Revision 63cae12bce9861cec309798d34701cf3da20bc71 authored by Peter Zijlstra on 09 December 2016, 13:59:00 UTC, committed by Ingo Molnar on 14 January 2017, 09:56:10 UTC
There is problem with installing an event in a task that is 'stuck' on
an offline CPU.
Blocked tasks are not dis-assosciated from offlined CPUs, after all, a
blocked task doesn't run and doesn't require a CPU etc.. Only on
wakeup do we ammend the situation and place the task on a available
CPU.
If we hit such a task with perf_install_in_context() we'll loop until
either that task wakes up or the CPU comes back online, if the task
waking depends on the event being installed, we're stuck.
While looking into this issue, I also spotted another problem, if we
hit a task with perf_install_in_context() that is in the middle of
being migrated, that is we observe the old CPU before sending the IPI,
but run the IPI (on the old CPU) while the task is already running on
the new CPU, things also go sideways.
Rework things to rely on task_curr() -- outside of rq->lock -- which
is rather tricky. Imagine the following scenario where we're trying to
install the first event into our task 't':
CPU0 CPU1 CPU2
(current == t)
t->perf_event_ctxp[] = ctx;
smp_mb();
cpu = task_cpu(t);
switch(t, n);
migrate(t, 2);
switch(p, t);
ctx = t->perf_event_ctxp[]; // must not be NULL
smp_function_call(cpu, ..);
generic_exec_single()
func();
spin_lock(ctx->lock);
if (task_curr(t)) // false
add_event_to_ctx();
spin_unlock(ctx->lock);
perf_event_context_sched_in();
spin_lock(ctx->lock);
// sees event
So its CPU0's store of t->perf_event_ctxp[] that must not go 'missing'.
Because if CPU2's load of that variable were to observe NULL, it would
not try to schedule the ctx and we'd have a task running without its
counter, which would be 'bad'.
As long as we observe !NULL, we'll acquire ctx->lock. If we acquire it
first and not see the event yet, then CPU0 must observe task_curr()
and retry. If the install happens first, then we must see the event on
sched-in and all is well.
I think we can translate the first part (until the 'must not be NULL')
of the scenario to a litmus test like:
C C-peterz
{
}
P0(int *x, int *y)
{
int r1;
WRITE_ONCE(*x, 1);
smp_mb();
r1 = READ_ONCE(*y);
}
P1(int *y, int *z)
{
WRITE_ONCE(*y, 1);
smp_store_release(z, 1);
}
P2(int *x, int *z)
{
int r1;
int r2;
r1 = smp_load_acquire(z);
smp_mb();
r2 = READ_ONCE(*x);
}
exists
(0:r1=0 /\ 2:r1=1 /\ 2:r2=0)
Where:
x is perf_event_ctxp[],
y is our tasks's CPU, and
z is our task being placed on the rq of CPU2.
The P0 smp_mb() is the one added by this patch, ordering the store to
perf_event_ctxp[] from find_get_context() and the load of task_cpu()
in task_function_call().
The smp_store_release/smp_load_acquire model the RCpc locking of the
rq->lock and the smp_mb() of P2 is the context switch switching from
whatever CPU2 was running to our task 't'.
This litmus test evaluates into:
Test C-peterz Allowed
States 7
0:r1=0; 2:r1=0; 2:r2=0;
0:r1=0; 2:r1=0; 2:r2=1;
0:r1=0; 2:r1=1; 2:r2=1;
0:r1=1; 2:r1=0; 2:r2=0;
0:r1=1; 2:r1=0; 2:r2=1;
0:r1=1; 2:r1=1; 2:r2=0;
0:r1=1; 2:r1=1; 2:r2=1;
No
Witnesses
Positive: 0 Negative: 7
Condition exists (0:r1=0 /\ 2:r1=1 /\ 2:r2=0)
Observation C-peterz Never 0 7
Hash=e427f41d9146b2a5445101d3e2fcaa34
And the strong and weak model agree.
