David S. Miller [Fri, 10 Feb 2006 00:12:22 +0000 (16:12 -0800)]
[SPARC64]: Implement sun4v TSB miss handlers.
When we register a TSB with the hypervisor, so that it or hardware can
handle TLB misses and do the TSB walk for us, the hypervisor traps
down to these trap when it incurs a TSB miss.
Processing is simple, we load the missing virtual address and context,
and do a full page table walk.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 8 Feb 2006 10:53:50 +0000 (02:53 -0800)]
[SPARC64]: Sun4v interrupt handling.
Sun4v has 4 interrupt queues: cpu, device, resumable errors,
and non-resumable errors. A set of head/tail offset pointers
help maintain a work queue in physical memory. The entries
are 64-bytes in size.
Each queue is allocated then registered with the hypervisor
as we bring cpus up.
The two error queues each get a kernel side buffer that we
use to quickly empty the main interrupt queue before we
call up to C code to log the event and possibly take evasive
action.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 6 Feb 2006 06:27:28 +0000 (22:27 -0800)]
[SPARC64]: Sanitize %pstate writes for sun4v.
If we're just switching between different alternate global
sets, nop it out on sun4v. Also, get rid of all of the
alternate global save/restore in the OBP CIF trampoline code.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 6 Feb 2006 04:47:26 +0000 (20:47 -0800)]
[SPARC64]: Fill dead cycles on trap entry with real work.
As we save trap state onto the stack, the store buffer fills up
mid-way through and we stall for several cycles as the store buffer
trickles out to the L2 cache. Meanwhile we can do some privileged
register reads and other calculations, essentially for free.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 4 Feb 2006 08:10:01 +0000 (00:10 -0800)]
[SPARC64]: Refine register window trap handling.
When saving and restoing trap state, do the window spill/fill
handling inline so that we never trap deeper than 2 trap levels.
This is important for chips like Niagara.
The window fixup code is massively simplified, and many more
improvements are now possible.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 27 Feb 2006 07:27:19 +0000 (23:27 -0800)]
[SPARC64]: Refine code sequences to get the cpu id.
On uniprocessor, it's always zero for optimize that.
On SMP, the jmpl to the stub kills the return address stack in the cpu
branch prediction logic, so expand the code sequence inline and use a
code patching section to fix things up. This also always better and
explicit register selection, which will be taken advantage of in a
future changeset.
The hard_smp_processor_id() function is big, so do not inline it.
Fix up tests for Jalapeno to also test for Serrano chips too. These
tests want "jbus Ultra-IIIi" cases to match, so that is what we should
test for.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 3 Feb 2006 00:16:24 +0000 (16:16 -0800)]
[SPARC64]: Turn off TSB growing for now.
There are several tricky races involved with growing the TSB. So just
use base-size TSBs for user contexts and we can revisit enabling this
later.
One part of the SMP problems is that tsb_context_switch() can see
partially updated TSB configuration state if tsb_grow() is running in
parallel. That's easily solved with a seqlock taken as a writer by
tsb_grow() and taken as a reader to capture all the TSB config state
in tsb_context_switch().
Then there is flush_tsb_user() running in parallel with a tsb_grow().
In theory we could take the seqlock as a reader there too, and just
resample the TSB pointer and reflush but that looks really ugly.
Lastly, I believe there is a case with threads that results in a TSB
entry lock bit being set spuriously which will cause the next access
to that TSB entry to wedge the cpu (since the TSB entry lock bit will
never clear). It's either copy_tsb() or some bug elsewhere in the TSB
assembly.
Signed-off-by: David S. Miller <davem@davemloft.net>
The are distrupting, which by the sparc v9 definition means they
can only occur when interrupts are enabled in the %pstate register.
This never occurs in any of the trap handling code running at
trap levels > 0.
So just mark it as an unexpected trap.
This allows us to kill off the cee_stuff member of struct thread_info.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 23:55:21 +0000 (15:55 -0800)]
[SPARC64]: Access TSB with physical addresses when possible.
This way we don't need to lock the TSB into the TLB.
The trick is that every TSB load/store is registered into
a special instruction patch section. The default uses
virtual addresses, and the patch instructions use physical
address load/stores.
