Skip to content

K
kernel_linux

OpenHarmony / kernel_linux
上一次同步 接近 4 年

通知 14
Star 8
Fork 2
K
kernel_linux
切换分支/标签
  1. 13 3月, 2012 4 次提交
    • S
      sched/x86: Fix overflow in cyc2ns_offset · 9993bc63
      由 Salman Qazi 提交于 
      When a machine boots up, the TSC generally gets reset.  However,
when kexec is used to boot into a kernel, the TSC value would be
carried over from the previous kernel.  The computation of
cycns_offset in set_cyc2ns_scale is prone to an overflow, if the
machine has been up more than 208 days prior to the kexec.  The
overflow happens when we multiply *scale, even though there is
enough room to store the final answer.

We fix this issue by decomposing tsc_now into the quotient and
remainder of division by CYC2NS_SCALE_FACTOR and then performing
the multiplication separately on the two components.

Refactor code to share the calculation with the previous
fix in __cycles_2_ns().
Signed-off-by: NSalman Qazi <sqazi@google.com>
Acked-by: NJohn Stultz <john.stultz@linaro.org>
Acked-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Turner <pjt@google.com>
Cc: john stultz <johnstul@us.ibm.com>
Link: http://lkml.kernel.org/r/20120310004027.19291.88460.stgit@dungbeetle.mtv.corp.google.comSigned-off-by: NIngo Molnar <mingo@elte.hu>
      9993bc63
    • P
      perf/x86: Prettify pmu config literals · f9b4eeb8
      由 Peter Zijlstra 提交于 
      I got somewhat tired of having to decode hex numbers..
Signed-off-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Stephane Eranian <eranian@google.com>
Cc: Robert Richter <robert.richter@amd.com>
Link: http://lkml.kernel.org/n/tip-0vsy1sgywc4uar3mu1szm0rg@git.kernel.orgSigned-off-by: NIngo Molnar <mingo@elte.hu>
      f9b4eeb8
    • P
      perf/x86: Fix local vs remote memory events for NHM/WSM · 87e24f4b
      由 Peter Zijlstra 提交于 
      Verified using the below proglet.. before:

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 0
remote write

 Performance counter stats for './numa 0':

         2,101,554 node-stores
         2,096,931 node-store-misses

       5.021546079 seconds time elapsed

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 1
local write

 Performance counter stats for './numa 1':

           501,137 node-stores
               199 node-store-misses

       5.124451068 seconds time elapsed

After:

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 0
remote write

 Performance counter stats for './numa 0':

         2,107,516 node-stores
         2,097,187 node-store-misses

       5.012755149 seconds time elapsed

[root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 1
local write

 Performance counter stats for './numa 1':

         2,063,355 node-stores
               165 node-store-misses

       5.082091494 seconds time elapsed

#define _GNU_SOURCE

#include <sched.h>
#include <stdio.h>
#include <errno.h>
#include <sys/mman.h>
#include <sys/types.h>
#include <dirent.h>
#include <signal.h>
#include <unistd.h>
#include <numaif.h>
#include <stdlib.h>

#define SIZE (32*1024*1024)

volatile int done;

void sig_done(int sig)
{
	done = 1;
}

int main(int argc, char **argv)
{
	cpu_set_t *mask, *mask2;
	size_t size;
	int i, err, t;
	int nrcpus = 1024;
	char *mem;
	unsigned long nodemask = 0x01; /* node 0 */
	DIR *node;
	struct dirent *de;
	int read = 0;
	int local = 0;

	if (argc < 2) {
		printf("usage: %s [0-3]\n", argv[0]);
		printf("  bit0 - local/remote\n");
		printf("  bit1 - read/write\n");
		exit(0);
	}

	switch (atoi(argv[1])) {
	case 0:
		printf("remote write\n");
		break;
	case 1:
		printf("local write\n");
		local = 1;
		break;
	case 2:
		printf("remote read\n");
		read = 1;
		break;
	case 3:
		printf("local read\n");
		local = 1;
		read = 1;
		break;
	}

