Michael Chan [Tue, 21 Mar 2006 01:47:44 +0000 (17:47 -0800)]
[TG3]: Enable TSO by default
Enable TSO by default on newer chips that support TSO in hardware.
Leave TSO off by default on older chips that do firmware TSO because
performance is slightly lower.
Signed-off-by: Michael Chan <mchan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrea Bittau <a.bittau@cs.ucl.ac.uk> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Original work by Andrea Bittau, Arnaldo Melo cleaned up and fixed several
issues on the merge process.
For now CCID2 was turned the default for all SOCK_DCCP connections, but this
will be remedied soon with the merge of the feature negotiation code.
Signed-off-by: Andrea Bittau <a.bittau@cs.ucl.ac.uk> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Andrea Bittau [Tue, 21 Mar 2006 01:19:55 +0000 (17:19 -0800)]
[DCCP] ackvec: Introduce ack vector records
Based on a patch by Andrea Bittau.
Signed-off-by: Andrea Bittau <a.bittau@cs.ucl.ac.uk> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Harald Welte [Tue, 21 Mar 2006 01:15:11 +0000 (17:15 -0800)]
[NETFILTER] nfnetlink_log: add sequence numbers for log events
By using a sequence number for every logged netfilter event, we can
determine from userspace whether logging information was lots somewhere
downstream.
The user has a choice of either having per-instance local sequence
counters, or using a global sequence counter, or both.
Signed-off-by: Harald Welte <laforge@netfilter.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Harald Welte [Tue, 21 Mar 2006 01:14:12 +0000 (17:14 -0800)]
[NETFILTER] NAT sequence adjustment: Save eight bytes per conntrack
This patch reduces the size of 'struct ip_conntrack' on systems with NAT
by eight bytes. The sequence number delta values can be int16_t, since
we only support one sequence number modification per window anyway, and
one such modification is not going to exceed 32kB ;)
Signed-off-by: Harald Welte <laforge@netfilter.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Patrick McHardy [Tue, 21 Mar 2006 01:12:12 +0000 (17:12 -0800)]
[NET]: Reduce size of struct sk_buff on 64 bit architectures
Move skb->nf_mark next to skb->tc_index to remove a 4 byte hole between
skb->nfmark and skb->nfct and another one between skb->users and skb->head
when CONFIG_NETFILTER, CONFIG_NET_SCHED and CONFIG_NET_CLS_ACT are enabled.
For all other combinations the size stays the same.
Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Stefan Rompf [Tue, 21 Mar 2006 01:11:41 +0000 (17:11 -0800)]
[VLAN]: translate IF_OPER_DORMANT to netif_dormant_on()
this patch adds support to the VLAN driver to translate IF_OPER_DORMANT of the
underlying device to netif_dormant_on(). Beside clean state forwarding, this
allows running independant userspace supplicants on both the real device and
the stacked VLAN. It depends on my RFC2863 patch.
Signed-off-by: Stefan Rompf <stefan@loplof.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Stefan Rompf [Tue, 21 Mar 2006 01:09:11 +0000 (17:09 -0800)]
[NET] core: add RFC2863 operstate
this patch adds a dormant flag to network devices, RFC2863 operstate derived
from these flags and possibility for userspace interaction. It allows drivers
to signal that a device is unusable for user traffic without disabling
queueing (and therefore the possibility for protocol establishment traffic to
flow) and a userspace supplicant (WPA, 802.1X) to mark a device unusable
without changes to the driver.
It is the result of our long discussion. However I must admit that it
represents what Jamal and I agreed on with compromises towards Krzysztof, but
Thomas and Krzysztof still disagree with some parts. Anyway I think it should
be applied.
Signed-off-by: Stefan Rompf <stefan@loplof.de> Signed-off-by: David S. Miller <davem@davemloft.net>
[IPV6]: ADDRCONF: Use our standard algorithm for randomized ifid.
RFC 3041 describes an algorithm to generate random interface
identifier. In RFC 3041bis, it is allowed to use different
algorithm than one described in RFC 3041.
So, let's use our standard pseudo random algorithm to simplify
our implementation.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: David S. Miller <davem@davemloft.net>
[IPV6]: TUNNEL6: Don't try to add multicast route twice.
Since addrconf_add_dev() has already called addrconf_add_mroute()
to added route for multicast prefix, there's no point to call it
again in addrconf_ip6_tnl_config().
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6: (230 commits)
[SPARC64]: Update defconfig.
[SPARC64]: Fix 2 bugs in huge page support.
[SPARC64]: CONFIG_BLK_DEV_RAM fix
[SPARC64]: Optimized TSB table initialization.
[SPARC64]: Allow CONFIG_MEMORY_HOTPLUG to build.
