David S. Miller [Mon, 23 Feb 2004 06:09:29 +0000 (22:09 -0800)]
[NET]: Set default socket rmem/wmem values more sanely and consistently.
1) Take into account the fact that struct sk_buff, which counts towards
socket buffer limits, changes across different platforms.
2) Give ipv4/ipv6 ICMP sockets more accurately the wmem limits they
want.
Andrew Morton [Sun, 22 Feb 2004 15:07:34 +0000 (07:07 -0800)]
[PATCH] Fix fs/partitions/efi.c printk warnings
Lots of these:
fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 2)
fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 3)
fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 2)
fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 3)
fs/partitions/efi.c:395: warning: long unsigned int format, u64 arg (arg 2)
fs/partitions/efi.c:395: warning: long unsigned int format, u64 arg (arg 3)
because it is assuming that you can print a u64 with %ld on 64-bit machines.
Andrew Morton [Sun, 22 Feb 2004 15:07:24 +0000 (07:07 -0800)]
[PATCH] Fix ISDN v.110.
From: David Woodhouse <dwmw2@infradead.org>
Now that we coalesce ACKs for outgoing packets, the v.110 code needs to
look at the size it's given rather than assuming it'll get precisely one
callback for each packet sent...
Andrew Morton [Sun, 22 Feb 2004 15:06:42 +0000 (07:06 -0800)]
[PATCH] security oops fix
From: Stephen Smalley <sds@epoch.ncsc.mil>
Looking at d_splice_alias, it appears that the security_d_instantiate call
in the first case (where it ends up moving an entry in place of the given
dentry) should be passing 'new' rather than 'dentry' to ensure that the
passed dentry is associated with the inode. This patch should fix the
problem.
Andrew Morton [Sun, 22 Feb 2004 15:06:20 +0000 (07:06 -0800)]
[PATCH] Another x86-64 fix for problems from the recent merge
From: Andi Kleen <ak@suse.de>,
From Suresh.B.Siddha
Fix another bug caused by the recent IA32e merge. The full L1 cache size
padding of the GDT didn't work with smaller cache line sizes. Just revert
that.
I must admit I still don't understand why it didn't work, but it's better
to have a working system for now.
Andrew Morton [Sun, 22 Feb 2004 15:05:27 +0000 (07:05 -0800)]
[PATCH] nbd: remove PARANOIA and other cleanups
From: Paul Clements <Paul.Clements@SteelEye.com>
This patch removes the PARANOIA define (it was always defined anyway). It
also removes the requests_in/out counters (which weren't always accurate,
and are superfluous anyway, since the block layer keeps its own in_flight
counter). Also some minor cleanup of comments.
Andrew Morton [Sun, 22 Feb 2004 15:05:05 +0000 (07:05 -0800)]
[PATCH] Kconfig help: dm-crypto && cryptoloop
From: bert hubert <ahu@ds9a.nl>
Current bitkeeper features a cryptoloop which is not safe for journaled
file systems and also a Device Mapper target which is. The patch below
updates Kconfig to that effect.
Andrew Morton [Sun, 22 Feb 2004 15:03:18 +0000 (07:03 -0800)]
[PATCH] m68k: M68k configuration
From: Geert Uytterhoeven <geert@linux-m68k.org>
Update M68k configuration logic:
- Introduce CONFIG_MMU_MOTOROLA and CONFIG_MMU_SUN3
- Move Sun-3 config selection to the top since all other platforms conflict
with it
- Sun-3 implies MMU_SUN3
- All other platforms imply MMU_MOTOROLA
- Sun-3 implies M68020
- Sun-3x implies M68030
- Sun-3 kernels cannot support 68030, 68040, or 68060
Andrew Morton [Sun, 22 Feb 2004 15:02:58 +0000 (07:02 -0800)]
[PATCH] m68k: M68k uses drivers/Kconfig
From: Geert Uytterhoeven <geert@linux-m68k.org>
M68k: Use drivers/Kconfig and fix up some remaining dependencies:
- M68k no longer uses rtc.c
- M68k never has AGP
- CONFIG_ZORRO depends on CONFIG_AMIGS
Andrew Morton [Sun, 22 Feb 2004 15:00:49 +0000 (07:00 -0800)]
[PATCH] Intel i830 AGP fix
From: Terence Ripperda <tripperda@nvidia.com>
When coming out of standby, some chipsets lose their configuration in pci
config space. this includes not just the agp master register being lost, but
chipset-specific information, such as the physical address of the gatt table
and other settings. Calling this configure routine again restores those
settings.
Dave said he was adding chipsets to this list on an "as needed and tested"
basis. I ran across this problem I believe on a Dell Inspiron 4100 with
nv11. (there had been some similar problems with specific nv11 systems when
doing apm with the 2.4 kernel, I suspect, but haven't been able to verify,
it's the same problem).