	mask = CPU_ALLOC(nrcpus);
	size = CPU_ALLOC_SIZE(nrcpus);
	CPU_ZERO_S(size, mask);

	node = opendir("/sys/devices/system/node/node0/");
	if (!node)
		perror("opendir");
	while ((de = readdir(node))) {
		int cpu;

		if (sscanf(de->d_name, "cpu%d", &cpu) == 1)
			CPU_SET_S(cpu, size, mask);
	}
	closedir(node);

	mask2 = CPU_ALLOC(nrcpus);
	CPU_ZERO_S(size, mask2);
	for (i = 0; i < size; i++)
		CPU_SET_S(i, size, mask2);
	CPU_XOR_S(size, mask2, mask2, mask); // invert

	if (!local)
		mask = mask2;

	err = sched_setaffinity(0, size, mask);
	if (err)
		perror("sched_setaffinity");

	mem = mmap(0, SIZE, PROT_READ|PROT_WRITE,
			MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
	err = mbind(mem, SIZE, MPOL_BIND, &nodemask, 8*sizeof(nodemask), MPOL_MF_MOVE);
	if (err)
		perror("mbind");

	signal(SIGALRM, sig_done);
	alarm(5);

	if (!read) {
		while (!done) {
			for (i = 0; i < SIZE; i++)
				mem[i] = 0x01;
		}
	} else {
		while (!done) {
			for (i = 0; i < SIZE; i++)
				t += *(volatile char *)(mem + i);
		}
	}

	return 0;
}
Signed-off-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Cc: <stable@kernel.org>
Link: http://lkml.kernel.org/n/tip-tq73sxus35xmqpojf7ootxgs@git.kernel.orgSigned-off-by: NIngo Molnar <mingo@elte.hu>
      87e24f4b
    • P
      sched: Cleanup cpu_active madness · 5fbd036b
      由 Peter Zijlstra 提交于 
      Stepan found:

CPU0		CPUn

_cpu_up()
  __cpu_up()

		boostrap()
		  notify_cpu_starting()
		  set_cpu_online()
		  while (!cpu_active())
		    cpu_relax()

<PREEMPT-out>

smp_call_function(.wait=1)
  /* we find cpu_online() is true */
  arch_send_call_function_ipi_mask()

  /* wait-forever-more */

<PREEMPT-in>
		  local_irq_enable()

  cpu_notify(CPU_ONLINE)
    sched_cpu_active()
      set_cpu_active()

Now the purpose of cpu_active is mostly with bringing down a cpu, where
we mark it !active to avoid the load-balancer from moving tasks to it
while we tear down the cpu. This is required because we only update the
sched_domain tree after we brought the cpu-down. And this is needed so
that some tasks can still run while we bring it down, we just don't want
new tasks to appear.

On cpu-up however the sched_domain tree doesn't yet include the new cpu,
so its invisible to the load-balancer, regardless of the active state.
So instead of setting the active state after we boot the new cpu (and
consequently having to wait for it before enabling interrupts) set the
cpu active before we set it online and avoid the whole mess.
Reported-by: NStepan Moskovchenko <stepanm@codeaurora.org>
Signed-off-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: NThomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1323965362.18942.71.camel@twinsSigned-off-by: NIngo Molnar <mingo@elte.hu>
      5fbd036b
  3. 06 3月, 2012 3 次提交
    • M
      x86/kprobes: Split out optprobe related code to kprobes-opt.c · 3f33ab1c
      由 Masami Hiramatsu 提交于 
      Split out optprobe related code to arch/x86/kernel/kprobes-opt.c
for maintenanceability.
Signed-off-by: NMasami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Suggested-by: NIngo Molnar <mingo@elte.hu>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: yrl.pp-manager.tt@hitachi.com
Cc: systemtap@sourceware.org
Cc: anderson@redhat.com
Link: http://lkml.kernel.org/r/20120305133222.5982.54794.stgit@localhost.localdomain
[ Tidied up the code a tiny bit ]
Signed-off-by: NIngo Molnar <mingo@elte.hu>
      3f33ab1c
    • M
      x86/kprobes: Fix a bug which can modify kernel code permanently · 46484688
      由 Masami Hiramatsu 提交于 
      Fix a bug in kprobes which can modify kernel code
permanently at run-time. In the result, kernel can
crash when it executes the modified code.