Reported-by: Mark Rutland <mark.rutland@arm.com>
Tested-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Will Deacon <will.deacon@arm.com>
Cc: jeremy.linton@arm.com
Link: http://lkml.kernel.org/r/20161209135900.GU3174@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
1 parent ad5013d
xcbc.c
/*
* Copyright (C)2006 USAGI/WIDE Project
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* Author:
* Kazunori Miyazawa <miyazawa@linux-ipv6.org>
*/
#include <crypto/internal/hash.h>
#include <linux/err.h>
#include <linux/kernel.h>
#include <linux/module.h>
static u_int32_t ks[12] = {0x01010101, 0x01010101, 0x01010101, 0x01010101,
0x02020202, 0x02020202, 0x02020202, 0x02020202,
0x03030303, 0x03030303, 0x03030303, 0x03030303};
/*
* +------------------------
* | <parent tfm>
* +------------------------
* | xcbc_tfm_ctx
* +------------------------
* | consts (block size * 2)
* +------------------------
*/
struct xcbc_tfm_ctx {
struct crypto_cipher *child;
u8 ctx[];
};
/*
* +------------------------
* | <shash desc>
* +------------------------
* | xcbc_desc_ctx
* +------------------------
* | odds (block size)
* +------------------------
* | prev (block size)
* +------------------------
*/
struct xcbc_desc_ctx {
unsigned int len;
u8 ctx[];
};
static int crypto_xcbc_digest_setkey(struct crypto_shash *parent,
const u8 *inkey, unsigned int keylen)
{
unsigned long alignmask = crypto_shash_alignmask(parent);
struct xcbc_tfm_ctx *ctx = crypto_shash_ctx(parent);
int bs = crypto_shash_blocksize(parent);
u8 *consts = PTR_ALIGN(&ctx->ctx[0], alignmask + 1);
int err = 0;
u8 key1[bs];
if ((err = crypto_cipher_setkey(ctx->child, inkey, keylen)))
return err;
crypto_cipher_encrypt_one(ctx->child, consts, (u8 *)ks + bs);
crypto_cipher_encrypt_one(ctx->child, consts + bs, (u8 *)ks + bs * 2);
crypto_cipher_encrypt_one(ctx->child, key1, (u8 *)ks);
return crypto_cipher_setkey(ctx->child, key1, bs);
}
static int crypto_xcbc_digest_init(struct shash_desc *pdesc)
{
unsigned long alignmask = crypto_shash_alignmask(pdesc->tfm);
struct xcbc_desc_ctx *ctx = shash_desc_ctx(pdesc);
int bs = crypto_shash_blocksize(pdesc->tfm);
u8 *prev = PTR_ALIGN(&ctx->ctx[0], alignmask + 1) + bs;
ctx->len = 0;
memset(prev, 0, bs);
return 0;
}
static int crypto_xcbc_digest_update(struct shash_desc *pdesc, const u8 *p,
unsigned int len)
{
struct crypto_shash *parent = pdesc->tfm;
unsigned long alignmask = crypto_shash_alignmask(parent);
struct xcbc_tfm_ctx *tctx = crypto_shash_ctx(parent);
struct xcbc_desc_ctx *ctx = shash_desc_ctx(pdesc);
struct crypto_cipher *tfm = tctx->child;
int bs = crypto_shash_blocksize(parent);
u8 *odds = PTR_ALIGN(&ctx->ctx[0], alignmask + 1);
u8 *prev = odds + bs;
/* checking the data can fill the block */
if ((ctx->len + len) <= bs) {
memcpy(odds + ctx->len, p, len);
ctx->len += len;
return 0;
}
/* filling odds with new data and encrypting it */
memcpy(odds + ctx->len, p, bs - ctx->len);
len -= bs - ctx->len;
p += bs - ctx->len;
crypto_xor(prev, odds, bs);
crypto_cipher_encrypt_one(tfm, prev, prev);
/* clearing the length */
ctx->len = 0;
/* encrypting the rest of data */
while (len > bs) {
crypto_xor(prev, p, bs);
crypto_cipher_encrypt_one(tfm, prev, prev);
p += bs;
len -= bs;
}
/* keeping the surplus of blocksize */
if (len) {
memcpy(odds, p, len);
ctx->len = len;
}
return 0;
}
static int crypto_xcbc_digest_final(struct shash_desc *pdesc, u8 *out)
