load time it must
not be marked __init.

Noted by Petr Vandrovec

23 years ago[PATCH] tty cleanups (12/12)
Alexander Viro [Wed, 23 Apr 2003 11:19:54 +0000 (04:19 -0700)]
[PATCH] tty cleanups (12/12)

* we allow tty_driver to cover more than 256 devices
* pty.c cleaned up - now we only one driver for UNIX98 masters and
only one driver for UNIX98 slaves, so a lot of ugliness can be killed.
* get_tty_driver() became an analog of get_gendisk() - it does
a lookup by device number and gives (pointer to tty_driver,index).
* registration/unregistration of tty_driver updated
* /proc/tty/drivers code updated (now one structure can be responsible
for several lines)

23 years ago[PATCH] tty cleanups (11/12)
Alexander Viro [Wed, 23 Apr 2003 11:19:46 +0000 (04:19 -0700)]
[PATCH] tty cleanups (11/12)

tty->device switched to dev_t
There are very few uses of tty->device left by now; most of
them actually want dev_t (process accounting, proc/<pid>/stat, several
ioctls, slip.c logics, etc.) and the rest will go away shortly.

23 years ago[PATCH] tty cleanups (10/12)
Alexander Viro [Wed, 23 Apr 2003 11:19:36 +0000 (04:19 -0700)]
[PATCH] tty cleanups (10/12)

Preparations to cleanup:
* call of get_tty_driver() moved from init_dev() to its callers
* instead of kdev_t dev we pass struct tty_struct *driver and int index

23 years ago[PATCH] tty cleanups (9/12)
Alexander Viro [Wed, 23 Apr 2003 11:19:27 +0000 (04:19 -0700)]
[PATCH] tty cleanups (9/12)

* drivers/char/rio/* supports up to 4 boards, each with up to 128
lines.  It used to share termios for 1st/3rd and 2nd/4th boards,  Fixed.
* cleanups and kdev_t removals - we pass tty instead of tty->device
in a couple of helper functions and instead of comparisons on major(tty->device)
we check where does tty->driver point to.

23 years ago[PATCH] tty cleanups (8/12)
Alexander Viro [Wed, 23 Apr 2003 11:19:18 +0000 (04:19 -0700)]
[PATCH] tty cleanups (8/12)

* generic_serial.c typo fix (->driver used instead of correct
->driver_data)
* tubio cleaned up

23 years ago[PATCH] tty cleanups (7/12)
Alexander Viro [Wed, 23 Apr 2003 11:19:08 +0000 (04:19 -0700)]
[PATCH] tty cleanups (7/12)

sanitized driver->driver_name initialization and use

23 years ago[PATCH] tty cleanups (6/12)
Alexander Viro [Wed, 23 Apr 2003 11:18:59 +0000 (04:18 -0700)]
[PATCH] tty cleanups (6/12)

tty->tty_index added; we initialize it with minor(tty->device) -
tty->driver->minor_start.  Majority of remaining tty->device uses had
that form and are switched to use of tty->index.

23 years ago[PATCH] tty cleanups (5/12)
Alexander Viro [Wed, 23 Apr 2003 11:18:48 +0000 (04:18 -0700)]
[PATCH] tty cleanups (5/12)

new field - tty->tty_name;
initialized to <driver->name><tty index+driver->base_name>
when we allocate tty_struct.  Drivers code switched to use of that
beast (in debugging printks, mostly).  Large, but trivial.

23 years ago[PATCH] tty cleanups (4/12)
Alexander Viro [Wed, 23 Apr 2003 11:18:37 +0000 (04:18 -0700)]
[PATCH] tty cleanups (4/12)

Instead of registering "drivers" for /dev/tty, /dev/vc/0, /dev/ptmx
and /dev/console (they are never looked up since tty_open() special-cases
them and they should not be looked up - these devices are remapped on open)
we register corresponding chrdev ranges and devfs nodes directly.
/proc/tty/drivers code updated to keep the contents unchanged

23 years ago[PATCH] tty cleanups (3/12)
Alexander Viro [Wed, 23 Apr 2003 11:18:28 +0000 (04:18 -0700)]
[PATCH] tty cleanups (3/12)