[SPARC64]: Use SLAB caches for TSB tables.
[SPARC64]: Don't kill the page allocator when growing a TSB.
[SPARC64]: Randomize mm->mmap_base when PF_RANDOMIZE is set.
[SPARC64]: Increase top of 32-bit process stack.
[SPARC64]: Top-down address space allocation for 32-bit tasks.
[SPARC64] bbc_i2c: Fix cpu check and add missing module license.
[SPARC64]: Fix and re-enable dynamic TSB sizing.
[SUNSU]: Fix missing spinlock initialization.
[TG3]: Do not try to access NIC_SRAM_DATA_SIG on Sun parts.
[SPARC64]: First cut at VIS simulator for Niagara.
[SPARC64]: Fix system type in /proc/cpuinfo and remove bogus OBP check.
[SPARC64]: Add SMT scheduling support for Niagara.
[SPARC64]: Fix 32-bit truncation which broke sparsemem.
[SPARC64]: Move over to sparsemem.
[SPARC64]: Fix new context version SMP handling.
...
Linus Torvalds [Mon, 20 Mar 2006 18:32:33 +0000 (10:32 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/shaggy/jfs-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/shaggy/jfs-2.6:
JFS: add uid, gid, and umask mount options
JFS: Take logsync lock before testing mp->lsn
JFS: kzalloc conversion
JFS: Add missing file from fa3241d24cf1182b0ffb6e4d412c3bc2a2ab7bf6
JFS: Use the kthread_ API
JFS: Fix regression. fsck complains if symlinks do not have INLINEEA attribute
JFS: ext2 inode attributes for jfs
JFS: semaphore to mutex conversion.
JFS: make buddy table static
JFS: Add back directory i_size calculations for legacy partitions
Linus Torvalds [Mon, 20 Mar 2006 18:30:31 +0000 (10:30 -0800)]
Merge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6: (150 commits)
[PATCH] ipw2100: Update version ipw2100 stamp to 1.2.2
[PATCH] ipw2100: move mutex.h include from ipw2100.c to ipw2100.h
[PATCH] ipw2100: semaphore to mutexes conversion
[PATCH] ipw2100: Fix radiotap code gcc warning
[PATCH] ipw2100: add radiotap headers to packtes captured in monitor mode
[PATCH] ipw2x00: expend Copyright to 2006
[PATCH] drivers/net/wireless/ipw2200.c: fix an array overun
[PATCH] ieee80211: Don't update network statistics from off-channel packets.
[PATCH] ipw2200: Update ipw2200 version stamp to 1.1.1
[PATCH] ipw2200: switch to the new ipw2200-fw-3.0 image format
[PATCH] ipw2200: wireless extension sensitivity threshold support
[PATCH] ipw2200: Enables the "slow diversity" algorithm
[PATCH] ipw2200: Set a meaningful silence threshold value
[PATCH] ipw2200: export `debug' module param only if CONFIG_IPW2200_DEBUG
[PATCH] ipw2200: Change debug level for firmware error logging
[PATCH] ipw2200: Filter unsupported channels out in ad-hoc mode
[PATCH] ipw2200: Fix ipw_sw_reset() implementation inconsistent with comment
[PATCH] ipw2200: Fix rf_kill is activated after mode change with 'disable=1'
[PATCH] ipw2200: remove the WPA card associates to non-WPA AP checking
[PATCH] ipw2200: Add signal level to iwlist scan output
...
Linus Torvalds [Mon, 20 Mar 2006 18:28:51 +0000 (10:28 -0800)]
Merge branch 'block-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/block
* 'block-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/block:
[PATCH] fix rmmod problems with elevator attributes, clean them up
[PATCH] elevator_t lifetime rules and sysfs fixes
[PATCH] noise removal: cfq-iosched.c
[PATCH] don't bother with refcounting for cfq_data
[PATCH] fix sysfs interaction and lifetime rules handling for queues
[PATCH] regularize blk_cleanup_queue() use
[PATCH] fix cfq_get_queue()/ioprio_set(2) races
[PATCH] deal with rmmod/put_io_context() races
[PATCH] stop elv_unregister() from rogering other iosched's data, fix locking
[PATCH] stop cfq from pinning queue down
[PATCH] make cfq_exit_queue() prune the cfq_io_context for that queue
[PATCH] fix the exclusion for ioprio_set()
[PATCH] keep sync and async cfq_queue separate
[PATCH] switch to use of ->key to get cfq_data by cfq_io_context
[PATCH] stop leaking cfq_data in cfq_set_request()
[PATCH] fix cfq hash lookups
[PATCH] fix locking in queue_requests_store()
[PATCH] fix double-free in blk_init_queue_node()
[PATCH] don't do exit_io_context() until we know we won't be doing any IO
David S. Miller [Sun, 19 Mar 2006 07:55:11 +0000 (23:55 -0800)]
[SPARC64]: Optimized TSB table initialization.
