[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH v2 25/25] plugins: move reset of plugin data to tb_start
From: |
Alex Bennée |
Subject: |
[PATCH v2 25/25] plugins: move reset of plugin data to tb_start |
Date: |
Tue, 1 Feb 2022 18:20:50 +0000 |
We can't always guarantee we get to the end of a translator loop.
Although this can happen for a variety of reasons it does happen more
often on x86 system emulation when an instruction spans across to an
un-faulted page. This caused confusion of the instruction tracking
data resulting in apparent reverse execution (at least from the
plugins point of view).
Fix this by moving the reset code to plugin_gen_tb_start so we always
start with a clean slate.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/824
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220124201608.604599-23-alex.bennee@linaro.org>
---
accel/tcg/plugin-gen.c | 29 ++++++++++++++++++-----------
1 file changed, 18 insertions(+), 11 deletions(-)
diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c
index 22d95fe1c3..ae70e4a5b0 100644
--- a/accel/tcg/plugin-gen.c
+++ b/accel/tcg/plugin-gen.c
@@ -854,10 +854,21 @@ static void plugin_gen_inject(const struct qemu_plugin_tb
*plugin_tb)
bool plugin_gen_tb_start(CPUState *cpu, const TranslationBlock *tb, bool
mem_only)
{
- struct qemu_plugin_tb *ptb = tcg_ctx->plugin_tb;
bool ret = false;
if (test_bit(QEMU_PLUGIN_EV_VCPU_TB_TRANS, cpu->plugin_mask)) {
+ struct qemu_plugin_tb *ptb = tcg_ctx->plugin_tb;
+ int i;
+
+ /* reset callbacks */
+ for (i = 0; i < PLUGIN_N_CB_SUBTYPES; i++) {
+ if (ptb->cbs[i]) {
+ g_array_set_size(ptb->cbs[i], 0);
+ }
+ }
+ ptb->n = 0;
+ tcg_ctx->plugin_insn = NULL;
+
ret = true;
ptb->vaddr = tb->pc;
@@ -904,23 +915,19 @@ void plugin_gen_insn_end(void)
plugin_gen_empty_callback(PLUGIN_GEN_AFTER_INSN);
}
+/*
+ * There are cases where we never get to finalise a translation - for
+ * example a page fault during translation. As a result we shouldn't
+ * do any clean-up here and make sure things are reset in
+ * plugin_gen_tb_start.
+ */
void plugin_gen_tb_end(CPUState *cpu)
{
struct qemu_plugin_tb *ptb = tcg_ctx->plugin_tb;
- int i;
/* collect instrumentation requests */
qemu_plugin_tb_trans_cb(cpu, ptb);
/* inject the instrumentation at the appropriate places */
plugin_gen_inject(ptb);
-
- /* clean up */
- for (i = 0; i < PLUGIN_N_CB_SUBTYPES; i++) {
- if (ptb->cbs[i]) {
- g_array_set_size(ptb->cbs[i], 0);
- }
- }
- ptb->n = 0;
- tcg_ctx->plugin_insn = NULL;
}
--
2.30.2
- [PATCH v2 00/25] testing and plugin updates, Alex Bennée, 2022/02/01
- [PATCH v2 04/25] .gitignore: add .gcov pattern, Alex Bennée, 2022/02/01
- [PATCH v2 01/25] tests/Makefile.include: clean-up old code, Alex Bennée, 2022/02/01
- [PATCH v2 06/25] gitmodules: Correct libvirt-ci submodule URL, Alex Bennée, 2022/02/01
- [PATCH v2 03/25] Makefile: also remove .gcno files when cleaning, Alex Bennée, 2022/02/01
- [PATCH v2 02/25] tests/qtest: enable more vhost-user tests by default, Alex Bennée, 2022/02/01
- [PATCH v2 11/25] tests/lcitool: Install libibumad to cover RDMA on Debian based distros, Alex Bennée, 2022/02/01
- [PATCH v2 25/25] plugins: move reset of plugin data to tb_start,
Alex Bennée <=
- [PATCH v2 08/25] drop libxml2 checks since libxml is not actually used (for parallels), Alex Bennée, 2022/02/01
- [PATCH v2 12/25] docs/devel: mention our .editorconfig, Alex Bennée, 2022/02/01
- [PATCH v2 09/25] tests/lcitool: Refresh submodule and remove libxml2, Alex Bennée, 2022/02/01
- [PATCH v2 16/25] docs: remove references to TCG tracing, Alex Bennée, 2022/02/01
- [PATCH v2 13/25] gitlab: fall back to commit hash in qemu-setup filename, Alex Bennée, 2022/02/01
- [PATCH v2 07/25] tests/lcitool: Include local qemu.yml when refreshing cirrus-ci files, Alex Bennée, 2022/02/01
- [PATCH v2 15/25] block: fix FreeBSD build failure with fallocate, Alex Bennée, 2022/02/01
- [PATCH v2 05/25] MAINTAINERS: Cover lcitool submodule with build test / automation, Alex Bennée, 2022/02/01