From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: "Emilio G. Cota" <cota@braap.org>
Subject: [Qemu-devel] [PULL 18/31] atomics: emit an smp_read_barrier_depends() barrier only for Alpha and Thread Sanitizer
Date: Fri, 27 May 2016 12:06:31 +0200 [thread overview]
Message-ID: <1464343604-517-19-git-send-email-pbonzini@redhat.com> (raw)
In-Reply-To: <1464343604-517-1-git-send-email-pbonzini@redhat.com>
From: "Emilio G. Cota" <cota@braap.org>
For correctness, smp_read_barrier_depends() is only required to
emit a barrier on Alpha hosts. However, we are currently emitting
a consume fence unconditionally, and most compilers currently treat
consume and acquire fences as equivalent.
Fix it by keeping the consume fence if we're compiling with Thread
Sanitizer, since this might help prevent false warnings. Otherwise,
only emit the barrier for Alpha hosts. Note that we still guarantee
that smp_read_barrier_depends() is a compiler barrier.
Signed-off-by: Emilio G. Cota <cota@braap.org>
Message-Id: <1464120374-8950-3-git-send-email-cota@braap.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
include/qemu/atomic.h | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/include/qemu/atomic.h b/include/qemu/atomic.h
index 5bc4d6c..96db6e9 100644
--- a/include/qemu/atomic.h
+++ b/include/qemu/atomic.h
@@ -36,7 +36,18 @@
#define smp_wmb() ({ barrier(); __atomic_thread_fence(__ATOMIC_RELEASE); barrier(); })
#define smp_rmb() ({ barrier(); __atomic_thread_fence(__ATOMIC_ACQUIRE); barrier(); })
+/* Most compilers currently treat consume and acquire the same, but really
+ * no processors except Alpha need a barrier here. Leave it in if
+ * using Thread Sanitizer to avoid warnings, otherwise optimize it away.
+ */
+#if defined(__SANITIZE_THREAD__)
#define smp_read_barrier_depends() ({ barrier(); __atomic_thread_fence(__ATOMIC_CONSUME); barrier(); })
+#elsif defined(__alpha__)
+#define smp_read_barrier_depends() asm volatile("mb":::"memory")
+#else
+#define smp_read_barrier_depends() barrier()
+#endif
+
/* Weak atomic operations prevent the compiler moving other
* loads/stores past the atomic operation load/store. However there is
--
2.5.5
next prev parent reply other threads:[~2016-05-27 10:07 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-05-27 10:06 [Qemu-devel] [PULL 00/31] Misc changes for 2016-05-27 Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 01/31] Add optionrom compatible with fw_cfg DMA version Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 02/31] Revert "memory: Drop FlatRange.romd_mode" Paolo Bonzini
2016-05-27 10:51 ` Laszlo Ersek
2016-05-27 10:06 ` [Qemu-devel] [PULL 03/31] hw/char: QOM'ify escc.c Paolo Bonzini
2016-05-31 22:13 ` Mark Cave-Ayland
2016-06-01 3:06 ` xiaoqiang zhao
2016-06-01 7:04 ` Mark Cave-Ayland
2016-06-01 7:08 ` xiaoqiang zhao
2016-05-27 10:06 ` [Qemu-devel] [PULL 04/31] hw/char: QOM'ify etraxfs_ser.c Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 05/31] hw/char: QOM'ify lm32_juart.c Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 06/31] hw/char: QOM'ify lm32_uart.c Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 07/31] hw/char: QOM'ify milkymist-uart.c Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 08/31] nbd: Don't trim unrequested bytes Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 09/31] kvm_stat: Remove Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 10/31] scsi: pvscsi: check command descriptor ring buffer size (CVE-2016-4952) Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 11/31] scsi: mptsas: infinite loop while fetching requests Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 12/31] scsi: megasas: use appropriate property buffer size Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 13/31] scsi: megasas: initialise local configuration data buffer Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 14/31] scsi: megasas: check 'read_queue_head' index value Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 15/31] block/iscsi: avoid potential overflow of acb->task->cdb Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 16/31] bt: rewrite csrhci_write to avoid out-of-bounds writes Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 17/31] docs/atomics: update atomic_read/set comparison with Linux Paolo Bonzini
2016-05-27 10:06 ` Paolo Bonzini [this message]
2016-05-27 10:06 ` [Qemu-devel] [PULL 19/31] atomics: do not emit consume barrier for atomic_rcu_read Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 20/31] docs/atomics: update comparison with Linux Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 21/31] xen-hvm: ignore background I/O sections Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 22/31] scsi-disk: introduce a common base class Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 23/31] scsi-disk: introduce dma_readv and dma_writev Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 24/31] scsi-disk: add need_fua_emulation to SCSIDiskClass Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 25/31] scsi-disk: introduce scsi_disk_req_check_error Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 26/31] scsi-block: always use SG_IO Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 27/31] scsi-generic: Merge block max xfer len in INQUIRY response Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 28/31] memory: remove qemu_get_ram_fd, qemu_set_ram_fd, qemu_ram_block_host_ptr Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 29/31] exec: remove ram_addr argument from qemu_ram_block_from_host Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 30/31] memory: split memory_region_from_host from qemu_ram_addr_from_host Paolo Bonzini
2016-05-27 10:06 ` [Qemu-devel] [PULL 31/31] exec: hide mr->ram_addr from qemu_get_ram_ptr users Paolo Bonzini
2016-05-27 12:53 ` [Qemu-devel] [PULL 00/31] Misc changes for 2016-05-27 Peter Maydell
2016-05-27 13:04 ` Peter Maydell
2016-05-27 13:38 ` Richard W.M. Jones
2016-05-27 14:04 ` Peter Maydell
2016-05-27 14:07 ` Paolo Bonzini
2016-05-27 14:06 ` Paolo Bonzini
2016-05-27 14:11 ` Richard W.M. Jones
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1464343604-517-19-git-send-email-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=cota@braap.org \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).