This bug can happen when we put two probes enough near
and the first probe is optimized. When the second probe
is set up, it copies a byte which is already modified
by the first probe, and executes it when the probe is hit.
Even worse, the first probe and the second probe are removed
respectively, the second probe writes back the copied
(modified) instruction.

To fix this bug, kprobes always recovers the original
code and copies the first byte from recovered instruction.
Signed-off-by: NMasami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: yrl.pp-manager.tt@hitachi.com
Cc: systemtap@sourceware.org
Cc: anderson@redhat.com
Link: http://lkml.kernel.org/r/20120305133215.5982.31991.stgit@localhost.localdomainSigned-off-by: NIngo Molnar <mingo@elte.hu>
      46484688
    • M
      x86/kprobes: Fix instruction recovery on optimized path · 86b4ce31
      由 Masami Hiramatsu 提交于 
      Current probed-instruction recovery expects that only breakpoint
instruction modifies instruction. However, since kprobes jump
optimization can replace original instructions with a jump,
that expectation is not enough. And it may cause instruction
decoding failure on the function where an optimized probe
already exists.

This bug can reproduce easily as below:

1) find a target function address (any kprobe-able function is OK)

 $ grep __secure_computing /proc/kallsyms
   ffffffff810c19d0 T __secure_computing

2) decode the function
   $ objdump -d vmlinux --start-address=0xffffffff810c19d0 --stop-address=0xffffffff810c19eb

  vmlinux:     file format elf64-x86-64

Disassembly of section .text:

ffffffff810c19d0 <__secure_computing>:
ffffffff810c19d0:       55                      push   %rbp
ffffffff810c19d1:       48 89 e5                mov    %rsp,%rbp
ffffffff810c19d4:       e8 67 8f 72 00          callq
ffffffff817ea940 <mcount>
ffffffff810c19d9:       65 48 8b 04 25 40 b8    mov    %gs:0xb840,%rax
ffffffff810c19e0:       00 00
ffffffff810c19e2:       83 b8 88 05 00 00 01    cmpl $0x1,0x588(%rax)
ffffffff810c19e9:       74 05                   je     ffffffff810c19f0 <__secure_computing+0x20>

3) put a kprobe-event at an optimize-able place, where no
 call/jump places within the 5 bytes.
 $ su -
 # cd /sys/kernel/debug/tracing
 # echo p __secure_computing+0x9 > kprobe_events

4) enable it and check it is optimized.
 # echo 1 > events/kprobes/p___secure_computing_9/enable
 # cat ../kprobes/list
 ffffffff810c19d9  k  __secure_computing+0x9    [OPTIMIZED]

5) put another kprobe on an instruction after previous probe in
  the same function.
 # echo p __secure_computing+0x12 >> kprobe_events
 bash: echo: write error: Invalid argument
 # dmesg | tail -n 1
 [ 1666.500016] Probing address(0xffffffff810c19e2) is not an instruction boundary.

6) however, if the kprobes optimization is disabled, it works.
 # echo 0 > /proc/sys/debug/kprobes-optimization
 # cat ../kprobes/list
 ffffffff810c19d9  k  __secure_computing+0x9
 # echo p __secure_computing+0x12 >> kprobe_events
 (no error)

This is because kprobes doesn't recover the instruction
which is overwritten with a relative jump by another kprobe
when finding instruction boundary.
It only recovers the breakpoint instruction.

This patch fixes kprobes to recover such instructions.

With this fix:

 # echo p __secure_computing+0x9 > kprobe_events
 # echo 1 > events/kprobes/p___secure_computing_9/enable
 # cat ../kprobes/list
 ffffffff810c1aa9  k  __secure_computing+0x9    [OPTIMIZED]
 # echo p __secure_computing+0x12 >> kprobe_events
 # cat ../kprobes/list
 ffffffff810c1aa9  k  __secure_computing+0x9    [OPTIMIZED]
 ffffffff810c1ab2  k  __secure_computing+0x12    [DISABLED]

Changes in v4:
 - Fix a bug to ensure optimized probe is really optimized
   by jump.
 - Remove kprobe_optready() dependency.
 - Cleanup code for preparing optprobe separation.

