[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH v3 27/28] tcg: When allocating for !splitwx, begin with PROT_NONE
From: |
Richard Henderson |
Subject: |
[PATCH v3 27/28] tcg: When allocating for !splitwx, begin with PROT_NONE |
Date: |
Sun, 2 May 2021 16:18:43 -0700 |
There's a change in mprotect() behaviour [1] in the latest macOS
on M1 and it's not yet clear if it's going to be fixed by Apple.
In this case, instead of changing permissions of N guard pages,
we change permissions of N rwx regions. The same number of
syscalls are required either way.
[1] https://gist.github.com/hikalium/75ae822466ee4da13cbbe486498a191f
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
tcg/region.c | 19 +++++++++----------
1 file changed, 9 insertions(+), 10 deletions(-)
diff --git a/tcg/region.c b/tcg/region.c
index 604530b902..5e00db4cfb 100644
--- a/tcg/region.c
+++ b/tcg/region.c
@@ -765,12 +765,15 @@ static int alloc_code_gen_buffer(size_t size, int
splitwx, Error **errp)
error_free_or_abort(errp);
}
- prot = PROT_READ | PROT_WRITE | PROT_EXEC;
+ /*
+ * macOS 11.2 has a bug (Apple Feedback FB8994773) in which mprotect
+ * rejects a permission change from RWX -> NONE when reserving the
+ * guard pages later. We can go the other way with the same number
+ * of syscalls, so always begin with PROT_NONE.
+ */
+ prot = PROT_NONE;
flags = MAP_PRIVATE | MAP_ANONYMOUS;
-#ifdef CONFIG_TCG_INTERPRETER
- /* The tcg interpreter does not need execute permission. */
- prot = PROT_READ | PROT_WRITE;
-#elif defined(CONFIG_DARWIN)
+#ifdef CONFIG_DARWIN
/* Applicable to both iOS and macOS (Apple Silicon). */
if (!splitwx) {
flags |= MAP_JIT;
@@ -901,11 +904,7 @@ void tcg_region_init(size_t tb_size, int splitwx, unsigned
max_cpus)
}
}
if (have_prot != 0) {
- /*
- * macOS 11.2 has a bug (Apple Feedback FB8994773) in which
mprotect
- * rejects a permission change from RWX -> NONE. Guard pages are
- * nice for bug detection but are not essential; ignore any
failure.
- */
+ /* Guard pages are nice for bug detection but are not essential. */
(void)qemu_mprotect_none(end, page_size);
}
}
--
2.25.1
- [PATCH v3 15/28] tcg: Move MAX_CODE_GEN_BUFFER_SIZE to tcg-target.h, (continued)
- [PATCH v3 15/28] tcg: Move MAX_CODE_GEN_BUFFER_SIZE to tcg-target.h, Richard Henderson, 2021/05/02
- [PATCH v3 11/28] tcg: Create tcg_init, Richard Henderson, 2021/05/02
- [PATCH v3 18/28] tcg: Tidy tcg_n_regions, Richard Henderson, 2021/05/02
- [PATCH v3 16/28] tcg: Replace region.end with region.total_size, Richard Henderson, 2021/05/02
- [PATCH v3 24/28] util/osdep: Add qemu_mprotect_rw, Richard Henderson, 2021/05/02
- [PATCH v3 17/28] tcg: Rename region.start to region.after_prologue, Richard Henderson, 2021/05/02
- [PATCH v3 22/28] tcg: Return the map protection from alloc_code_gen_buffer, Richard Henderson, 2021/05/02
- [PATCH v3 21/28] tcg: Allocate code_gen_buffer into struct tcg_region_state, Richard Henderson, 2021/05/02
- [PATCH v3 20/28] tcg: Move in_code_gen_buffer and tests to region.c, Richard Henderson, 2021/05/02
- [PATCH v3 23/28] tcg: Sink qemu_madvise call to common code, Richard Henderson, 2021/05/02
- [PATCH v3 27/28] tcg: When allocating for !splitwx, begin with PROT_NONE,
Richard Henderson <=
- [PATCH v3 28/28] tcg: Move tcg_init_ctx and tcg_ctx from accel/tcg/, Richard Henderson, 2021/05/02
- [PATCH v3 19/28] tcg: Tidy split_cross_256mb, Richard Henderson, 2021/05/02
- [PATCH v3 26/28] tcg: Merge buffer protection and guard page protection, Richard Henderson, 2021/05/02
- [PATCH v3 09/28] accel/tcg: Move alloc_code_gen_buffer to tcg/region.c, Richard Henderson, 2021/05/02
- [PATCH v3 12/28] accel/tcg: Merge tcg_exec_init into tcg_init_machine, Richard Henderson, 2021/05/02
- [PATCH v3 13/28] accel/tcg: Pass down max_cpus to tcg_init, Richard Henderson, 2021/05/02
- [PATCH v3 14/28] tcg: Introduce tcg_max_ctxs, Richard Henderson, 2021/05/02
- [PATCH v3 25/28] tcg: Round the tb_size default from qemu_get_host_physmem, Richard Henderson, 2021/05/02
- Re: [PATCH v3 00/28] tcg: Clean up code_gen_buffer allocation, no-reply, 2021/05/02