Andrew Morton [Sun, 22 Feb 2004 15:00:39 +0000 (07:00 -0800)]
[PATCH] more serial driver fixups
From: Russell King <rmk@arm.linux.org.uk>
Here is a patch which updates various serial drivers in the kernel to make
them less broken than they were before. Nevertheless, they are still
broken.
The improvement is to make these drivers use the tiocmget/tiocmset methods,
which are present in the 2.6 kernel.
Many of these have been incorrectly converted from the old global IRQ
locking without regard for SMP issues, or still use the old global IRQ
locking methods which are no longer present in 2.6 kernels.
Russell King [Sun, 22 Feb 2004 23:40:31 +0000 (23:40 +0000)]
[ARM] Fix late abort handler for Thumb code.
ARM720T processors use the "late" abort option, which means we have to
correct the base address of any page fault ourselves. However, we
were mis-correcting the thumb "push" instruction, causing the retried
instruction to fail in a spectacular way.
Peter Teichmann [Sun, 22 Feb 2004 17:16:19 +0000 (17:16 +0000)]
[ARM PATCH 1719/1] Acornfb update
Patch from: Peter Teichmann, edited by Russell King.
* a bugfix to the pseudo palette handling in 16/32 bit modes: the
pseudo palette is always kept in 32 bit integers
* a bugfix to setcolreg() in 16 bit modes: there have to be 32 colour
registers (this has nothing to do with the 16 pseudo palette
colours, it is a bit misleading that setcolreg() is used both for
pseudo palette setup and DIRECTCOLOR colour translation)
(The behaviour should be selected using the visual and the rgb
representation rather than the current BPP setting. --rmk)
* change 16 bit visual from TRUECOLOR to DIRECTCOLOR
* take the limited bandwidth of the RiscPC video memory into account
for acornfb_valid_pixrate()
* acornfb_vidc20_find_rates() now takes into account the differences
between VRAM and DRAM when setting the FIFO preload value. With
VRAM, the FIFO can be set to the highest possible setting because
there are no latency considerations for other memory accesses.
However, in 64 bit bus mode the FIFO preload value must not be set
to VIDC20_CTRL_FIFO_28 because this will let the FIFO overflow.
See VIDC20 manual page 33 (6.0 Setting the FIFO preload value).
* select Multifreq monitor as default, increase frequency range for
Multifreq monitor (this is of course a philosophical question, but
I believe a sensible default setting should not impose too much
limits to the user)
(I dropped this change because it is known that some monitor
hardware shipped with Acorn machines does not take kindly to
wrong timing signals. If people really care, they should use
the kernel command line to select the correct monitor type. --rmk)
Marcel Holtmann [Sun, 22 Feb 2004 17:37:36 +0000 (18:37 +0100)]
[Bluetooth] Copy all L2CAP signal frames to the raw sockets
Currently only very few responses travel back to a L2CAP RAW socket, which
is very bad if you need to see command responses. This patch copies every
signal frame to all attached raw sockets.
Patch from Collin R. Mulliner <collin@betaversion.net>
Bjorn Helgaas [Sat, 21 Feb 2004 22:39:59 +0000 (22:39 +0000)]
[SERIAL] Discover ACPI serial ports before plug-in ports
Patch from: Bjorn Helgaas
This makes us discover ACPI serial ports before PCI ones. The ACPI
ports tend to be built-in, and discovering them late means their ttyS
names move around if PCI serial ports are added or removed. I think
it makes more sense to have relatively predictable and consistent
names for the built-in hardware.
This patch removes the code that tweaked the L1 cache when setting
up the L2 one. That was added a while ago in the intend of making
things more robust but ended up breaking earlier 750 CPU setup.
Also fix some crap in the L1 cache code that is only used for the
powermac sleep at this point.
Alexander Viro [Fri, 20 Feb 2004 10:41:57 +0000 (02:41 -0800)]
[PKT_SCHED]: Convert to {subsys,module}_initcall(), fix init failure bugs in sch_teql.c
pktsched_init() and tc_filter_init() converted to subsys_initcall().
initialization of individual qdisc and tcf_proto switched to
module_init(). Some of them used to be registered twice if built-in, BTW.
init failure handling in sch_teql.c fixed - it used to leave objects
(both qdisc and netdev) registered if insmod failed.
Jeremy Higdon [Fri, 20 Feb 2004 09:47:40 +0000 (01:47 -0800)]
[PATCH] ia64: Fix 64 bit DMA mapping problem with PCI cards on SN2
PCI cards were forced to 32 bit addresses. There is significant
extra overhead to DMA setup with 32 bit, and it is fairly easy to
run out of mapping resources.
This patch reenables 64 bit DMA mapping for PCI cards.