/proc/tty/drivers converted to seq_file

23 years ago[PATCH] tty cleanups (2/12)
Alexander Viro [Wed, 23 Apr 2003 11:18:18 +0000 (04:18 -0700)]
[PATCH] tty cleanups (2/12)

Instead of copying tty_driver into tty_struct we put a reference
in there.  tty->driver turned into a pointer, users updated.  Large, but
trivial

23 years ago[PATCH] tty cleanups (1/12)
Alexander Viro [Wed, 23 Apr 2003 11:18:07 +0000 (04:18 -0700)]
[PATCH] tty cleanups (1/12)

Christoph's fix for devfs problems with pty.

23 years ago[PATCH] C99 initializers for fs/proc/proc_misc.c
Art Haas [Wed, 23 Apr 2003 11:09:04 +0000 (04:09 -0700)]
[PATCH] C99 initializers for fs/proc/proc_misc.c

23 years ago[PATCH] Fix C99 initializers in fs/nfs/nfs4proc.c
Art Haas [Wed, 23 Apr 2003 11:08:57 +0000 (04:08 -0700)]
[PATCH] Fix C99 initializers in fs/nfs/nfs4proc.c

23 years ago[PATCH] C99 initializers for drivers/block/genhd.c
Art Haas [Wed, 23 Apr 2003 11:08:48 +0000 (04:08 -0700)]
[PATCH] C99 initializers for drivers/block/genhd.c

23 years ago[PATCH] Avast there ye swabs, prepare to fire a broadside!
Steven Cole [Wed, 23 Apr 2003 11:08:40 +0000 (04:08 -0700)]
[PATCH] Avast there ye swabs, prepare to fire a broadside!

Chuck Ebbert notes that kernel developers are apparently pirates in the
_original_ meaning of the word, and like to "cannonicalize".

I prepared a patch in case anyone cares enough about this.  I tested on
i386 by building and booting.

23 years ago[PATCH] Fix SWSUSP & !SWAP
Pavel Machek [Wed, 23 Apr 2003 11:08:30 +0000 (04:08 -0700)]
[PATCH] Fix SWSUSP & !SWAP

Swsusp without swap makes no sense, and leads to compilation
failure.  So make the dependency clear in the config files.

23 years ago[PATCH] IEEE1394/Firewire updates
Ben Collins [Wed, 23 Apr 2003 10:50:15 +0000 (03:50 -0700)]
[PATCH] IEEE1394/Firewire updates

- Workaround possible reset loop trying to get IRM sanity
- Logical Unit DIrectory (LUN) support for sbp2(scsi) devices.
- Fix hostnum allocation.
- Whitespace and formatting sync with 2.4 branch.
- Make pcilynx return IRQ_NONE for no interrupts to handle.

23 years agoMerge bk://bk.arm.linux.org.uk/linux-2.5-pcmcia
Linus Torvalds [Wed, 23 Apr 2003 08:35:33 +0000 (01:35 -0700)]
Merge bk://bk.arm.linux.org.uk/linux-2.5-pcmcia
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[PCMCIA] Make cb_release_cis_mem() local to cardbus.c
Russell King [Wed, 23 Apr 2003 23:04:13 +0000 (00:04 +0100)]
[PCMCIA] Make cb_release_cis_mem() local to cardbus.c

The cardbus CIS parsing code does not use the PCMCIA resource
subsystem, so there isn't any point in freeing its memory when
we remove PCMCIA memory resources.  We also free CIS resources
immediately prior to calling cb_free().  We might as well move
the function call into cb_free(), thereby making all references
to cb_release_cis_mem() local to cardbus.c

23 years ago[PCMCIA] Don't cache CIS bytes found to be invalid.
Russell King [Wed, 23 Apr 2003 22:18:16 +0000 (23:18 +0100)]
[PCMCIA] Don't cache CIS bytes found to be invalid.

Several PCMCIA cards I have here do not work correctly over a
suspend/resume cycle; the PCMCIA code believes that the card has
been changed in the slot, and therefore performs a remove/insert
cycle.

