qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Richard Henderson <richard.henderson@linaro.org>
To: qemu-devel@nongnu.org
Cc: "Philippe Mathieu-Daudé" <philmd@linaro.org>
Subject: [PULL 17/22] tcg: Do not elide memory barriers for !CF_PARALLEL in system mode
Date: Mon, 26 Jun 2023 17:39:40 +0200	[thread overview]
Message-ID: <20230626153945.76180-18-richard.henderson@linaro.org> (raw)
In-Reply-To: <20230626153945.76180-1-richard.henderson@linaro.org>

The virtio devices require proper memory ordering between
the vcpus and the iothreads.

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
 tcg/tcg-op.c | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)

diff --git a/tcg/tcg-op.c b/tcg/tcg-op.c
index c07de5d9f8..7aadb37756 100644
--- a/tcg/tcg-op.c
+++ b/tcg/tcg-op.c
@@ -102,7 +102,19 @@ void tcg_gen_br(TCGLabel *l)
 
 void tcg_gen_mb(TCGBar mb_type)
 {
-    if (tcg_ctx->gen_tb->cflags & CF_PARALLEL) {
+#ifdef CONFIG_USER_ONLY
+    bool parallel = tcg_ctx->gen_tb->cflags & CF_PARALLEL;
+#else
+    /*
+     * It is tempting to elide the barrier in a uniprocessor context.
+     * However, even with a single cpu we have i/o threads running in
+     * parallel, and lack of memory order can result in e.g. virtio
+     * queue entries being read incorrectly.
+     */
+    bool parallel = true;
+#endif
+
+    if (parallel) {
         tcg_gen_op1(INDEX_op_mb, mb_type);
     }
 }
-- 
2.34.1



  parent reply	other threads:[~2023-06-26 15:44 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-26 15:39 [PULL 00/22] tcg patch queue Richard Henderson
2023-06-26 15:39 ` [PULL 01/22] accel: Replace target_ulong in tlb_*() Richard Henderson
2023-06-26 15:39 ` [PULL 02/22] accel/tcg/translate-all.c: Widen pc and cs_base Richard Henderson
2023-07-11 16:39   ` Peter Maydell
2023-06-26 15:39 ` [PULL 03/22] target: Widen pc/cs_base in cpu_get_tb_cpu_state Richard Henderson
2023-06-26 15:39 ` [PULL 04/22] accel/tcg/cputlb.c: Widen CPUTLBEntry access functions Richard Henderson
2023-06-26 15:39 ` [PULL 05/22] accel/tcg/cputlb.c: Widen addr in MMULookupPageData Richard Henderson
2023-06-26 15:39 ` [PULL 06/22] accel/tcg/cpu-exec.c: Widen pc to vaddr Richard Henderson
2023-07-11 16:40   ` Peter Maydell
2023-06-26 15:39 ` [PULL 07/22] accel/tcg: Widen pc to vaddr in CPUJumpCache Richard Henderson
2023-06-26 15:39 ` [PULL 08/22] accel: Replace target_ulong with vaddr in probe_*() Richard Henderson
2023-06-26 15:39 ` [PULL 09/22] accel/tcg: Replace target_ulong with vaddr in *_mmu_lookup() Richard Henderson
2023-06-26 15:39 ` [PULL 10/22] accel/tcg: Replace target_ulong with vaddr in translator_*() Richard Henderson
2023-06-26 15:39 ` [PULL 11/22] cpu: Replace target_ulong with hwaddr in tb_invalidate_phys_addr() Richard Henderson
2023-06-26 15:39 ` [PULL 12/22] softfloat: use QEMU_FLATTEN to avoid mistaken isra inlining Richard Henderson
2023-06-26 15:39 ` [PULL 13/22] tests/plugin: Remove duplicate insn log from libinsn.so Richard Henderson
2023-06-26 15:39 ` [PULL 14/22] accel/tcg: remove CONFIG_PROFILER Richard Henderson
2023-06-26 15:39 ` [PULL 15/22] tcg: Fix temporary variable in tcg_gen_gvec_andcs Richard Henderson
2023-06-26 15:39 ` [PULL 16/22] target/microblaze: Define TCG_GUEST_DEFAULT_MO Richard Henderson
2023-06-26 15:39 ` Richard Henderson [this message]
2023-06-26 15:39 ` [PULL 18/22] tcg: Add host memory barriers to cpu_ldst.h interfaces Richard Henderson
2023-06-26 15:39 ` [PULL 19/22] accel/tcg: Remove check_tcg_memory_orders_compatible Richard Henderson
2023-06-26 15:39 ` [PULL 20/22] accel/tcg: Store some tlb flags in CPUTLBEntryFull Richard Henderson
2023-06-26 15:39 ` [PULL 21/22] accel/tcg: Move TLB_WATCHPOINT to TLB_SLOW_FLAGS_MASK Richard Henderson
2023-06-26 15:39 ` [PULL 22/22] accel/tcg: Renumber TLB_DISCARD_WRITE Richard Henderson
2023-06-26 18:11 ` [PULL 00/22] tcg patch queue Richard Henderson

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=20230626153945.76180-18-richard.henderson@linaro.org \
    --to=richard.henderson@linaro.org \
    --cc=philmd@linaro.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).