All of lore.kernel.org
 help / color / mirror / Atom feed
From: Emilio Cota <cota@braap.org>
To: qemu-devel@nongnu.org
Cc: "Alex Bennée" <alex.bennee@linaro.org>,
	"Richard Henderson" <richard.henderson@linaro.org>,
	"Aaron Lindsay" <aaron@os.amperecomputing.com>,
	"Emilio Cota" <cota@braap.org>
Subject: [PATCH 1/4] plugins: fix optimization in plugin_gen_disable_mem_helpers
Date: Sun,  8 Jan 2023 11:47:28 -0500	[thread overview]
Message-ID: <20230108164731.61469-2-cota@braap.org> (raw)
In-Reply-To: <20230108164731.61469-1-cota@braap.org>

We were mistakenly checking tcg_ctx->plugin_insn as a canary to know
whether the TB had emitted helpers that might have accessed memory.

The problem is that tcg_ctx->plugin_insn gets updated on every
instruction in the TB, which results in us wrongly performing the
optimization (i.e. not clearing cpu->plugin_mem_cbs) way too often,
since it's not rare that the last instruction in the TB doesn't
use helpers.

Fix it by tracking a per-TB canary.

While at it, expand documentation.

Related: #1381

Signed-off-by: Emilio Cota <cota@braap.org>
---
 accel/tcg/plugin-gen.c | 26 ++++++++++++++++++--------
 include/qemu/plugin.h  |  7 +++++++
 2 files changed, 25 insertions(+), 8 deletions(-)

diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c
index c7d6514840..17a686bd9e 100644
--- a/accel/tcg/plugin-gen.c
+++ b/accel/tcg/plugin-gen.c
@@ -579,7 +579,8 @@ static void inject_mem_helper(TCGOp *begin_op, GArray *arr)
  * is possible that the code we generate after the instruction is
  * dead, we also add checks before generating tb_exit etc.
  */
-static void inject_mem_enable_helper(struct qemu_plugin_insn *plugin_insn,
+static void inject_mem_enable_helper(struct qemu_plugin_tb *ptb,
+                                     struct qemu_plugin_insn *plugin_insn,
                                      TCGOp *begin_op)
 {
     GArray *cbs[2];
@@ -599,6 +600,7 @@ static void inject_mem_enable_helper(struct qemu_plugin_insn *plugin_insn,
         rm_ops(begin_op);
         return;
     }
+    ptb->mem_helper = true;
 
     arr = g_array_sized_new(false, false,
                             sizeof(struct qemu_plugin_dyn_cb), n_cbs);
@@ -626,15 +628,22 @@ void plugin_gen_disable_mem_helpers(void)
 {
     TCGv_ptr ptr;
 
-    if (likely(tcg_ctx->plugin_insn == NULL ||
-               !tcg_ctx->plugin_insn->mem_helper)) {
+    /*
+     * We could emit the clearing unconditionally and be done. However, this can
+     * be wasteful if for instance plugins don't track memory accesses, or if
+     * most TBs don't use helpers. Instead, emit the clearing iff the TB calls
+     * helpers that might access guest memory.
+     *
+     * Note: we do not reset plugin_tb->mem_helper here; a TB might have several
+     * exit points, and we want to emit the clearing from all of them.
+     */
+    if (!tcg_ctx->plugin_tb->mem_helper) {
         return;
     }
     ptr = tcg_const_ptr(NULL);
     tcg_gen_st_ptr(ptr, cpu_env, offsetof(CPUState, plugin_mem_cbs) -
                                  offsetof(ArchCPU, env));
     tcg_temp_free_ptr(ptr);
-    tcg_ctx->plugin_insn->mem_helper = false;
 }
 
 static void plugin_gen_tb_udata(const struct qemu_plugin_tb *ptb,
@@ -682,14 +691,14 @@ static void plugin_gen_mem_inline(const struct qemu_plugin_tb *ptb,
     inject_inline_cb(cbs, begin_op, op_rw);
 }
 