This seems to be because the card returns more or less random data
when reading memory space, leading to the CIS cache mismatching
the card data.  This in turn is caused because we try to read CIS
data from both the attribute and memory spaces, and we add the result
to the CIS cache whether or not the returned data was valid.

We therefore convert the CIS cache to use a linked list, and provide
a way to remove cached data from that list.  We also replace the
"s->cis_used=0;" construct with a function "destroy_cis_cache(s)"
which clearly describes what we're doing.

23 years ago[PCMCIA] Fix oops in validate_mem when CONFIG_PCMCIA_PROBE=n
Pavel Roskin [Wed, 23 Apr 2003 21:59:33 +0000 (22:59 +0100)]
[PCMCIA] Fix oops in validate_mem when CONFIG_PCMCIA_PROBE=n

If I compile a recent 2.5.x kernel without CONFIG_ISA defined, I get
an oops in validate_mem().  Stack trace contains 0x6b6b6b6 - a clear
sign that freed memory is being accessed.

It's the second validate_mem() in drivers/pcmcia/rsrc_mgr.c - the one
used when CONFIG_PCMCIA_PROBE is not defined.  It turns out the memory
is freed in do_mem_probe() when it's called from validate_mem().

The solution is to use the same trick as in the first validate_mem().
This problem is quite serious and it's not specific to the plx9052
driver. I see it with yenta_socket as well.

23 years ago[PCMCIA] remove unused files
Christoph Hellwig [Wed, 23 Apr 2003 21:55:14 +0000 (22:55 +0100)]
[PCMCIA] remove unused files

From Christoph Hellwig

There's no need to keep the stubs around.

23 years ago[PCMCIA] Fix compilation of cardmgr
Pavel Roskin [Wed, 23 Apr 2003 21:51:31 +0000 (22:51 +0100)]
[PCMCIA] Fix compilation of cardmgr

Patch from Pavel Roskin

ds.h should not be including linux/device.h when compiling userspace
code.

23 years agoMerge bk://kernel.bkbits.net/davem/net-2.5
Linus Torvalds [Wed, 23 Apr 2003 05:50:53 +0000 (22:50 -0700)]
Merge bk://kernel.bkbits.net/davem/net-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoMerge bk://kernel.bkbits.net/davem/sparc-2.5
Linus Torvalds [Wed, 23 Apr 2003 05:50:06 +0000 (22:50 -0700)]
Merge bk://kernel.bkbits.net/davem/sparc-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoACPI: Indicate whether we handled the interrupt or not
Andy Grover [Wed, 23 Apr 2003 03:57:23 +0000 (20:57 -0700)]
ACPI: Indicate whether we handled the interrupt or not

23 years agoMerge groveronline.com:/root/bk/linux-2.5
Andy Grover [Wed, 23 Apr 2003 03:32:32 +0000 (20:32 -0700)]
Merge groveronline.com:/root/bk/linux-2.5
into groveronline.com:/root/bk/linux-acpi

23 years agoACPI: Add missing include
Andy Grover [Wed, 23 Apr 2003 03:22:16 +0000 (20:22 -0700)]
ACPI: Add missing include

23 years agoNever merge vma's that have mapping-private data.
Linus Torvalds [Wed, 23 Apr 2003 01:50:36 +0000 (18:50 -0700)]
Never merge vma's that have mapping-private data.

23 years ago[NET]: SG without checksum support is illegal.
David S. Miller [Tue, 22 Apr 2003 17:19:42 +0000 (10:19 -0700)]
[NET]: SG without checksum support is illegal.

23 years ago[SPARC]: Kill initialize_secondary, unused.
Rob Radez [Tue, 22 Apr 2003 17:07:21 +0000 (10:07 -0700)]
[SPARC]: Kill initialize_secondary, unused.

23 years ago[IPV6]: dst_alloc() clean-up.
Hideaki Yoshifuji [Tue, 22 Apr 2003 16:57:57 +0000 (09:57 -0700)]
[IPV6]: dst_alloc() clean-up.

