The waitqueue API is used to replace a custom wait mechanism. Only one
global waitqueue (instead of per-device waitqueues or completions) is
added because there is usually just one waiter.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
- Add checks for the (very unlikely) cases that the target writes too
little or too much status data or writes unsolicited status.
- Indicate that these and similar conditions are unlikely().
- Check the 'resp' and 'sbp_status' fields for possible failure status.
- Slightly optimize access macros for the status block bitfields.
- Unify a few related log messages.
TODO: Check if 'src'==1, then withhold the respective ORB from reuse
until status for any subsequent ORB was received. This is an old bug
whose fix requires more complex command queue handling.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Sbp2's copy of the status fifo was cleared when management ORBs or new
command ORBs were prepared. The latter had potential for a race
condition if the block layer's soft IRQ and the 1394 LLD's interrupt
handler ran on different CPUs. It would also yield wrong status if a
command was completed with non-zero completion status before other
commands that had zero completion status, and no new command was
enqueued in the meantime.
Now, the status buffer is cleared right before it is written. Thus it
ends up in the following simpler and safer access pattern:
- sbp2_alloc_device: allocates and implicitly clears once,
- sbp2_handle_status_write: clears, writes, and reads,
- sbp2_query_logins, sbp2_login_device, sbp2_reconnect_device: read.
The latter three do not race with sbp2_handle_status_write because of
how the protocol works.
As a tiny optimization, the first two quadlets of the status never need
to be cleared.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Only the driver writes ORBs, the device just reads them. Therefore
PCI_DMA_BIDIRECTIONAL can be replaced by PCI_DMA_TODEVICE which may be
cheaper on some architectures.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Since sbp2 is at the moment unable to do anything with the return value
of sbp2_link_orb_command, just discard it.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
The sbp2 initiator has two ways to tell a target's fetch agent about new
command ORBs:
- Write the ORB's address to the ORB_POINTER register. This must not
be done while the fetch agent is active.
- Put the ORB's address into the previously submitted ORB's next_ORB
field and write to the DOORBELL register. This may be done while the
fetch agent is active or suspended. It must not be done while the
fetch agent is in reset state.
Sbp2 has a last_orb pointer which indicates in what way a new command
should be announced. That pointer is concurrently accessed at various
occasions. Furthermore, initiator and target are accessing the next_ORB
field of ORBs concurrently and asynchronously.
This patch does:
- Protect all initiator accesses to last_orb by sbp2_command_orb_lock.
- Add pci_dma_sync_single_for_device before a previously submitted
ORB's next_ORB field is overwritten.
- Insert a memory barrier between when next_ORB_lo and next_ORB_hi are
overwritten. Next_ORB_hi must not be updated before next_ORB_lo.
- Remove the rather unspecific and now superfluous qualifier "volatile"
from the next_ORB fields.
- Add comments on how last_orb is connected with what is known about
the target's fetch agent's state.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
This patch reduces the size of struct hpsb_host and also removes
semaphores from ieee1394_transactions.c. On i386, struct hpsb_host
shrinks from 10656 bytes to 6688 bytes. This is accomplished by
- using a single wait_queue for hpsb_get_tlabel instead of many
instances of semaphores,
- using a single lock to serialize access to all tlabel pools (the
protected code regions are small, i.e. lock contention very low),
- omitting the sysfs attribute tlabels_allocations.
Drawback: In the rare case that a process needs to sleep because all
transaction labels for the node are temporarily exhausted, it is also
woken up if a tlabel for a different node became free, checks for an
available tlabel, and is put to sleep again. The check is not costly
and the situation occurs extremely rarely. (Tlabels are typically
only exhausted if there was no context switch to the khpsbpkt thread
which recycles tlables.) Therefore the benefit of reduced tpool size
outweighs this drawback.
The sysfs attributes tlabels_free and tlabels_mask are not compiled
anymore unless CONFIG_IEEE1394_VERBOSEDEBUG is set.
The by far biggest member of struct hpsb_host, the struct csr_control
csr (5272 bytes on i386), is now placed at the end of struct hpsb_host.