Changes in v3:
 - Fix a build error when CONFIG_OPTPROBE=n. (Thanks, Ingo!)
   To fix the error, split optprobe instruction recovering
   path from kprobes path.
 - Cleanup comments/styles.

Changes in v2:
 - Fix a bug to recover original instruction address in
   RIP-relative instruction fixup.
 - Moved on tip/master.
Signed-off-by: NMasami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: yrl.pp-manager.tt@hitachi.com
Cc: systemtap@sourceware.org
Cc: anderson@redhat.com
Link: http://lkml.kernel.org/r/20120305133209.5982.36568.stgit@localhost.localdomainSigned-off-by: NIngo Molnar <mingo@elte.hu>
      86b4ce31
  5. 05 3月, 2012 10 次提交
  7. 02 3月, 2012 1 次提交
  9. 01 3月, 2012 1 次提交
  11. 27 2月, 2012 1 次提交
  13. 24 2月, 2012 1 次提交
    • I
      static keys: Introduce 'struct static_key', static_key_true()/false() and... · c5905afb
      由 Ingo Molnar 提交于 
      static keys: Introduce 'struct static_key', static_key_true()/false() and static_key_slow_[inc|dec]()

So here's a boot tested patch on top of Jason's series that does
all the cleanups I talked about and turns jump labels into a
more intuitive to use facility. It should also address the
various misconceptions and confusions that surround jump labels.

Typical usage scenarios:

        #include <linux/static_key.h>

        struct static_key key = STATIC_KEY_INIT_TRUE;

        if (static_key_false(&key))
                do unlikely code
        else
                do likely code

Or:

        if (static_key_true(&key))
                do likely code
        else
                do unlikely code

The static key is modified via:

        static_key_slow_inc(&key);
        ...
        static_key_slow_dec(&key);

The 'slow' prefix makes it abundantly clear that this is an
expensive operation.

I've updated all in-kernel code to use this everywhere. Note
that I (intentionally) have not pushed through the rename
blindly through to the lowest levels: the actual jump-label
patching arch facility should be named like that, so we want to
decouple jump labels from the static-key facility a bit.

On non-jump-label enabled architectures static keys default to
likely()/unlikely() branches.
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Acked-by: NJason Baron <jbaron@redhat.com>
Acked-by: NSteven Rostedt <rostedt@goodmis.org>
Cc: a.p.zijlstra@chello.nl
Cc: mathieu.desnoyers@efficios.com
Cc: davem@davemloft.net
Cc: ddaney.cavm@gmail.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20120222085809.GA26397@elte.huSigned-off-by: NIngo Molnar <mingo@elte.hu>
      c5905afb
  15. 22 2月, 2012 2 次提交
    • B
      x86/mce/AMD: Fix UP build error · 3f806e50
      由 Borislav Petkov 提交于 
      141168c3 ("x86: Simplify code by removing a !SMP #ifdefs
from 'struct cpuinfo_x86'") removed a bunch of CONFIG_SMP ifdefs
around code touching struct cpuinfo_x86 members but also caused
the following build error with Randy's randconfigs:

mce_amd.c:(.cpuinit.text+0x4723): undefined reference to `cpu_llc_shared_map'

Restore the #ifdef in threshold_create_bank() which creates
symlinks on the non-BSP CPUs.

There's a better patch series being worked on by Kevin Winchester
which will solve this in a cleaner fashion, but that series is
too ambitious for v3.3 merging - so we first queue up this trivial
fix and then do the rest for v3.4.
Signed-off-by: NBorislav Petkov <bp@alien8.de>
Acked-by: NKevin Winchester <kjwinchester@gmail.com>
Cc: Randy Dunlap <rdunlap@xenotime.net>
Cc: Nick Bowler <nbowler@elliptictech.com>
Link: http://lkml.kernel.org/r/20120203191801.GA2846@x1.osrc.amd.comSigned-off-by: NIngo Molnar <mingo@elte.hu>
      3f806e50
    • S
      x86/tsc: Reduce the TSC sync check time for core-siblings · b0e5c779
      由 Suresh Siddha 提交于 
      For each logical CPU that is coming online, we spend 20msec for
checking the TSC synchronization. And as this is done
sequentially for each logical CPU boot, this time gets added up
depending on the number of logical CPU's supported by the
platform.

