Linux Kernel Documents

Fork

• linux-2.6
• linux-2.4.36

[GCC4] fix build error in net/khttpd/prototypes.h

This patch fixes this error with gcc 4 :

In file included from main.c:68:
prototypes.h:52: error: array type has incomplete element type

[GCC4] fix build error in net/ipv6/sysctl_net_ipv6.c

This patch fixes these errors with gcc 4 :

sysctl_net_ipv6.c:38: error: non-static declaration of 'ipv6_root_table' follows static declaration
sysctl_net_ipv6.c:34: error: previous declaration of 'ipv6_root_table' was here
sysctl_net_ipv6.c:43: error: non-static declaration of 'ipv6_net_table' follows static declaration
sysctl_net_ipv6.c:35: error: previous declaration of 'ipv6_net_table' was here

[GCC4] fix build error in net/ipv6/ip6_fib.c

This patch fixes this error with gcc4 :
ip6_fib.c:98: error: static declaration of 'fib6_walker_list' follows non-static declaration
/usr/src/git/linux-2.4/include/net/ip6_fib.h:98: error: previous declaration of 'fib6_walker_list' was here
make[3]: *** [ip6_fib.o] Error 1
make[3]: Leaving directory `/usr/src/git/linux-2.4/net/ipv6'
make[2]: *** [first_rule] Error 2
make[2]: Leaving directory `/usr/src/git/linux-2.4/net/ipv6'
make[1]: *** [_subdir_ipv6] Error 2
make[1]: Leaving directory `/usr/src/git/linux-2.4/net'
make: *** [_dir_net] Error 2

[GCC4] fix build error in fs/intermezzo/presto.c

This patch fixes these errors with gcc 4 :

presto.c: In function 'izo_mark_cache':
presto.c:347: error: invalid lvalue in assignment
presto.c:348: error: invalid lvalue in assignment
presto.c: In function 'izo_mark_fset':
presto.c:386: error: invalid lvalue in assignment
presto.c:387: error: invalid lvalue in assignment

[GCC4] fix build error in include/asm-ppc/spinlock.h

asm-ppc/spinlock.h references an extern __spin_trylock() function
which is defined as static in arch/ppc/lib/locks.c.

[GCC4] fix build error in include/net/irda/irlan_event.h

This patch fixes this error with gcc 4 :

irlan_common.c:74: error: static declaration of 'irlan_state' follows non-static declaration
/usr/src/git/linux-2.4/include/net/irda/irlan_event.h:70: error: previous declaration of 'irlan_state' was here

[GCC4] fix build error in include/net/irda/irttp.h

This patch fixes this error with gcc 4 :

irttp.c:41: error: static declaration of 'irttp' follows non-static declaration
/usr/src/git/linux-2.4/include/net/irda/irttp.h:212: error: previous declaration of 'irttp' was here

[GCC4] fix build error in include/net/udp.h

This patch fixes this error with gcc 4 :

In file included from af_econet.c:39:
/usr/src/git/linux-2.4/include/net/udp.h:75: error: array type has incomplete element type

[GCC4] fix build error in include/linux/intermezzo_fs.h

This patch fixes these errors with gcc 4 :

super.c:339: error: static declaration of 'presto_fs_type' follows non-static declaration
/usr/src/git/linux-2.4/include/linux/intermezzo_fs.h:324: error: previous declaration of 'presto_fs_type' was here

[GCC4] fix build error in include/linux/fsfilter.h

This patch fixes this error with gcc 4 :

methods.c:60: error: static declaration of 'filter_oppar' follows non-static declaration
/usr/src/git/linux-2.4/include/linux/fsfilter.h:73: error: previous declaration of 'filter_oppar' was here

[GCC4] fix build error in include/net/irda/qos.h

This patch fixes this error with gcc 4 :

actisys.c:65: error: static declaration of 'baud_rates' follows non-static declaration
/usr/src/git/linux-2.4/include/net/irda/qos.h:86: error: previous declaration of 'baud_rates' was here

[GCC4] fix build error in include/linux/generic_serial.h

This patch fixes this error with gcc 4 :

generic_serial.c:36: error: static declaration of 'gs_debug' follows non-static declaration
/usr/src/git/linux-2.4/include/linux/generic_serial.h:101: error: previous declaration of 'gs_debug' was here

[GCC4] add preliminary support for GCC 4 (Mikael Pettersson)

This is the well tested part of Mikael Pettersson's GCC4 patches.
It offers support for GCC4 on x86, x86_64 and ppc. Other archs
untested. The risk of regression on older compilers at this stage
should be close to zero. At least, x86 with allmodconfig still builds
with gcc-2.95.3.