Note, hpsb_get_tlabel calls the macro wait_event_interruptible with a
condition argument which has a side effect (allocation of a tlabel and
manipulation of the packet). This side effect happens only if the
condition is true. The patch relies on wait_event_interruptible not
evaluating the condition again after it became true.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
[PATCH 9/9] s390: qeth driver fixes [6/6]
From: Frank Pavlic <fpavlic@de.ibm.com>
- Hipersockets has no IPV6 support, thus prevent issueing
SETRTG_IPV6 control commands on Hipersockets devices.
- fixed error handling in qeth_sysfs_(un)register
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
[PATCH 8/9] s390: qeth driver fixes [5/6]
From: Frank Pavlic <fpavlic@de.ibm.com>
fix kernel panic in qdio queue handling.
qeth_qdio_clear_card() could be invoked by 2 CPUs
simultaneously (for example reboot event and recovery).
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
[PATCH 7/9] s390: qeth driver fixes [4/6]
From: Frank Pavlic <fpavlic@de.ibm.com>
- fix kernel crash due to race,
set card->state to SOFTSETUP after
card and card->dev are initialized properly.
- remove CONFIG_QETH_PERF_STATS, use sysfs attribute instead,
as we want to have the ability to turn on/off the
statistics at runtime.
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
[PATCH 6/9] s390: qeth driver fixes [3/6]
From: Frank Pavlic <fpavlic@de.ibm.com>
fixed kernel panic caused by qeth driver:
Using a bonding device qeth driver will realloc
headroom for every skb coming from the bond device.
Once this happens qeth frees the original skb and
set the skb pointer to the new realloced skb.
Under heavy transmit workload (e.g.UDP streams) through bond
network device the qdio output queue might get full.
In this case we return with EBUSY from qeth_send_packet.
Returning to qeth_hard_start_xmit routine
the skb address on the stack still points to the old address,
which has been freed before.
Returning from qeth_hard_start_xmit with EBUSY results in
requeuing the skb. In this case it corrupts the qdisc queue
and results in kernel panic.
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
[PATCH 5/9] s390: qeth driver fixes [2/6]
From: Frank Pavlic <fpavlic@de.ibm.com>
- fixed error handling in create_device_attributes
- fixed some minor bugs in IPv4
and IPv6 address checking
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
[PATCH 4/9] s390: qeth driver fixes [1/6]
From: Frank Pavlic <fpavlic@de.ibm.com>
- Drop incoming packets with vlan_tag set
if card->vlangrp is not set.
- use always vlan_hwaccel_rx to pass
vlan frames to the stack.
- fix recovery problem. Device was recovered
properly but still not working.
netif_carrier_on call right before
recovery start fixes it.
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
[PATCH 3/9] s390: Makefile cleanup
From: Frank Pavlic <fpavlic@de.ibm.com>
remove CONFIG_MPC from Makefile which was
introduced accidently in the past.
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
[PATCH 2/9] s390: netiucv driver fixes
From: Frank Pavlic <fpavlic@de.ibm.com>
- missing lock initialization added
- avoid duplicate iucv-interfaces to the same peer
- rw-lock added for manipulating the list of
defined iucv connections
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Hi Jeff,
this is a RESEND of the nine s390 network driver patches.
I finally found that my kmail corrupted almost every patch
I sent the last time. Please apply these 9 patches and forget
about my first attempt! Sorry for the delay, I had some fights
with sendmail, IMAP and mutt configuration.
Frank
[RESEND PATCH 1/9] s390: minor s390 network driver fixes
From: Frank Pavlic <fpavlic@de.ibm.com>
- iucv driver:
use do { } while (0) constructs
instead of empty defines to avoid compile bugs.
- ctc driver:
missing lock initialization added
- lcs driver:
BUG_ON usage was removed accidently
with the last lcs patch.
Put them back in place.
Signed-off-by: Frank Pavlic <fpavlic@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
* 'for-linus' of master.kernel.org:/pub/scm/linux/kernel/git/roland/infiniband:
RDMA/cma: Increase the IB CM retry count in CMA
IPoIB: Retry failed send-only multicast group joins
IB/srp: Don't schedule reconnect from srp
In some special case (padding because of sync or umount) it can be possible
that summary information is not fit to the end of the erase block. In
these cases the collecting of summary is disabled for this erase block.
The problem was that this was not respected by jffs2_sum_add_kvec(). This
patch fix this bug.