We can't do this on all chips because only cheetah+ and later
have the physical variant of the atomic quad load.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 02:35:05 +0000 (18:35 -0800)]
[SPARC64]: Don't clobber alt-global %g4 on window fixups.
If we are returning back to kernel mode, %g4 could be live
(for example, in the case where we window spill in the etrap
code). So do not change it's value if going back to kernel.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 02:34:51 +0000 (18:34 -0800)]
[SPARC64]: Fix race in LOAD_PER_CPU_BASE()
Since we use %g5 itself as a temporary, it can get clobbered
if we take an interrupt mid-stream and thus cause end up with
the final %g5 value too early as a result of rtrap processing.
Set %g5 at the very end, atomically, to avoid this problem.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 02:34:21 +0000 (18:34 -0800)]
[SPARC64]: Fix too early reference to %g6
%g6 is not necessarily set to current_thread_info()
at sparc64_realfault_common. So store the fault
code and address after we invoke etrap and %g6 is
properly set up.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 02:33:00 +0000 (18:33 -0800)]
[SPARC64]: Fix bogus flush instruction usage.
Some of the trap code was still assuming that alternate
global %g6 was hard coded with current_thread_info().
Let's just consistently flush at KERNBASE when we need
a pipeline synchronization. That's locked into the TLB
and will always work.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 02:31:20 +0000 (18:31 -0800)]
[SPARC64]: Add infrastructure for dynamic TSB sizing.
This also cleans up tsb_context_switch(). The assembler
routine is now __tsb_context_switch() and the former is
an inline function that picks out the bits from the mm_struct
and passes it into the assembler code as arguments.
setup_tsb_parms() computes the locked TLB entry to map the
TSB. Later when we support using the physical address quad
load instructions of Cheetah+ and later, we'll simply use
the physical address for the TSB register value and set
the map virtual and PTE both to zero.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 02:31:06 +0000 (18:31 -0800)]
[SPARC64]: TSB refinements.
Move {init_new,destroy}_context() out of line.
Do not put huge pages into the TSB, only base page size translations.
There are some clever things we could do here, but for now let's be
correct instead of fancy.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 27 Feb 2006 07:24:22 +0000 (23:24 -0800)]
[SPARC64]: Elminate all usage of hard-coded trap globals.
UltraSPARC has special sets of global registers which are switched to
for certain trap types. There is one set for MMU related traps, one
set of Interrupt Vector processing, and another set (called the
Alternate globals) for all other trap types.
For what seems like forever we've hard coded the values in some of
these trap registers. Some examples include:
1) Interrupt Vector global %g6 holds current processors interrupt
work struct where received interrupts are managed for IRQ handler
dispatch.
2) MMU global %g7 holds the base of the page tables of the currently
active address space.
3) Alternate global %g6 held the current_thread_info() value.
Such hardcoding has resulted in some serious issues in many areas.
There are some code sequences where having another register available
would help clean up the implementation. Taking traps such as
cross-calls from the OBP firmware requires some trick code sequences
wherein we have to save away and restore all of the special sets of
global registers when we enter/exit OBP.
We were also using the IMMU TSB register on SMP to hold the per-cpu
area base address, which doesn't work any longer now that we actually
use the TSB facility of the cpu.
The implementation is pretty straight forward. One tricky bit is
getting the current processor ID as that is different on different cpu
variants. We use a stub with a fancy calling convention which we
patch at boot time. The calling convention is that the stub is
branched to and the (PC - 4) to return to is in register %g1. The cpu
number is left in %g6. This stub can be invoked by using the
__GET_CPUID macro.
We use an array of per-cpu trap state to store the current thread and
physical address of the current address space's page tables. The
TRAP_LOAD_THREAD_REG loads %g6 with the current thread from this
table, it uses __GET_CPUID and also clobbers %g1.
TRAP_LOAD_IRQ_WORK is used by the interrupt vector processing to load
the current processor's IRQ software state into %g6. It also uses
__GET_CPUID and clobbers %g1.
Finally, TRAP_LOAD_PGD_PHYS loads the physical address base of the
current address space's page tables into %g7, it clobbers %g1 and uses
__GET_CPUID.
Many refinements are possible, as well as some tuning, with this stuff
in place.