23 years ago[NETFILTER IPV4]: Fix typo in Kconfig.
Jeff Smith [Tue, 22 Apr 2003 16:56:10 +0000 (09:56 -0700)]
[NETFILTER IPV4]: Fix typo in Kconfig.

23 years ago[IGMP]: Fix bug in broadcast handling.
David Stevens [Tue, 22 Apr 2003 16:53:39 +0000 (09:53 -0700)]
[IGMP]: Fix bug in broadcast handling.

23 years ago[IGMPv3/MPDv2]: Bug fixes and ipv4 multiprotocol API.
David Stevens [Tue, 22 Apr 2003 16:51:48 +0000 (09:51 -0700)]
[IGMPv3/MPDv2]: Bug fixes and ipv4 multiprotocol API.

23 years ago[SPARC64]: Kill unnecessary MOD_{INC,DEC}_USE_COUNT in cpwatchdog and envctrl drivers.
David S. Miller [Tue, 22 Apr 2003 16:40:17 +0000 (09:40 -0700)]
[SPARC64]: Kill unnecessary MOD_{INC,DEC}_USE_COUNT in cpwatchdog and envctrl drivers.

23 years ago[NET]: Do not let GCC reload pointers after NULL checks.
David S. Miller [Tue, 22 Apr 2003 16:14:38 +0000 (09:14 -0700)]
[NET]: Do not let GCC reload pointers after NULL checks.

23 years agoo net: module refcounting for sk_alloc/sk_free
Arnaldo Carvalho de Melo [Tue, 22 Apr 2003 23:53:59 +0000 (20:53 -0300)]
o net: module refcounting for sk_alloc/sk_free

I had to move the rtnetlink_init and init_netlink calls to af_netlink init time, so that
the sk_alloc called down the rtnetlink_init callchain is done after the PF_NETLINK
net_proto_family is sock_registered, and because of that the af_netlink init function
call had to be moved to earlier by means of subsys_initcall (DaveM's suggestion).

23 years ago[SPARC]: Refactor AUXIO support.
Eric Brower [Tue, 22 Apr 2003 15:43:16 +0000 (08:43 -0700)]
[SPARC]: Refactor AUXIO support.

23 years ago[SPARC]: Colin Gibbs gcc-3.x support.
Pete Zaitcev [Tue, 22 Apr 2003 14:37:21 +0000 (07:37 -0700)]
[SPARC]: Colin Gibbs gcc-3.x support.

23 years ago[PATCH] Fix IRDA irq handler prototype
Shachar Shemesh [Tue, 22 Apr 2003 02:34:47 +0000 (19:34 -0700)]
[PATCH] Fix IRDA irq handler prototype

This fixes a mismatch in declaration between "irport_interrupt" in the
header files (returning void) and in the definition (returning
irqreturn_t).

23 years agoMerge bk://kernel.bkbits.net/davem/net-2.5
Linus Torvalds [Tue, 22 Apr 2003 01:35:26 +0000 (18:35 -0700)]
Merge bk://kernel.bkbits.net/davem/net-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoMerge bk://kernel.bkbits.net/davem/sparc-2.5
Linus Torvalds [Tue, 22 Apr 2003 01:34:12 +0000 (18:34 -0700)]
Merge bk://kernel.bkbits.net/davem/sparc-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[PATCH] Update umem driver for newer cards.
Neil Brown [Tue, 22 Apr 2003 01:33:21 +0000 (18:33 -0700)]
[PATCH] Update umem driver for newer cards.

23 years ago[PATCH] Update umem to new request_irq interface
Neil Brown [Tue, 22 Apr 2003 01:33:10 +0000 (18:33 -0700)]
[PATCH] Update umem to new request_irq interface

request_irq requires a handler that returns irqreturn_t,
so mm_interrupt now returns the appropriate value

23 years agoMerge http://jfs.bkbits.net/linux-2.5
Linus Torvalds [Tue, 22 Apr 2003 01:29:36 +0000 (18:29 -0700)]
Merge http://jfs.bkbits.net/linux-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoMerge jfs@jfs.bkbits.net:linux-2.5
Dave Kleikamp [Tue, 22 Apr 2003 03:46:10 +0000 (22:46 -0500)]
Merge jfs@jfs.bkbits.net:linux-2.5
into shaggy.austin.ibm.com:/shaggy/bk/jfs-2.5

23 years ago[PATCH] request structure stack corruption
Jens Axboe [Tue, 22 Apr 2003 01:18:01 +0000 (18:18 -0700)]
[PATCH] request structure stack corruption

This fixes a problem with drivers that have request on the stack for
some operations, like IDE.  If we wake before releasing the request, the
stack may have already disappeared beneath us when the rest of
end_that_request_last() is run.