-static void plugin_gen_enable_mem_helper(const struct qemu_plugin_tb *ptb,
+static void plugin_gen_enable_mem_helper(struct qemu_plugin_tb *ptb,
                                          TCGOp *begin_op, int insn_idx)
 {
     struct qemu_plugin_insn *insn = g_ptr_array_index(ptb->insns, insn_idx);
-    inject_mem_enable_helper(insn, begin_op);
+    inject_mem_enable_helper(ptb, insn, begin_op);
 }
 
-static void plugin_gen_disable_mem_helper(const struct qemu_plugin_tb *ptb,
+static void plugin_gen_disable_mem_helper(struct qemu_plugin_tb *ptb,
                                           TCGOp *begin_op, int insn_idx)
 {
     struct qemu_plugin_insn *insn = g_ptr_array_index(ptb->insns, insn_idx);
@@ -750,7 +759,7 @@ static void pr_ops(void)
 #endif
 }
 
-static void plugin_gen_inject(const struct qemu_plugin_tb *plugin_tb)
+static void plugin_gen_inject(struct qemu_plugin_tb *plugin_tb)
 {
     TCGOp *op;
     int insn_idx = -1;
@@ -870,6 +879,7 @@ bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db,
         ptb->haddr1 = db->host_addr[0];
         ptb->haddr2 = NULL;
         ptb->mem_only = mem_only;
+        ptb->mem_helper = false;
 
         plugin_gen_empty_callback(PLUGIN_GEN_FROM_TB);
     }
diff --git a/include/qemu/plugin.h b/include/qemu/plugin.h
index a772e14193..e0ebedef84 100644
--- a/include/qemu/plugin.h
+++ b/include/qemu/plugin.h
@@ -118,7 +118,10 @@ struct qemu_plugin_insn {
     void *haddr;
     GArray *cbs[PLUGIN_N_CB_TYPES][PLUGIN_N_CB_SUBTYPES];
     bool calls_helpers;
+
+    /* if set, the instruction calls helpers that might access guest memory */
     bool mem_helper;
+
     bool mem_only;
 };
 
@@ -158,6 +161,10 @@ struct qemu_plugin_tb {
     void *haddr1;
     void *haddr2;
     bool mem_only;
+
+    /* if set, the TB calls helpers that might access guest memory */
+    bool mem_helper;
+
     GArray *cbs[PLUGIN_N_CB_SUBTYPES];
 };
 
-- 
2.34.1



  reply	other threads:[~2023-01-08 16:48 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-08 16:47 [PATCH 0/4] plugin patches to fix #1381 Emilio Cota
2023-01-08 16:47 ` Emilio Cota [this message]
2023-01-10 15:29   ` [PATCH 1/4] plugins: fix optimization in plugin_gen_disable_mem_helpers Aaron Lindsay
2023-01-08 16:47 ` [PATCH 2/4] translator: always pair plugin_gen_insn_{start, end} calls Emilio Cota
2023-01-08 19:56   ` Philippe Mathieu-Daudé
2023-01-10 15:26   ` [PATCH 2/4] translator: always pair plugin_gen_insn_{start,end} calls Aaron Lindsay
2023-01-08 16:47 ` [PATCH 3/4] tcg: exclude lookup_tb_ptr from helper instrumentation Emilio Cota
2023-01-08 16:51 ` [PATCH 4/4] cpu-exec: assert that plugin_mem_cbs is NULL after execution Emilio Cota
2023-01-09 13:52   ` Alex Bennée
2023-01-10  2:16     ` Emilio Cota
2023-01-10 13:02       ` Alex Bennée

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=20230108164731.61469-2-cota@braap.org \
    --to=cota@braap.org \
    --cc=aaron@os.amperecomputing.com \
    --cc=alex.bennee@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.