Commit Graph

7 Commits

Author SHA1 Message Date
Emilio G. Cota 76b553b308 qht: fix unlock-after-free segfault upon resizing
The old map's bucket locks are being unlocked *after*
that same old map has been passed to RCU for destruction.
This is a bug that can cause a segfault, since there's
no guarantee that the deletion will be deferred (e.g.
there may be no concurrent readers).

The segfault is easily triggered in RHEL6/CentOS6 with qht-test,
particularly on a single-core system or by pinning qht-test
to a single core.

Fix it by unlocking the map's bucket locks right after having
published the new map, and (crucially) before marking the map
for deletion via call_rcu().

While at it, expand qht_do_resize() to atomically do (1) a reset,
(2) a resize, or (3) a reset+resize. This simplifies the calling
code, since the new function (qht_do_resize_reset()) acquires
and releases the buckets' locks.

Note that no qht_do_reset inline is provided, since it would have
no users--qht_reset() already performs a reset without taking
ht->lock.

Reported-by: Peter Maydell <peter.maydell@linaro.org>
Reported-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Message-Id: <1475706880-10667-3-git-send-email-cota@braap.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-10-06 18:04:13 +02:00
Emilio G. Cota f555a9d0b3 qht: simplify qht_reset_size
Sometimes gcc doesn't pick up the fact that 'new' is properly
set if 'resize == true', which may generate an unnecessary
build warning.

Fix it by removing 'resize' and directly checking that 'new'
is non-NULL.

Signed-off-by: Emilio G. Cota <cota@braap.org>
Message-Id: <1475706880-10667-2-git-send-email-cota@braap.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-10-06 18:04:13 +02:00
Alex Bennée a890643958 util/qht: atomically set b->hashes
ThreadSanitizer detects a possible race between reading/writing the
hashes. The ordering semantics are already documented for QHT however
for true C11 compliance we should use relaxed atomic primitives for
accesses that are done across threads. On x86 this slightly changes to
the code to not do a load/compare in a single instruction leading to a
slight performance degradation.

Running 'taskset -c 0 tests/qht-bench -n 1 -d 10' (i.e. all lookups) 10
times, we get:

before the patch:
 $ ./mean.pl 34.04 34.24 34.38 34.25 34.18 34.51 34.46 34.44 34.29 34.08
 34.287 +- 0.160072900059109
after:
 $ ./mean.pl 33.94 34.00 33.52 33.46 33.55 33.71 34.27 34.06 34.28 34.58
 33.937 +- 0.374731014640279

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Emilio G. Cota <cota@braap.org>
Message-Id: <20160930213106.20186-10-alex.bennee@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-10-04 10:00:26 +02:00
Emilio G. Cota 7266ae91a1 qht: do not segfault when gathering stats from an uninitialized qht
So far, QHT functions assume that the passed qht has previously been
initialized--otherwise they segfault.

This patch makes an exception for qht_statistics_init, with the goal
of simplifying calling code. For instance, qht_statistics_init is
called from the 'info jit' dump, and given that under KVM the TB qht
is never initialized, we get a segfault. Thus, instead of complicating
the 'info jit' code with additional checks, let's allow passing an
uninitialized qht to qht_statistics_init.

While at it, add a test for this to test-qht.

Before the patch (for $ qemu -enable-kvm [...]):
(qemu) info jit
[...]
direct jump count   0 (0%) (2 jumps=0 0%)
Program received signal SIGSEGV, Segmentation fault.

After the patch the "TB hash buckets", "TB hash occupancy"
and "TB hash avg chain" lines are omitted.
(qemu) info jit
[...]
direct jump count   0 (0%) (2 jumps=0 0%)
TB hash buckets     0/0 (-nan% head buckets used)
TB hash occupancy   nan% avg chain occ. Histogram: (null)
TB hash avg chain   nan buckets. Histogram: (null)
[...]

Reported by: Changlong Xie <xiecl.fnst@cn.fujitsu.com>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Message-Id: <1469205390-14369-1-git-send-email-cota@braap.org>
[Extract printing statistics to an entirely separate function. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-08-02 12:03:58 +02:00
Paolo Bonzini 34506b30e4 util/qht: Document memory ordering assumptions
It is naturally expected that some memory ordering should be provided
around qht_insert() and qht_lookup(). Document these assumptions in the
header file and put some comments in the source to denote how that
memory ordering requirements are fulfilled.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
[Sergey Fedorov: commit title and message provided;
comment on qht_remove() elided]
Signed-off-by: Sergey Fedorov <serge.fdrv@gmail.com>
Message-Id: <20160715175852.30749-2-sergey.fedorov@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-08-02 12:03:58 +02:00
Paolo Bonzini e9abfcb57f clean-includes: run it once more
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-06-16 18:39:03 +02:00
Emilio G. Cota 2e11264aaf qht: QEMU's fast, resizable and scalable Hash Table
This is a fast, scalable chained hash table with optional auto-resizing, allowing
reads that are concurrent with reads, and reads/writes that are concurrent
with writes to separate buckets.

A hash table with these features will be necessary for the scalability
of the ongoing MTTCG work; before those changes arrive we can already
benefit from the single-threaded speedup that qht also provides.

Signed-off-by: Emilio G. Cota <cota@braap.org>
Message-Id: <1465412133-3029-11-git-send-email-cota@braap.org>
Signed-off-by: Richard Henderson <rth@twiddle.net>
2016-06-11 23:10:20 +00:00