Minimize this by using the socket topology information.

If the target CPU coming online doesn't have any of its
core-siblings online, a timeout of 20msec will be used for the
TSC-warp measurement loop. Otherwise a smaller timeout of 2msec
will be used, as we have some information about this socket
already (and this information grows as we have more and more
logical-siblings in that socket).

Ideally we should be able to skip the TSC sync check on the
other core-siblings, if the first logical CPU in a socket passed
the sync test. But as the TSC is per-logical CPU and can
potentially be modified wrongly by the bios before the OS boot,
TSC sync test for smaller duration should be able to catch such
errors. Also this will catch the condition where all the cores
in the socket doesn't get reset at the same time.

For example, with this modification, time spent in TSC sync
checks on a 4 socket 10-core with HT system gets reduced from
1580msec to 212msec.
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Acked-by: NArjan van de Ven <arjan@linux.intel.com>
Acked-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Jack Steiner <steiner@sgi.com>
Cc: venki@google.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/1328581940.29790.20.camel@sbsiddha-desk.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>
      b0e5c779
  17. 21 2月, 2012 5 次提交
    • L
      i387: export 'fpu_owner_task' per-cpu variable · 27e74da9
      由 Linus Torvalds 提交于 
      (And define it properly for x86-32, which had its 'current_task'
declaration in separate from x86-64)

Bitten by my dislike for modules on the machines I use, and the fact
that apparently nobody else actually wanted to test the patches I sent
out.

Snif. Nobody else cares.

Anyway, we probably should uninline the 'kernel_fpu_begin()' function
that is what modules actually use and that references this, but this is
the minimal fix for now.
Reported-by: NJosh Boyer <jwboyer@gmail.com>
Reported-and-tested-by: NJongman Heo <jongman.heo@samsung.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      27e74da9
    • S
      x86: Specify a size for the cmp in the NMI handler · a38449ef
      由 Steven Rostedt 提交于 
      Linus noticed that the cmp used to check if the code segment is
__KERNEL_CS or not did not specify a size. Perhaps it does not matter
as H. Peter Anvin noted that user space can not set the bottom two
bits of the %cs register. But it's best not to let the assembly choose
and change things between different versions of gas, but instead just
pick the size.

Four bytes are used to compare the saved code segment against
__KERNEL_CS. Perhaps this might mess up Xen, but we can fix that when
the time comes.

Also I noticed that there was another non-specified cmp that checks
the special stack variable if it is 1 or 0. This too probably doesn't
matter what cmp is used, but this patch uses cmpl just to make it non
ambiguous.

Link: http://lkml.kernel.org/r/CA+55aFxfAn9MWRgS3O5k2tqN5ys1XrhSFVO5_9ZAoZKDVgNfGA@mail.gmail.comSuggested-by: NLinus Torvalds <torvalds@linux-foundation.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>
      a38449ef
    • L
      i387: support lazy restore of FPU state · 7e16838d
      由 Linus Torvalds 提交于 
      This makes us recognize when we try to restore FPU state that matches
what we already have in the FPU on this CPU, and avoids the restore
entirely if so.

To do this, we add two new data fields:

 - a percpu 'fpu_owner_task' variable that gets written any time we
   update the "has_fpu" field, and thus acts as a kind of back-pointer
   to the task that owns the CPU.  The exception is when we save the FPU
   state as part of a context switch - if the save can keep the FPU
   state around, we leave the 'fpu_owner_task' variable pointing at the
   task whose FP state still remains on the CPU.