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 1 Feb 2006 02:29:18 +0000 (18:29 -0800)]
[SPARC64]: Move away from virtual page tables, part 1.
We now use the TSB hardware assist features of the UltraSPARC
MMUs.
SMP is currently knowingly broken, we need to find another place
to store the per-cpu base pointers. We hid them away in the TSB
base register, and that obviously will not work any more :-)
Another known broken case is non-8KB base page size.
Also noticed that flush_tlb_all() is not referenced anywhere, only
the internal __flush_tlb_all() (local cpu only) is used by the
sparc64 port, so we can get rid of flush_tlb_all().
The kernel gets it's own 8KB TSB (swapper_tsb) and each address space
gets it's own private 8K TSB. Later we can add code to dynamically
increase the size of per-process TSB as the RSS grows. An 8KB TSB is
good enough for up to about a 4MB RSS, after which the TSB starts to
incur many capacity and conflict misses.
We even accumulate OBP translations into the kernel TSB.
Another area for refinement is large page size support. We could use
a secondary address space TSB to handle those.
Signed-off-by: David S. Miller <davem@davemloft.net>
The patch "[SPARC64]: Get rid of fast IRQ feature"
moved the the code from arch/sparc64/kernel/entry.S:
lduba [%g7] ASI_PHYS_BYPASS_EC_E, %g5
or %g5, AUXIO_AUX1_FTCNT, %g5
stba %g5, [%g7] ASI_PHYS_BYPASS_EC_E
andn %g5, AUXIO_AUX1_FTCNT, %g5
stba %g5, [%g7] ASI_PHYS_BYPASS_EC_E
to arch/sparc64/kernel/irq.c:
val = readb(auxio_register);
val |= AUXIO_AUX1_FTCNT;
writeb(val, auxio_register);
val &= AUXIO_AUX1_FTCNT;
writeb(val, auxio_register);
This looks like it it missing a bitwise not, which is reintroduced
by this patch.
Due to lack of a floppy device, I could not test it, but it looks
evident.
Signed-off-by: Bernhard R Link <brlink@debian.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Mon, 20 Mar 2006 05:12:00 +0000 (21:12 -0800)]
Merge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
[MIPS] SB1: Check for -mno-sched-prolog if building corelis debug kernel.
[MIPS] Sibyte: Fix race in sb1250_gettimeoffset().
[MIPS] Sibyte: Fix interrupt timer off by one bug.
[MIPS] Sibyte: Fix M_SCD_TIMER_INIT and M_SCD_TIMER_CNT wrong field width.
[MIPS] Protect more of timer_interrupt() by xtime_lock.
[MIPS] Work around bad code generation for <asm/io.h>.
[MIPS] Simple patch to power off DBAU1200
[MIPS] Fix DBAu1550 software power off.
[MIPS] local_r4k_flush_cache_page fix
[MIPS] SB1: Fix interrupt disable hazard.
[MIPS] Get rid of the IP22-specific code in arclib.
Update MAINTAINERS entry for MIPS.
Michael Chan [Sun, 19 Mar 2006 21:21:12 +0000 (13:21 -0800)]
[TG3]: 40-bit DMA workaround part 2
The 40-bit DMA workaround recently implemented for 5714, 5715, and
5780 needs to be expanded because there may be other tg3 devices
behind the EPB Express to PCIX bridge in the 5780 class device.
For example, some 4-port card or mother board designs have 5704 behind
the 5714.
All devices behind the EPB require the 40-bit DMA workaround.
Thanks to Chris Elmquist again for reporting the problem and testing
the patch.
Signed-off-by: Michael Chan <mchan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
If the AX.25 dialect chosen by the sysadmin is set to DAMA master / 3
(or DAMA slave / 2, if CONFIG_AX25_DAMA_SLAVE=n) ax25_kick() will fall
through the switch statement without calling ax25_send_iframe() or any
other function that would eventually free skbn thus leaking the packet.
Fix by restricting the sysctl inferface to allow only actually supported
AX.25 dialects.
The system administration mistake needed for this to happen is rather
unlikely, so this is an uncritical hole.