Signed-off-by: Ferenc Havasi <havasi@inf.u-szeged.hu>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
In the case of data-pad-ecc-pad-data... layout the oob start position has
to be sizeof(data) in nand_write_oob_syndrom().
In nand_fill_oob() we need to copy to buf + buffer offset instead of buf +
write offset.
Signed-off-by: Vitaly Wool <vwool@ru.mvista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
ext3-get-blocks support caused ~20% degrade in Sequential read
performance (tiobench). Problem is with marking the buffer boundary
so IO can be submitted right away. Here is the patch to fix it.
2.6.18-rc6:
-----------
# ./iotest
1048576+0 records in
1048576+0 records out
4294967296 bytes (4.3 GB) copied, 75.2726 seconds, 57.1 MB/s
real 1m15.285s
user 0m0.276s
sys 0m3.884s
2.6.18-rc6 + fix:
-----------------
[root@elm3a241 ~]# ./iotest
1048576+0 records in
1048576+0 records out
4294967296 bytes (4.3 GB) copied, 62.9356 seconds, 68.2 MB/s
The boundary block check in ext3_get_blocks_handle needs to be adjusted
against the count of blocks mapped in this call, now that it can map
more than one block.
Signed-off-by: Suparna Bhattacharya <suparna@in.ibm.com>
Tested-by: Badari Pulavarty <pbadari@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
I think there is a bug in kmod.c: In __call_usermodehelper(), when
kernel_thread(wait_for_helper, ...) return success, since wait_for_helper()
might call complete() at any time, the sub_info should not be used any
more.
Normally wait_for_helper() take a long time to finish, you may not get
problem for most of the case. But if you remove /sbin/modprobe, it may
become easier for you to get a oop in khelper.
Cc: Matt Helsley <matthltc@us.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
`make headers_check' wants to go and write stuff in /lib/modules, which
requires root, whic is unfortunate.
In fact, there's no _particular_ reason for headers_install to put it there
either -- it can go into a subdirectory of the build tree in both cases.
It's not intended to go directly into /usr/include, which is why we didn't
put it there -- and we certainly don't want people screwing around with
symlinking to it. It's for distributors to take away and do stuff with, so
leaving it in $(objtree) is fine, even in the headers_install case.
I picked $(objtree)/usr/include but I have no _particular_ preference
for that; it just seemed reasonable.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Fix two problems with the CONFIG_EMBEDDED submenu:
(1) The menu was split in two by the rt_mutex patch, which moved
half the items into the "General setup" menu.
(2) CONFIG_SYSCTL and CONFIG_UID16 were added to the main menu
instead of the submenu.
Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Alpha currently fails 'make headers_check' in the 2.6.18-rc kernels. This
patch fixes it, by moving the existing #ifdef __KERNEL__ in asm/page.h so that
it covers everything that userspace shouldn't so, and by adding asm/compiler.h
to the list of exported files so that its use within asm/byteorder.h is
successful.
[ Note that at least with GCC 4, <linux/compiler.h> doesn't do the forced
inlining about which there are nasty comments (and a workaround) in
<asm/compiler.h>, unless you set CONFIG_FORCED_INLINING. Rather than keep
the mess you have in <asm/compiler.h> you could perhaps just make sure
CONFIG_FORCED_INLINING=n is also honoured with GCC3, and make sure it cannot
be set for Alpha? ]
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Richard Henderson <rth@twiddle.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
On Tue, 2006-09-12 at 17:44 +0100, David Woodhouse wrote:
> asm-x86_64/elf.h requires asm/processor.h, which does not exist
> asm-x86_64/signal.h requires linux/linkage.h, which does not exist
> asm-x86_64/unistd.h requires linux/linkage.h, which does not exist
> asm-x86_64/vsyscall.h requires linux/seqlock.h, which does not exist
Again, move stuff which shouldn't be visible inside (mostly already existing)
#ifdef __KERNEL__.