Fix by making sure the completion is done _last_.

23 years agoMerge davem@nuts.ninka.net:/home/davem/src/BK/net-2.5
David S. Miller [Mon, 21 Apr 2003 18:22:28 +0000 (11:22 -0700)]
Merge davem@nuts.ninka.net:/home/davem/src/BK/net-2.5
into kernel.bkbits.net:/home/davem/net-2.5

23 years agoMerge davem@nuts.ninka.net:/home/davem/src/BK/sparc-2.5
David S. Miller [Mon, 21 Apr 2003 18:21:16 +0000 (11:21 -0700)]
Merge davem@nuts.ninka.net:/home/davem/src/BK/sparc-2.5
into kernel.bkbits.net:/home/davem/sparc-2.5

23 years ago[IP_GRE]: Kill duplicate update_pmtu call.
Steven Whitehouse [Mon, 21 Apr 2003 17:47:19 +0000 (10:47 -0700)]
[IP_GRE]: Kill duplicate update_pmtu call.

23 years ago[BRIDGE]: New maintainership.
Stephen Hemminger [Mon, 21 Apr 2003 17:41:06 +0000 (10:41 -0700)]
[BRIDGE]: New maintainership.

23 years ago[SPARC64]: Add LOOP_{GET,SET}_STATUS64 to ioctl32.
David S. Miller [Mon, 21 Apr 2003 17:36:52 +0000 (10:36 -0700)]
[SPARC64]: Add LOOP_{GET,SET}_STATUS64 to ioctl32.

23 years ago[SPARC]: CLOCK_MONOTONIC fixes, from x86.
David S. Miller [Mon, 21 Apr 2003 17:09:33 +0000 (10:09 -0700)]
[SPARC]: CLOCK_MONOTONIC fixes, from x86.

23 years ago[NETLINK]: Fix minor numbers in netlink_dev.c
David S. Miller [Mon, 21 Apr 2003 17:00:59 +0000 (10:00 -0700)]
[NETLINK]: Fix minor numbers in netlink_dev.c

23 years ago[SPARC64]: Update defconfig.
David S. Miller [Mon, 21 Apr 2003 16:58:39 +0000 (09:58 -0700)]
[SPARC64]: Update defconfig.

23 years agoMerge bk://kernel.bkbits.net/acme/net-2.5
David S. Miller [Mon, 21 Apr 2003 16:53:23 +0000 (09:53 -0700)]
Merge bk://kernel.bkbits.net/acme/net-2.5
into nuts.ninka.net:/home/davem/src/BK/net-2.5

23 years agoo rtnetlink_dev: use C99 struct init style
Arnaldo Carvalho de Melo [Mon, 21 Apr 2003 21:16:52 +0000 (18:16 -0300)]
o rtnetlink_dev: use C99 struct init style

23 years agoo atm/lec.c: use C99 struct init style
Arnaldo Carvalho de Melo [Mon, 21 Apr 2003 21:07:15 +0000 (18:07 -0300)]
o atm/lec.c: use C99 struct init style

23 years agoo rtnetlink: use C99 struct init style
Arnaldo Carvalho de Melo [Mon, 21 Apr 2003 20:48:15 +0000 (17:48 -0300)]
o rtnetlink: use C99 struct init style

23 years ago[SPARC64]: Spelling fixes.
Steven Cole [Mon, 21 Apr 2003 16:13:33 +0000 (09:13 -0700)]
[SPARC64]: Spelling fixes.