Other places have been carefully fixed but the code has not been run
yet. They will be merged individually in further patches.

Merge branch 'next'

Change VERSION to 2.4.33

[PATCH] export memchr() which is used by smbfs and lp driver.

Recently, an smbfs fix added a dependency on memchr() which is
not exported if smbfs is built as a module.

Signed-Off-By: Willy Tarreau <w@1wt.eu>

[PKTGEN] : fix an oops when used with bonding driver (Tien ChenLi)

Back-port from Tien ChenLi's 2.6 patch described this below.

"I fixed a bug in pktgen so it won't cause oops when used with
balance-tlb or balance-alb bonding driver".

The root cause is that the bond_alb_xmit in bonding will peek the
destination address in packet via the skb->nh.iph pointer, generally
this will be filled by upper layer network driver, but the packet
generated by pktgen will be sent to device driver so it will need to
set this pointer correctly. The other two pointers are not necessary
for now, they are set to avoid similar problem.

[PATCH] USB: unsigned long flags

After a similar problem surfaced in usbserial, I looked across the code base
and found this instance. The bug is obvious: local_irq_save takes a long,
not int.

Signed-Off-By: Pete Zaitcev <zaitcev@redhat.com>

[PATCH] USB: Little Rework for usbserial

This patch fixes various hangs and oopses which happen if serial devices
are handled roughly (e.g. disconnected while open), open-close-open
races and hangs, and issues with getty running on ttyUSBx.

Finally, I got rid of the "#ifdef I_AM_A_DARING_HACKER". Originally,
I thought it would be there for a week or two, and it was stuck in the
code for two years.

Signed-off-by: Pete Zaitcev <zaitcev@redhat.com>

[PATCH] AVM C4 ISDN card : use cpu_relax() in busy loops

As suggested by Alan, use cpu_relax() in 3 busy loops : "It's a
polled busy loop so you want other CPU threads to run if possible".

Signed-off-by: Willy Tarreau <w@1wt.eu>

[PATCH] Fix AVM C4 ISDN card init problems with newer CPUs

AVM C4 ISDN NIC: Add three memory barriers, taken from 2.6.7,
(they are there in 2.6.17.7 too), to fix module initialization
problems appearing with at least some newer Celerons and
Pentium III.

Signed-off-by: Jukka Partanen <jspartanen@gmail.com>
Acked-by: Karsten Keil <kkeil@suse.de>

[PATCH] Bug with USB proc_bulk in 2.4 kernel

Replace the semaphore exclusive_access with an open-coded lock for
the special use. The lock can be taken for: read, write, and both.
This way, two bulk URBs can be submitted simultaneously with ioctl
USBDEVFS_BULK: one read and one write. Such access was possible
before 2.4.28.

The semaphore was introduced in 2.4.28 for the purpose of exclusion
between access from device.c (cat /proc/bus/usb/devices), devio.c
through libusb, and in-kernel driver. Currently, only usb-storage
observes the locking protocol. The most popular device which locks
in case of simultaneous access is TEAC CD-210PU.

Signed-off-By: Pete Zaitcev <zaitcev@redhat.com>

[PATCH] [BLKMTD] : missing offset sometimes causes panics

Using JFFS2 on CompactFlash devices via blkmtd on kernel 2.4.32 causes
occasionnal kernel panics.

It happens that every few days (5-10 days), we encounter a kernel BUG in
set_bh_page() during a write because ->offset is bigger than PAGE_SIZE.
The call chain passes via jffs2 and blkmtd. I finally tracked it down to
the commit_pages() function in blkmtd not setting the ->offset member
in the kiovec structure. This one only gets initialized on a previous
read, because kiobuf_init() does not set it either.