 - a per-thread 'last_cpu' field, that indicates which CPU that thread
   used its FPU on last.  We update this on every context switch
   (writing an invalid CPU number if the last context switch didn't
   leave the FPU in a lazily usable state), so we know that *that*
   thread has done nothing else with the FPU since.

These two fields together can be used when next switching back to the
task to see if the CPU still matches: if 'fpu_owner_task' matches the
task we are switching to, we know that no other task (or kernel FPU
usage) touched the FPU on this CPU in the meantime, and if the current
CPU number matches the 'last_cpu' field, we know that this thread did no
other FP work on any other CPU, so the FPU state on the CPU must match
what was saved on last context switch.

In that case, we can avoid the 'f[x]rstor' entirely, and just clear the
CR0.TS bit.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      7e16838d
    • L
      i387: use 'restore_fpu_checking()' directly in task switching code · 80ab6f1e
      由 Linus Torvalds 提交于 
      This inlines what is usually just a couple of instructions, but more
importantly it also fixes the theoretical error case (can that FPU
restore really ever fail? Maybe we should remove the checking).

We can't start sending signals from within the scheduler, we're much too
deep in the kernel and are holding the runqueue lock etc.  So don't
bother even trying.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      80ab6f1e
    • L
      i387: fix up some fpu_counter confusion · cea20ca3
      由 Linus Torvalds 提交于 
      This makes sure we clear the FPU usage counter for newly created tasks,
just so that we start off in a known state (for example, don't try to
preload the FPU state on the first task switch etc).

It also fixes a thinko in when we increment the fpu_counter at task
switch time, introduced by commit 34ddc81a ("i387: re-introduce FPU
state preloading at context switch time").  We should increment the
*new* task fpu_counter, not the old task, and only if we decide to use
that state (whether lazily or preloaded).
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      cea20ca3
  19. 20 2月, 2012 2 次提交
  21. 19 2月, 2012 2 次提交
    • L
      i387: re-introduce FPU state preloading at context switch time · 34ddc81a
      由 Linus Torvalds 提交于 
      After all the FPU state cleanups and finally finding the problem that
caused all our FPU save/restore problems, this re-introduces the
preloading of FPU state that was removed in commit b3b0870e ("i387:
do not preload FPU state at task switch time").

However, instead of simply reverting the removal, this reimplements
preloading with several fixes, most notably

 - properly abstracted as a true FPU state switch, rather than as
   open-coded save and restore with various hacks.

   In particular, implementing it as a proper FPU state switch allows us
   to optimize the CR0.TS flag accesses: there is no reason to set the
   TS bit only to then almost immediately clear it again.  CR0 accesses
   are quite slow and expensive, don't flip the bit back and forth for
   no good reason.

 - Make sure that the same model works for both x86-32 and x86-64, so
   that there are no gratuitous differences between the two due to the
   way they save and restore segment state differently due to
   architectural differences that really don't matter to the FPU state.

 - Avoid exposing the "preload" state to the context switch routines,
   and in particular allow the concept of lazy state restore: if nothing
   else has used the FPU in the meantime, and the process is still on
   the same CPU, we can avoid restoring state from memory entirely, just
   re-expose the state that is still in the FPU unit.

   That optimized lazy restore isn't actually implemented here, but the
   infrastructure is set up for it.  Of course, older CPU's that use
   'fnsave' to save the state cannot take advantage of this, since the
   state saving also trashes the state.

In other words, there is now an actual _design_ to the FPU state saving,
rather than just random historical baggage.  Hopefully it's easier to
follow as a result.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      34ddc81a
    • L
      i387: move TS_USEDFPU flag from thread_info to task_struct · f94edacf
      由 Linus Torvalds 提交于 
      This moves the bit that indicates whether a thread has ownership of the
FPU from the TS_USEDFPU bit in thread_info->status to a word of its own
(called 'has_fpu') in task_struct->thread.has_fpu.

This fixes two independent bugs at the same time:

 - changing 'thread_info->status' from the scheduler causes nasty
   problems for the other users of that variable, since it is defined to
   be thread-synchronous (that's what the "TS_" part of the naming was
   supposed to indicate).