Coverity #651.
Signed-off-by: Ralf Baechle DL5RB <ralf@linux-mips.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Krufky [Wed, 15 Mar 2006 05:36:13 +0000 (02:36 -0300)]
[PATCH] Kconfig: swap VIDEO_CX88_ALSA and VIDEO_CX88_DVB
VIDEO_CX88_ALSA should not be between VIDEO_CX88_DVB and
VIDEO_CX88_DVB_ALL_FRONTENDS
When cx88-alsa was added to cx88/Kconfig, it was added in between
VIDEO_CX88_DVB and VIDEO_CX88_DVB_ALL_FRONTENDS. This caused
undesireable effects to the appearance of the menu options in
menuconfig.
This fix reorders cx88-alsa and cx88-dvb in Kconfig, to match saa7134,
and restore the correct menuconfig appearance.
Oleg Nesterov [Sat, 18 Mar 2006 17:41:10 +0000 (20:41 +0300)]
[PATCH] disable unshare(CLONE_VM) for now
sys_unshare() does mmput(new_mm). This is not enough if we have
mm->core_waiters.
This patch is a temporary fix for soon to be released 2.6.16.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
[ Checked with Uli: "I'm not planning to use unshare(CLONE_VM). It's
not needed for any functionality planned so far. What we (as in Red
Hat) need unshare() for now is the filesystem side." ] Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Ralf Baechle [Wed, 15 Mar 2006 00:03:29 +0000 (00:03 +0000)]
[MIPS] Sibyte: Fix race in sb1250_gettimeoffset().
From Dave Johnson <djohnson+linuxmips@sw.starentnetworks.com>:
sb1250_gettimeoffset() simply reads the current cpu 0 timer remaining
value, however once this counter reaches 0 and the interrupt is raised,
it immediately resets and begins to count down again.
If sb1250_gettimeoffset() is called on cpu 1 via do_gettimeofday() after
the timer has reset but prior to cpu 0 processing the interrupt and
taking write_seqlock() in timer_interrupt() it will return a full value
(or close to it) causing time to jump backwards 1ms. Once cpu 0 handles
the interrupt and timer_interrupt() gets far enough along it will jump
forward 1ms.
Fix this problem by implementing mips_hpt_*() on sb1250 using a spare
timer unrelated to the existing periodic interrupt timers. It runs at
1Mhz with a full 23bit counter. This eliminated the custom
do_gettimeoffset() for sb1250 and allowed use of the generic
fixed_rate_gettimeoffset() using mips_hpt_*() and timerhi/timerlo.
Ralf Baechle [Tue, 14 Mar 2006 23:46:58 +0000 (23:46 +0000)]
[MIPS] Protect more of timer_interrupt() by xtime_lock.
From Dave Johnson <djohnson+linuxmips@sw.starentnetworks.com>:
* do_timer() expects the arch-specific handler to take the lock as it
modifies jiffies[_64] and xtime.
* writing timerhi/lo in timer_interrupt() will mess up
fixed_rate_gettimeoffset() which reads timerhi/lo.
Ralf Baechle [Wed, 15 Mar 2006 11:36:31 +0000 (11:36 +0000)]
[MIPS] Work around bad code generation for <asm/io.h>.
If a call to set_io_port_base() was being followed by usage of
mips_io_port_base in the same function gcc was possibly using the old
value due to some clever abuse of const. Adding a barrier will keep
the optimization and result in correct code with latest gcc.
Atsushi Nemoto [Mon, 13 Mar 2006 09:23:03 +0000 (18:23 +0900)]
[MIPS] local_r4k_flush_cache_page fix
If dcache_size != icache_size or dcache_size != scache_size, or
set-associative cache, icache/scache does not flushed properly. Make
blast_?cache_page_indexed() masks its index value correctly. Also,
use physical address for physically indexed pcache/scache.
Hugh Dickins [Fri, 17 Mar 2006 07:04:09 +0000 (23:04 -0800)]
[PATCH] fix free swap cache latency
Lee Revell reported 28ms latency when process with lots of swapped memory
exits.