I found it logical to set it to zero in commit_pages() because this
function also sets the ->length parameter to a page-aligned value.

2.6 does not have this problem, because it does not use kiovecs.

I suspect not many people use jffs2 on top of blkmtd, and even in this
case, the conditions to meet to trigger the bug are not common at all.

Signed-Off-By: Willy Tarreau <w@1wt.eu>
Ack-By: David Woodhouse <dwmw2@infradead.org>

[PATCH] 2.4 NFS client - update d_cache when server reports ENOENT on an NFS remove

Anuwat Phrukphicharn of HP discovered and patched this problem in
RHEL-3, and asked that I push it upstream.

When the 2.4 NFS client does a REMOVE and gets an ENOENT back from the
server it does not remove the dentry from the d_cache. This can make it
inappropriately keep writing to an inode that has been renamed.

To reproduce, run this on an NFS server with /scratch exported:

#!/bin/sh
dir=/scratch/98201
mkdir -p $dir/recv
cat /dev/null > $dir/recv/x
while true; do
length=`wc -l $dir/recv/x | cut -d' ' -f1`
if [ ${length} -gt 1 ]; then
echo "problem occured!"
date
exit 1
fi
mv $dir/x $dir/recv/x
sleep 1
done

...and then do this on the client:

mount server:/scratch /mnt/server
while true; do
rm -f /mnt/server/98201/x
date >> /mnt/server/98201/x
usleep 100
done

The file "x" should never contain more than 1 line, but occasionally,
the server will report an ENOENT back to the client (indicating that the
server script has renamed the file before the rm could occur). After
this, the client will keep appending to the same inode, even though the
file has been renamed.

I've not seen this problem in 2.6 kernels, but I've not done any
extensive testing there as of yet so I can't confirm whether it's still
a problem there or not.

A patch to fix this follows. It just makes ENOENT a special case when
handling errors in the nfs_safe_remove function, and lets the client
update the dcache as if the remove had succeeded. The ENOENT is still
reported back to userspace. This corrected the problem on my test rig
and for the reporter as well:

Signed-off-by: Jeff Layton <jlayton@poochiereds.net>

Change VERSION to 2.4.33-rc3

[PATCH-2.4] Typo in cdrom.c

Discussion here :
http://bugzilla.kernel.org/show_bug.cgi?id=2966

The typo seems to exist in linux-2.4 too, at least in
2.4.32, 2.4.32-hf32.6 and 2.4.33pre3. The fix for
linux-2.4 would be just like the proposed patch for
linux-2.6.

[PATCH] ethtool: two oopses in ethtool_set_coalesce() and ethtool_set_pauseparam()

The function pointers which were checked were for their get_* counterparts.
Typically a copy-paste typo.

Signed-off-by: Willy Tarreau <w@1wt.eu>

[PATCH] EXT3: ext3 block bitmap leakage

This patch fixes ext3 block bitmap leakage,
which leads to the following fsck messages on
_healthy_ filesystem:
Block bitmap differences: -64159 -73707

All kernels up to 2.6.17 have this bug.

Found by
Vasily Averin <vvs@sw.ru> and Andrey Savochkin <saw@sawoct.com>
Test case triggered the issue was created by
Dmitry Monakhov <dmonakhov@sw.ru>

Signed-Off-By: Vasiliy Averin <vvs@sw.ru>
Signed-Off-By: Andrey Savochkin <saw@sawoct.com>
Signed-Off-By: Kirill Korotaev <dev@openvz.org>
CC: Dmitry Monakhov <dmonakhov@sw.ru>

[PATCH] range checking for sleep states sent to /proc/acpi/sleep

A range checking is missing in acpi_system_write_sleep() in kernel
2.4, and writing a large integer value to /proc/acpi/sleep will cause
an oops. I could reproduce one this way :

# echo 0x800000 >/proc/acpi/sleep

Fix extracted from the PaX patch.

Change VERSION to v2.4.33-rc2