   So perfectly valid code could (and did) do

	ti->status |= TS_RESTORE_SIGMASK;

   and the compiler was free to do that as separate load, or and store
   instructions.  Which can cause problems with preemption, since a task
   switch could happen in between, and change the TS_USEDFPU bit. The
   change to TS_USEDFPU would be overwritten by the final store.

   In practice, this seldom happened, though, because the 'status' field
   was seldom used more than once, so gcc would generally tend to
   generate code that used a read-modify-write instruction and thus
   happened to avoid this problem - RMW instructions are naturally low
   fat and preemption-safe.

 - On x86-32, the current_thread_info() pointer would, during interrupts
   and softirqs, point to a *copy* of the real thread_info, because
   x86-32 uses %esp to calculate the thread_info address, and thus the
   separate irq (and softirq) stacks would cause these kinds of odd
   thread_info copy aliases.

   This is normally not a problem, since interrupts aren't supposed to
   look at thread information anyway (what thread is running at
   interrupt time really isn't very well-defined), but it confused the
   heck out of irq_fpu_usable() and the code that tried to squirrel
   away the FPU state.

   (It also caused untold confusion for us poor kernel developers).

It also turns out that using 'task_struct' is actually much more natural
for most of the call sites that care about the FPU state, since they
tend to work with the task struct for other reasons anyway (ie
scheduling).  And the FPU data that we are going to save/restore is
found there too.

Thanks to Arjan Van De Ven <arjan@linux.intel.com> for pointing us to
the %esp issue.

Cc: Arjan van de Ven <arjan@linux.intel.com>
Reported-and-tested-by: NRaphael Prevost <raphael@buro.asia>
Acked-and-tested-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Tested-by: NPeter Anvin <hpa@zytor.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      f94edacf
  23. 17 2月, 2012 4 次提交
    • L
      i387: move AMD K7/K8 fpu fxsave/fxrstor workaround from save to restore · 4903062b
      由 Linus Torvalds 提交于 
      The AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception is
pending.  In order to not leak FIP state from one process to another, we
need to do a floating point load after the fxsave of the old process,
and before the fxrstor of the new FPU state.  That resets the state to
the (uninteresting) kernel load, rather than some potentially sensitive
user information.

We used to do this directly after the FPU state save, but that is
actually very inconvenient, since it

 (a) corrupts what is potentially perfectly good FPU state that we might
     want to lazy avoid restoring later and

 (b) on x86-64 it resulted in a very annoying ordering constraint, where
     "__unlazy_fpu()" in the task switch needs to be delayed until after
     the DS segment has been reloaded just to get the new DS value.

Coupling it to the fxrstor instead of the fxsave automatically avoids
both of these issues, and also ensures that we only do it when actually
necessary (the FP state after a save may never actually get used).  It's
simply a much more natural place for the leaked state cleanup.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      4903062b
    • L
      i387: do not preload FPU state at task switch time · b3b0870e
      由 Linus Torvalds 提交于 
      Yes, taking the trap to re-load the FPU/MMX state is expensive, but so
is spending several days looking for a bug in the state save/restore
code.  And the preload code has some rather subtle interactions with
both paravirtualization support and segment state restore, so it's not
nearly as simple as it should be.

Also, now that we no longer necessarily depend on a single bit (ie
TS_USEDFPU) for keeping track of the state of the FPU, we migth be able
to do better.  If we are really switching between two processes that
keep touching the FP state, save/restore is inevitable, but in the case
of having one process that does most of the FPU usage, we may actually
be able to do much better than the preloading.

In particular, we may be able to keep track of which CPU the process ran
on last, and also per CPU keep track of which process' FP state that CPU
has.  For modern CPU's that don't destroy the FPU contents on save time,
that would allow us to do a lazy restore by just re-enabling the
existing FPU state - with no restore cost at all!
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      b3b0870e
    • L
      i387: don't ever touch TS_USEDFPU directly, use helper functions · 6d59d7a9
      由 Linus Torvalds 提交于 
      This creates three helper functions that do the TS_USEDFPU accesses, and
makes everybody that used to do it by hand use those helpers instead.