2.6.15 introduced a latency regression when unmapping: in accounting the
zap_work latency breaker, pte_none counted 1, pte_present PAGE_SIZE, but a
swap entry counted nothing at all. We think of pages present as the slow
case, but Lee's trace shows that free_swap_and_cache's radix tree lookup
can make a lot of work - and we could have been doing it many thousands of
times without a latency break.
Move the zap_work update up to account swap entries like pages present.
This does account non-linear pte_file entries, and unmap_mapping_range
skipping over swap entries, by the same amount even though they're quick:
but neither of those cases deserves complicating the code (and they're
treated no worse than they were in 2.6.14).
Signed-off-by: Hugh Dickins <hugh@veritas.com> Acked-by: Nick Piggin <npiggin@suse.de> Acked-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Sam Ravnborg [Fri, 17 Mar 2006 07:04:08 +0000 (23:04 -0800)]
[PATCH] kbuild: fix buffer overflow in modpost
Jiri Benc <jbenc@suse.cz> reported that modpost would stop with SIGABRT if
used with long filepaths.
The error looked like:
> Building modules, stage 2.
> MODPOST
> *** glibc detected *** scripts/mod/modpost: realloc(): invalid next size:
+0x0809f588 ***
> [...]
Fix this by allocating at least the required memory + SZ bytes each time.
Before we sometimes ended up allocating too little memory resuting in the
glibc detected bug above. Based on patch originally submitted by: Jiri
Benc <jbenc@suse.cz>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This can happen because the arguments to the nfsservctl() system call are
versioned. This is a good thing. However, when a bad version is detected,
the kernel prints a message and then returns an error.
Signed-off-by: Peter Staubach <staubach@redhat.com> Cc: Trond Myklebust <trond.myklebust@fys.uio.no> Cc: Neil Brown <neilb@cse.unsw.edu.au> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We can call try_to_release_page() with PagePrivate off and a valid
page->mapping This may cause all sorts of trouble for the filesystem
*_releasepage() handlers. XFS bombs out in that case.
Lock the page before checking for page private.
Signed-off-by: Christoph Lameter <clameter@sgi.com> Cc: Nick Piggin <nickpiggin@yahoo.com.au> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Kevin Corry [Fri, 17 Mar 2006 07:04:03 +0000 (23:04 -0800)]
[PATCH] dm stripe: Fix bounds
The dm-stripe target currently does not enforce that the size of a stripe
device be a multiple of the chunk-size. Under certain conditions, this can
lead to I/O requests going off the end of an underlying device. This
test-case shows one example.
This will produce the output:
dd: writing '/dev/mapper/stripe0': Input/output error
97+0 records in
96+0 records out
And in the kernel log will be:
attempt to access beyond end of device
dm-0: rw=0, want=104, limit=100
The patch will check that the table size is a multiple of the stripe
chunk-size when the table is created, which will prevent the above striped
device from being created.
This should not affect tools like LVM or EVMS, since in all the cases I can
think of, striped devices are always created with the sizes being a
multiple of the chunk-size.
The size of a stripe device must be a multiple of its chunk-size.
(akpm: that typecast is quite gratuitous)
Signed-off-by: Kevin Corry <kevcorry@us.ibm.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
[PATCH] x86: check for online cpus before bringing them up
Bryce reported a bug wherein offlining CPU0 (on x86 box) and then
subsequently onlining it resulted in a lockup.
On x86, CPU0 is never offlined. The subsequent attempt to online CPU0
doesn't take that into account. It actually tries to bootup the already
booted CPU. Following patch fixes the problem (as acknowledged by Bryce).
Please consider for inclusion in 2.6.16.
[PATCH] v9fs: fix overzealous dropping of dentry which breaks dcache
There is a d_drop in dir_release which caused problems as it invalidates
dcache entries too soon. This was likely a part of the wierd cwd behavior
folks were seeing.
Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Roman Zippel [Fri, 17 Mar 2006 07:04:01 +0000 (23:04 -0800)]
[PATCH] posix-timers: fix requeue accounting when signal is ignored
When the posix-timer signal is ignored then the timer is rearmed by the
callback function. The requeue pending accounting has to be fixed up else
the state might be wrong.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
projects / linux-2.6 / log