This fixes a bunch of mislabelled and unlabelled #endifs in unistd.h and also
cleans that up to conform with what's visible on other architectures, since
the minimal fix for the error reported about would have involved a more
intrusive patch, renesting other ifdefs.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This brings i386 asm/unistd.h into consistency with other architectures by not
exporting functionality which is not necessary.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
On Tue, 2006-09-12 at 17:44 +0100, David Woodhouse wrote:
> asm-ia64/ptrace.h requires asm/asm-offsets.h, which does not exist
> asm-ia64/resource.h requires asm/ustack.h, which does not exist
Hide parts which shouldn't be visible to userspace.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
On Tue, 2006-09-12 at 17:44 +0100, David Woodhouse wrote:
> asm-s390/debug.h requires linux/string.h, which does not exist
> asm-s390/elf.h requires asm/system.h, which does not exist
Move things around slightly so the right things end up within
#ifdef __KERNEL__ and thus don't pollute the exported headers.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We generate an <asm/foo.h> which includes either <asm-$ARCH/foo.h> or
<asm-$ALTARCH/foo.h> as appropriate. But we were doing this dependent on
whether the file in question existed in the _unexported_ tree, not the
exported tree. So if a file was exported to userspace in one asm- directory
but not the other, the generated file in asm/ was incorrect.
This only changed the failure mode if it _was_ included from a nice #error to
a less explicable #include failure -- but it also gave false errors in 'make
headers_check' output. Fix it by looking in the right place instead.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The kernel build system supports making symbol type files (*.symtypes) from C
source files. Add these files to .gitignore.
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The kernel build system supports making mixed source and assembly listings
(*.lst) from C source files. Add these files to .gitignore.
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The kernel build system supports making preprocessed files (*.i) from C source
files. Add these files to .gitignore.
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
If one of the OEM flags becomes set in the flags from the hardware, the
driver could hang if no OEM handler was set. Fix the code to handle this.
This was tested by setting the flags by hand after they were fetched.
Signed-off-by: Corey Minyard <minyard@acm.org>
Ackde-by: Matt Domsch <Matt_Domsch@dell.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Inodes earlier than the 'first' inode (e.g. journal, resize) should be
rejected early - except the root inode. Also inode numbers that are too
big should be rejected early.
[akpm@osdl.org: cleanup]
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This prevents bad inode numbers from triggering errors in ext2_get_inode.
[akpm@osdl.org: speedup, cleanup]
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
When the ipmi_si module is loaded on a system without any ipmi device, it
fails with nodev. It would be fine if all resources were freed. A call to
device_unregister() is missing, resulting to a oops when you remove the
ipmi_msghandler.
Signed-off-by: Arnaud Patard <apatard@mandriva.com>
Acked-by: Corey Minyard <minyard@acm.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Fix a bug where the IRQ_PENDING flag is never cleared and the ISR is called
endlessly without an actual interrupt.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
3 seems like a low number of IB Communication Manager retries to set;
we see connections failing under stress, and in any case 3 just looks
like an arbitrary number. 15 is the max value allowed by the
InfiniBand spec.
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When a send-only multicast group join fails, mcast->query must be set
to NULL. Otherwise, IPoIB will never retry the join and the multicast
group will never be reachable.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If there is a problem in the connection, the SCSI mid-layer will
eventually call srp_reset_host(), which will call srp_reconnect(), so
we do not need to schedule a call to srp_reconnect_work() from
srp_completion().
Removing this prevents srp_reset_host() from failing if a reconnect
scheduled from srp_completion() is already in progress, which in turn
was causing crashes as both SCSI midlayer and srp_reconnect() were
cancelling commands.
Signed-off-by: Ishai Rabinovitz <ishai@mellanox.co.il>
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fix a missing call to dma_unmap_single() in the receive path. Without
this call, errors have been observed on non-cache-coherent systems.
Signed-off-by Dale Farnsworth <dale@farnsworth.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Add support for new hardware and bumps the version to 3.6.10. It seems
there were several changes introduced including soft_irq. I decided to
bump the major number to reflect these changes. Since we're still
supporting older vendor kernels I need some way differentiate between
kernel versions <=2.6.10 and newer kernels >=2.6.16.
Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Jens Axboe <axboe@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
MIPS asm/page.h unconditionally includes <asm-generic/memory_model.h>, which
doesn't exist in userspace. Move an #endif /* __KERNEL__ */ down a few lines
to prevent that.
Also, remove the broken definition of PAGE_SIZE which is never going to be
correct -- in the absence of PAGE_SIZE, non-broken userspace will fall back to
using sysconf() or getpagesize() instead.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