We only need to write an invalid tag every 16 bytes,
so taking advantage of this can save many instructions
compared to the simple memset() call we make now.
A prefetching implementation is implemented for sun4u
and a block-init store version if implemented for Niagara.
The next trick is to be able to perform an init and
a copy_tsb() in parallel when growing a TSB table.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 17 Mar 2006 22:41:03 +0000 (14:41 -0800)]
[SPARC64]: Top-down address space allocation for 32-bit tasks.
Currently allocations are very constrained for 32-bit processes.
It grows down-up from 0x70000000 to 0xf0000000 which gives about
2GB of stack + dynamic mmap() space.
So support the top-down method, and we need to override the
generic helper function in order to deal with D-cache coloring.
With these changes I was able to squeeze out a mmap() just over
3.6GB in size in a 32-bit process.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 16 Mar 2006 10:02:32 +0000 (02:02 -0800)]
[SPARC64]: Fix and re-enable dynamic TSB sizing.
This is good for up to %50 performance improvement of some test cases.
The problem has been the race conditions, and hopefully I've plugged
them all up here.
1) There was a serious race in switch_mm() wrt. lazy TLB
switching to and from kernel threads.
We could erroneously skip a tsb_context_switch() and thus
use a stale TSB across a TSB grow event.
There is a big comment now in that function describing
exactly how it can happen.
2) All code paths that do something with the TSB need to be
guarded with the mm->context.lock spinlock. This makes
page table flushing paths properly synchronize with both
TSB growing and TLB context changes.
3) TSB growing events are moved to the end of successful fault
processing. Previously it was in update_mmu_cache() but
that is deadlock prone. At the end of do_sparc64_fault()
we hold no spinlocks that could deadlock the TSB grow
sequence. We also have dropped the address space semaphore.
While we're here, add prefetching to the copy_tsb() routine
and put it in assembler into the tsb.S file. This piece of
code is quite time critical.
There are some small negative side effects to this code which
can be improved upon. In particular we grab the mm->context.lock
even for the tsb insert done by update_mmu_cache() now and that's
a bit excessive. We can get rid of that locking, and the same
lock taking in flush_tsb_user(), by disabling PSTATE_IE around
the whole operation including the capturing of the tsb pointer
and tsb_nentries value. That would work because anyone growing
the TSB won't free up the old TSB until all cpus respond to the
TSB change cross call.
I'm not quite so confident in that optimization to put it in
right now, but eventually we might be able to and the description
is here for reference.
This code seems very solid now. It passes several parallel GCC
bootstrap builds, and our favorite "nut cruncher" stress test which is
a full "make -j8192" build of a "make allmodconfig" kernel. That puts
about 256 processes on each cpu's run queue, makes lots of process cpu
migrations occur, causes lots of page table and TLB flushing activity,
incurs many context version number changes, and it swaps the machine
real far out to disk even though there is 16GB of ram on this test
system. :-)
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 9 Mar 2006 00:09:19 +0000 (16:09 -0800)]
[SPARC64]: Add SMT scheduling support for Niagara.
The mapping is a simple "(cpuid >> 2) == core" for now.
Later we'll add more sophisticated code that will walk
the sun4v machine description and figure this out from
there.
We should also add core mappings for jaguar and panther
processors.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 8 Mar 2006 23:57:03 +0000 (15:57 -0800)]
[SPARC64]: Fix 32-bit truncation which broke sparsemem.
The page->flags manipulations done by the D-cache dirty
state tracking was broken because the constants were not
marked with "UL" to make them 64-bit, which means we were
clobbering the upper 32-bits of page->flags all the time.
This doesn't jive well with sparsemem which stores the
section and indexing information in the top 32-bits of
page->flags.
This is yet another sparc64 bug which has been with us
forever.
While we're here, tidy up some things in bootmem_init()
and paginig_init():
1) Pass min_low_pfn to init_bootmem_node(), it's identical
to (phys_base >> PAGE_SHIFT) but we should use consistent
with the variable names we print in CONFIG_BOOTMEM_DEBUG
2) max_mapnr, although no longer used, was being set
inaccurately, we shouldn't subtract pfn_base any more.
3) All the games with phys_base in the zones_*[] arrays
we pass to free_area_init_node() are no longer necessary.
Thanks to Josh Grebe and Fabbione for the bug reports
and testing. Fix also verified locally on an SB2500
which had a memory layout that triggered the same problem.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 7 Mar 2006 06:50:44 +0000 (22:50 -0800)]
[SPARC64]: Fix new context version SMP handling.