{
struct crypto_shash *parent = pdesc->tfm;
unsigned long alignmask = crypto_shash_alignmask(parent);
struct xcbc_tfm_ctx *tctx = crypto_shash_ctx(parent);
struct xcbc_desc_ctx *ctx = shash_desc_ctx(pdesc);
struct crypto_cipher *tfm = tctx->child;
int bs = crypto_shash_blocksize(parent);
u8 *consts = PTR_ALIGN(&tctx->ctx[0], alignmask + 1);
u8 *odds = PTR_ALIGN(&ctx->ctx[0], alignmask + 1);
u8 *prev = odds + bs;
unsigned int offset = 0;
if (ctx->len != bs) {
unsigned int rlen;
u8 *p = odds + ctx->len;
*p = 0x80;
p++;
rlen = bs - ctx->len -1;
if (rlen)
memset(p, 0, rlen);
offset += bs;
}
crypto_xor(prev, odds, bs);
crypto_xor(prev, consts + offset, bs);
crypto_cipher_encrypt_one(tfm, out, prev);
return 0;
}
static int xcbc_init_tfm(struct crypto_tfm *tfm)
{
struct crypto_cipher *cipher;
struct crypto_instance *inst = (void *)tfm->__crt_alg;
struct crypto_spawn *spawn = crypto_instance_ctx(inst);
struct xcbc_tfm_ctx *ctx = crypto_tfm_ctx(tfm);
cipher = crypto_spawn_cipher(spawn);
if (IS_ERR(cipher))
return PTR_ERR(cipher);
ctx->child = cipher;
return 0;
};
static void xcbc_exit_tfm(struct crypto_tfm *tfm)
{
struct xcbc_tfm_ctx *ctx = crypto_tfm_ctx(tfm);
crypto_free_cipher(ctx->child);
}
static int xcbc_create(struct crypto_template *tmpl, struct rtattr **tb)
{
struct shash_instance *inst;
struct crypto_alg *alg;
unsigned long alignmask;
int err;
err = crypto_check_attr_type(tb, CRYPTO_ALG_TYPE_SHASH);
if (err)
return err;
alg = crypto_get_attr_alg(tb, CRYPTO_ALG_TYPE_CIPHER,
CRYPTO_ALG_TYPE_MASK);
if (IS_ERR(alg))
return PTR_ERR(alg);
switch(alg->cra_blocksize) {
case 16:
break;
default:
goto out_put_alg;
}
inst = shash_alloc_instance("xcbc", alg);
err = PTR_ERR(inst);
if (IS_ERR(inst))
goto out_put_alg;
err = crypto_init_spawn(shash_instance_ctx(inst), alg,
shash_crypto_instance(inst),
CRYPTO_ALG_TYPE_MASK);
if (err)
goto out_free_inst;
alignmask = alg->cra_alignmask | 3;
inst->alg.base.cra_alignmask = alignmask;
inst->alg.base.cra_priority = alg->cra_priority;
inst->alg.base.cra_blocksize = alg->cra_blocksize;
inst->alg.digestsize = alg->cra_blocksize;
inst->alg.descsize = ALIGN(sizeof(struct xcbc_desc_ctx),
crypto_tfm_ctx_alignment()) +
(alignmask &
~(crypto_tfm_ctx_alignment() - 1)) +
alg->cra_blocksize * 2;
inst->alg.base.cra_ctxsize = ALIGN(sizeof(struct xcbc_tfm_ctx),
alignmask + 1) +
alg->cra_blocksize * 2;
inst->alg.base.cra_init = xcbc_init_tfm;
inst->alg.base.cra_exit = xcbc_exit_tfm;
inst->alg.init = crypto_xcbc_digest_init;
inst->alg.update = crypto_xcbc_digest_update;
inst->alg.final = crypto_xcbc_digest_final;
inst->alg.setkey = crypto_xcbc_digest_setkey;
err = shash_register_instance(tmpl, inst);
if (err) {
out_free_inst:
shash_free_instance(shash_crypto_instance(inst));
}
out_put_alg:
crypto_mod_put(alg);
return err;
}
static struct crypto_template crypto_xcbc_tmpl = {
.name = "xcbc",
.create = xcbc_create,
.free = shash_free_instance,
.module = THIS_MODULE,
};
static int __init crypto_xcbc_module_init(void)
{
return crypto_register_template(&crypto_xcbc_tmpl);
}
static void __exit crypto_xcbc_module_exit(void)
{
crypto_unregister_template(&crypto_xcbc_tmpl);
}
module_init(crypto_xcbc_module_init);
module_exit(crypto_xcbc_module_exit);
MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("XCBC keyed hash algorithm");
MODULE_ALIAS_CRYPTO("xcbc");
Computing file changes ...