23 years agoFix irq event debug print-out, and add stack dump which can
Linus Torvalds [Mon, 21 Apr 2003 16:05:40 +0000 (09:05 -0700)]
Fix irq event debug print-out, and add stack dump which can
give a clue about what the context was that might have caused
the spurious interrupt.

23 years agoMerge nuts.ninka.net:/home/davem/src/BK/network-2.5
David S. Miller [Mon, 21 Apr 2003 12:16:49 +0000 (05:16 -0700)]
Merge nuts.ninka.net:/home/davem/src/BK/network-2.5
into nuts.ninka.net:/home/davem/src/BK/net-2.5

23 years agoMerge bk://bk.arm.linux.org.uk/linux-2.5-rmk
Linus Torvalds [Mon, 21 Apr 2003 09:34:20 +0000 (02:34 -0700)]
Merge bk://bk.arm.linux.org.uk/linux-2.5-rmk
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoMerge flint.arm.linux.org.uk:/usr/src/linux-bk-2.5/linux-2.5
Russell King [Tue, 22 Apr 2003 01:07:59 +0000 (02:07 +0100)]
Merge flint.arm.linux.org.uk:/usr/src/linux-bk-2.5/linux-2.5
into flint.arm.linux.org.uk:/usr/src/linux-bk-2.5/linux-2.5-rmk

23 years ago[ARM] spelling fixes for arm
Steven Cole [Tue, 22 Apr 2003 00:45:12 +0000 (01:45 +0100)]
[ARM] spelling fixes for arm

Patch from Steven Cole.

Here are some spelling fixes for arm.
This was diffed against the current 2.5 tree.

23 years ago[ARM PATCH] 1510/1: use a #define for asm jump address
Eli Carter [Mon, 21 Apr 2003 18:36:54 +0000 (19:36 +0100)]
[ARM PATCH] 1510/1: use a #define for asm jump address

Patch from Eli Carter

# Mon Apr 21 11:11:41 CDT 2003 ejc@rnd-linux-c84
# use-physoffset-define
#
# Use the PHYS_OFFSET #define instead of a hard-coded value.
#
# Diffed against linux-2.5.67-rmk1+1501-3+1506
#
#  arch/arm/boot/compressed/head-xscale.S |    2 +-
#  1 files changed, 1 insertion(+), 1 deletion(-)
#

23 years ago[ARM PATCH] 1511/1: iop321 #define cleanup
Eli Carter [Mon, 21 Apr 2003 18:31:47 +0000 (19:31 +0100)]
[ARM PATCH] 1511/1: iop321 #define cleanup

Patch from Eli Carter

# Mon Apr 21 11:20:06 CDT 2003 ejc@rnd-linux-c84
# pci-io-mem-size-defines
#
# Rename the IOP321_PCI_WINDOW_SIZE #defines to use
# IOP321_PCI_{IO,MEM}_{BASE,SIZE} instead.  This makes the #defines a bit more
# consistent.  No functional change, though it does bring up the question of
# whether res[].end should be BASE+SIZE-1 or BASE+SIZE.
#
# Diffed against linux-2.5.67-rmk1+1501-3+1506+1508-10
#
#  arch/arm/mach-iop3xx/iop321-pci.c    |    8 ++++----
#  include/asm-arm/arch-iop3xx/iop321.h |   10 ++++------
#  2 files changed, 8 insertions(+), 10 deletions(-)
#

23 years ago[ARM PATCH] 1508/1: use #define's for iq80321
Eli Carter [Mon, 21 Apr 2003 18:27:44 +0000 (19:27 +0100)]
[ARM PATCH] 1508/1: use #define's for iq80321

Patch from Eli Carter

# Thu Apr 17 16:37:58 CDT 2003 ejc@rnd-linux-c84
# iq80321-use-addr-defines
#
# Use #define's instead of hard-coded numbers for the device locations for the
# iq80321 board.
#
# Diffed against linux-2.5.65-rmk1+1472-4+1502-3+1506
# Applies cleanly to linux-2.5.67-rmk1+1502-3+1506
#
#  arch/arm/mach-iop3xx/mm-321.c |    4 ++--
#  1 files changed, 2 insertions(+), 2 deletions(-)
#

23 years agoAdd the Xeon variations (Pentium-III and P4-based) to the list
Linus Torvalds [Mon, 21 Apr 2003 09:31:33 +0000 (02:31 -0700)]
Add the Xeon variations (Pentium-III and P4-based) to the list
of Intel CPU optimizations. From Andi Kleen.