In addition, there's a couple of helper functions for the "change both
CR0.TS and TS_USEDFPU at the same time" case, and the places that do
that together have been changed to use those.  That means that we have
fewer random places that open-code this situation.

The intent is partly to clarify the code without actually changing any
semantics yet (since we clearly still have some hard to reproduce bug in
this area), but also to make it much easier to use another approach
entirely to caching the CR0.TS bit for software accesses.

Right now we use a bit in the thread-info 'status' variable (this patch
does not change that), but we might want to make it a full field of its
own or even make it a per-cpu variable.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      6d59d7a9
    • L
      i387: fix x86-64 preemption-unsafe user stack save/restore · 15d8791c
      由 Linus Torvalds 提交于 
      Commit 5b1cbac3 ("i387: make irq_fpu_usable() tests more robust")
added a sanity check to the #NM handler to verify that we never cause
the "Device Not Available" exception in kernel mode.

However, that check actually pinpointed a (fundamental) race where we do
cause that exception as part of the signal stack FPU state save/restore
code.

Because we use the floating point instructions themselves to save and
restore state directly from user mode, we cannot do that atomically with
testing the TS_USEDFPU bit: the user mode access itself may cause a page
fault, which causes a task switch, which saves and restores the FP/MMX
state from the kernel buffers.

This kind of "recursive" FP state save is fine per se, but it means that
when the signal stack save/restore gets restarted, it will now take the
'#NM' exception we originally tried to avoid.  With preemption this can
happen even without the page fault - but because of the user access, we
cannot just disable preemption around the save/restore instruction.

There are various ways to solve this, including using the
"enable/disable_page_fault()" helpers to not allow page faults at all
during the sequence, and fall back to copying things by hand without the
use of the native FP state save/restore instructions.

However, the simplest thing to do is to just allow the #NM from kernel
space, but fix the race in setting and clearing CR0.TS that this all
exposed: the TS bit changes and the TS_USEDFPU bit absolutely have to be
atomic wrt scheduling, so while the actual state save/restore can be
interrupted and restarted, the act of actually clearing/setting CR0.TS
and the TS_USEDFPU bit together must not.

Instead of just adding random "preempt_disable/enable()" calls to what
is already excessively ugly code, this introduces some helper functions
that mostly mirror the "kernel_fpu_begin/end()" functionality, just for
the user state instead.

Those helper functions should probably eventually replace the other
ad-hoc CR0.TS and TS_USEDFPU tests too, but I'll need to think about it
some more: the task switching functionality in particular needs to
expose the difference between the 'prev' and 'next' threads, while the
new helper functions intentionally were written to only work with
'current'.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      15d8791c
  25. 14 2月, 2012 2 次提交
    • L
      i387: make irq_fpu_usable() tests more robust · 5b1cbac3
      由 Linus Torvalds 提交于 
      Some code - especially the crypto layer - wants to use the x86
FP/MMX/AVX register set in what may be interrupt (typically softirq)
context.

That *can* be ok, but the tests for when it was ok were somewhat
suspect.  We cannot touch the thread-specific status bits either, so
we'd better check that we're not going to try to save FP state or
anything like that.

Now, it may be that the TS bit is always cleared *before* we set the
USEDFPU bit (and only set when we had already cleared the USEDFP
before), so the TS bit test may actually have been sufficient, but it
certainly was not obviously so.

So this explicitly verifies that we will not touch the TS_USEDFPU bit,
and adds a few related sanity-checks.  Because it seems that somehow
AES-NI is corrupting user FP state.  The cause is not clear, and this
patch doesn't fix it, but while debugging it I really wanted the code to
be more obviously correct and robust.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      5b1cbac3
    • L
      i387: math_state_restore() isn't called from asm · be98c2cd
      由 Linus Torvalds 提交于 
      It was marked asmlinkage for some really old and stale legacy reasons.
Fix that and the equally stale comment.

Noticed when debugging the irq_fpu_usable() bugs.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      be98c2cd
  27. 13 2月, 2012 1 次提交
  29. 09 2月, 2012 1 次提交