Don't piggy back the SMP receive signal code to do the
context version change handling.
Instead allocate another fixed PIL number for this
asynchronous cross-call. We can't use smp_call_function()
because this thing is invoked with interrupts disabled
and a few spinlocks held.
Also, fix smp_call_function_mask() to count "cpus" correctly.
There is no guarentee that the local cpu is in the mask
yet that is exactly what this code was assuming.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 7 Mar 2006 03:59:50 +0000 (19:59 -0800)]
[SPARC64]: Bulletproof MMU context locking.
1) Always spin_lock_init() in init_context(). The caller essentially
clears it out, or copies the mm info from the parent. In both
cases we need to explicitly initialize the spinlock.
2) Always do explicit IRQ disabling while taking mm->context.lock
and ctx_alloc_lock.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 3 Mar 2006 05:50:47 +0000 (21:50 -0800)]
[SPARC64]: More SUN4V cpu mondo bug fixing.
This cpu mondo sending interface isn't all that easy to
use correctly...
We were clearing out the wrong bits from the "mask" after getting
something other than EOK from the hypervisor.
It turns out the hypervisor can just be resent the same cpu_list[]
array, with the 0xffff "done" entries still in there, and it will do
the right thing.
So don't update or try to rebuild the cpu_list[] array to condense it.
This requires the "forward_progress" check to be done slightly
differently, but this new scheme is less bug prone than what we were
doing before.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 3 Mar 2006 02:12:27 +0000 (18:12 -0800)]
[SPARC64]: Do not allow mapping pages within 4GB of 64-bit VA hole.
The UltraSPARC T1 manual recommends this because the chip
could instruction prefetch into the VA hole, and this would
also make decoding certain kinds of memory access traps
more difficult (because the chip sign extends certain pieces
of trap state).
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 2 Mar 2006 06:42:18 +0000 (22:42 -0800)]
[SPARC64]: Fix _PAGE_EXEC handling.
First of all, use the known _PAGE_EXEC_{4U,4V} value instead
of loading _PAGE_EXEC from memory. We either know which one
to use by context, or we can code patch the test.
Next, we need to check executability of a PTE in the generic
TSB miss handler.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 28 Feb 2006 23:10:26 +0000 (15:10 -0800)]
[SPARC64]: Fix bugs in SUN4V cpu mondo dispatch.
There were several bugs in the SUN4V cpu mondo dispatch code.
In fact, if we ever got a EWOULDBLOCK or other error from
the hypervisor call, we'd potentially send a cpu mondo multiple
times to the same cpu and even worse we could loop until the
timeout resending the same mondo over and over to such cpus.
So let's bulletproof this thing as follows:
1) Implement cpu_mondo_send() and cpu_state() hypervisor calls
in arch/sparc64/kernel/entry.S, add prototypes to asm/hypervisor.h
2) Don't build and update the cpulist using inline functions, this
was causing the cpu mask to not get updated in the caller.
3) Disable interrupts during the entire mondo send, otherwise our
cpu list and/or mondo block could get overwritten if we take
an interrupt and do a cpu mondo send on the current cpu.
4) Check for all possible error return types from the cpu_mondo_send()
hypervisor call. In particular:
HV_EOK) Our work is done, all cpus have received the mondo.
HV_CPUERROR) One or more of the cpus in the cpu list we passed
to the hypervisor are in error state. Use cpu_state()
calls over the entries in the cpu list to see which
ones. Record them in "error_mask" and report this
after we are done sending the mondo to cpus which are
not in error state.
HV_EWOULDBLOCK) We need to keep trying.
Any other error we consider fatal, we report the event and exit
immediately.
5) We only timeout if forward progress is not made. Forward progress
is defined as having at least one cpu get the mondo successfully
in a given cpu_mondo_send() call. Otherwise we bump a counter
and delay a little. If the counter hits a limit, we signal an
error and report the event.
Also, smp_call_function_mask() error handling reports the number
of cpus incorrectly.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 23 Feb 2006 22:19:28 +0000 (14:19 -0800)]
[SPARC64]: Fix TLB context allocation with SMT style shared TLBs.
The context allocation scheme we use depends upon there being a 1<-->1
mapping from cpu to physical TLB for correctness. Chips like Niagara
break this assumption.
So what we do is notify all cpus with a cross call when the context
version number changes, and if necessary this makes them allocate
a valid context for the address space they are running at the time.
Stress tested with make -j1024, make -j2048, and make -j4096 kernel
builds on a 32-strand, 8 core, T2000 with 16GB of ram.
Signed-off-by: David S. Miller <davem@davemloft.net>
projects / linux-2.6 / log