23 years ago[PATCH] Runtime memory barrier patching
Andi Kleen [Mon, 21 Apr 2003 09:23:20 +0000 (02:23 -0700)]
[PATCH] Runtime memory barrier patching

This implements automatic code patching of memory barriers based
on the CPU capabilities. Normally lock ; addl $0,(%esp) barriers
are used, but these are a bit slow on the Pentium 4.

Linus proposed this a few weeks ago after the support for SSE1/SSE2
barriers was introduced. I now got around to implement it.

The main advantage is that it allows distributors to ship less binary
kernels but still get fast kernels. In particular it avoids the
need of a special Pentium 4 kernel.

The patching code is quite generic and could be used to patch
other instructions (like prefetches or specific other critical
instructions) too.
Thanks to Rusty's in kernel loader it also works seamlessly for modules.

The patching is done before other CPUs start to avoid potential
erratas with self modifying code on SMP systems. It makes no
attempt to automatically handle assymetric systems (an secondary
CPU having less capabilities than the boot CPU). In this
case just boot with "noreplacement"

23 years ago[PATCH] More careful about VMA merging
Andrew Morton [Mon, 21 Apr 2003 09:23:10 +0000 (02:23 -0700)]
[PATCH] More careful about VMA merging

The VMA merging code can merge vmas which have a ->vm_ops->close() handler.
But this means that per-VMA resources in one of the merged VMAs will not be
freed.

Fix that up by just refusing to merge any VMA's which look "complex": they
have special flags set or they have a ->close handler.

23 years agoMerge home:v2.5/linux
Linus Torvalds [Mon, 21 Apr 2003 06:25:31 +0000 (23:25 -0700)]
Merge home:v2.5/linux
into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux

23 years ago[PATCH] trivial task_prio() fix
Robert Love [Mon, 21 Apr 2003 06:24:30 +0000 (23:24 -0700)]
[PATCH] trivial task_prio() fix

Here is a trivial fix for task_prio() in the case MAX_RT_PRIO !=
MAX_USER_RT_PRIO.  In this case, all priorities are skewed by
(MAX_RT_PRIO - MAX_USER_RT_PRIO).

The fix is to subtract the full MAX_RT_PRIO value from p->prio, not just
MAX_USER_RT_PRIO.  This makes sense, as the full priority range is
unrelated to the maximum user value.  Only the real maximum RT value
matters.

This has been in Andrew's tree for awhile, with no issue.  Also, Ingo
acked it.

23 years agoAllow gcc to generate better code for irq handling.
Linus Torvalds [Mon, 21 Apr 2003 06:10:51 +0000 (23:10 -0700)]
Allow gcc to generate better code for irq handling.

Ok, now that most drivers have been converted to the new
irqreturn_t, we can remove the fascist type-checks and just
use a regular integer type which has a simpler calling
convention.

23 years ago[PATCH] n_hdlc update
Paul Fulghum [Mon, 21 Apr 2003 05:46:13 +0000 (22:46 -0700)]
[PATCH] n_hdlc update

With suggestions from Chritoph Hellwig

- Remove MODULE_USE_COUNT macros
- Add owner member to struct tty_ldisc
- Init tty_ldisc at compile time
- make some functions static

23 years ago[PATCH] synclink_cs update
Paul Fulghum [Mon, 21 Apr 2003 04:48:52 +0000 (21:48 -0700)]
[PATCH] synclink_cs update

- Remove MODULE_USE_COUNT macros
- Add owner member
- Add tiocmget/tiocmset tty callbacks

23 years ago[PATCH] synclinkmp update
Paul Fulghum [Mon, 21 Apr 2003 04:47:01 +0000 (21:47 -0700)]
[PATCH] synclinkmp update

- Remove MODULE_USE_COUNT macros
- Add owner member
- Add tiocmget/tiocmset tty callbacks

23 years ago[PATCH] synclink update
Paul Fulghum [Mon, 21 Apr 2003 04:45:56 +0000 (21:45 -0700)]
[PATCH] synclink update

- Remove MODULE_USE_COUNT macros
- Add owner member
- Add tiocmget/tiocmset tty callbacks

23 years ago[PATCH] 1394 updates
Ben Collins [Mon, 21 Apr 2003 04:28:14 +0000 (21:28 -0700)]
[PATCH] 1394 updates

- New irq handler prototypes.
- Lots of un-enumurated locking fixes/cleanups (thanks in large part to
  spinlock debug compile options in the kernel).
- Other various trivial fixes.

23 years agoReturn IRQ_NONE for ieee1394 driver when the interrupt was for
Linus Torvalds [Mon, 21 Apr 2003 04:27:24 +0000 (21:27 -0700)]
Return IRQ_NONE for ieee1394 driver when the interrupt was for
somebody else.

23 years ago[PATCH] Convert Alpha to the new 2.5 IRQ API
Marc Zyngier [Mon, 21 Apr 2003 04:23:36 +0000 (21:23 -0700)]
[PATCH] Convert Alpha to the new 2.5 IRQ API

This converts the Alpha architecture to the new IRQ API.  Tested on
Jensen.

23 years ago[PATCH] fix printk when an irq doesn't get responded to
Jeff Garzik [Mon, 21 Apr 2003 04:23:28 +0000 (21:23 -0700)]
[PATCH] fix printk when an irq doesn't get responded to

23 years agoMerge bk://kernel.bkbits.net/jgarzik/net-drivers-2.5
Linus Torvalds [Mon, 21 Apr 2003 03:41:10 +0000 (20:41 -0700)]
Merge bk://kernel.bkbits.net/jgarzik/net-drivers-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoResolve differences between Davem and Andrew on irq updates
Linus Torvalds [Mon, 21 Apr 2003 03:39:45 +0000 (20:39 -0700)]
Resolve differences between Davem and Andrew on irq updates

23 years ago[PATCH] irqs: hotplug drivers
Andrew Morton [Mon, 21 Apr 2003 03:32:35 +0000 (20:32 -0700)]
[PATCH] irqs: hotplug drivers

Update hotplug and pcmcia drivers to the new IRQ API.

23 years ago[PATCH] irqs: input drivers
Andrew Morton [Mon, 21 Apr 2003 03:21:28 +0000 (20:21 -0700)]
[PATCH] irqs: input drivers

Update the input and joystick drivers to the new IRQ API.

23 years ago[PATCH] irqs: ISDN
Andrew Morton [Mon, 21 Apr 2003 03:21:21 +0000 (20:21 -0700)]
[PATCH] irqs: ISDN

Update ISDN for the new IRQ API.

23 years ago[PATCH] irqs: IRDA
Andrew Morton [Mon, 21 Apr 2003 03:21:11 +0000 (20:21 -0700)]
[PATCH] irqs: IRDA

Update the IRDA drivers to the new IRQ API.

23 years ago[PATCH] ax25 build fix
Andrew Morton [Mon, 21 Apr 2003 03:21:04 +0000 (20:21 -0700)]
[PATCH] ax25 build fix

spin_lock() does not take a spinlock_t**

23 years ago[PATCH] parport_serial fix
Andrew Morton [Mon, 21 Apr 2003 03:20:56 +0000 (20:20 -0700)]
[PATCH] parport_serial fix

parport_serial is calling a couple of 8250 driver functions with no prototypes
in scope.

Turns out it was calling them with too many args, too.

23 years ago[PATCH] irqs: 1394
Andrew Morton [Mon, 21 Apr 2003 03:20:47 +0000 (20:20 -0700)]
[PATCH] irqs: 1394

Update 1394 to the new IRQ API

23 years ago[PATCH] irqs: video drivers
Andrew Morton [Mon, 21 Apr 2003 03:20:39 +0000 (20:20 -0700)]
[PATCH] irqs: video drivers

Update video drivers to the new IRQ API.

Almost none of them compile, so it was done on-spec.