Reviewers: Michael Starzinger,
Message:
PTAL.
Description:
OS::MemMove/OS::MemCopy: Don't call through to generated code when size ==
0 to
avoid prefetching invalid memory
BUG=chromium:233500
Please review this at https://codereview.chromium.org/14365011/
SVN Base: https://v8.googlecode.com/svn/branches/bleeding_edge
Affected files:
M src/ia32/codegen-ia32.cc
M src/platform-posix.cc
M src/platform-win32.cc
Index: src/ia32/codegen-ia32.cc
diff --git a/src/ia32/codegen-ia32.cc b/src/ia32/codegen-ia32.cc
index
caf808b953731c2a3adbfd56fe3cbf4ca9386526..7663c6a7fd37bd58b466a68db8b4eb86cc436cbc
100644
--- a/src/ia32/codegen-ia32.cc
+++ b/src/ia32/codegen-ia32.cc
@@ -635,6 +635,8 @@ OS::MemMoveFunction CreateMemMoveFunction() {
ASSERT(!RelocInfo::RequiresRelocation(desc));
CPU::FlushICache(buffer, actual_size);
OS::ProtectCode(buffer, actual_size);
+ // TODO(jkummerow): It would be nice to register this code creation event
+ // with the PROFILE / GDBJIT system.
return FUNCTION_CAST<OS::MemMoveFunction>(buffer);
}
Index: src/platform-posix.cc
diff --git a/src/platform-posix.cc b/src/platform-posix.cc
index
48898ed9aad0acdb4cc06ccba4b01c9818f26de4..9f650d1e5b5c3b7bcad6ee53fc2b8921cad7580a
100644
--- a/src/platform-posix.cc
+++ b/src/platform-posix.cc
@@ -334,6 +334,7 @@ OS::MemMoveFunction CreateMemMoveFunction();
// Copy memory area. No restrictions.
void OS::MemMove(void* dest, const void* src, size_t size) {
+ if (size == 0) return;
// Note: here we rely on dependent reads being ordered. This is true
// on all architectures we currently support.
(*memmove_function)(dest, src, size);
Index: src/platform-win32.cc
diff --git a/src/platform-win32.cc b/src/platform-win32.cc
index
272678fe649294017e9758000823da0706ab376c..94b78e45bed144d0a17b1aa904d48acf95e35f77
100644
--- a/src/platform-win32.cc
+++ b/src/platform-win32.cc
@@ -160,6 +160,7 @@ OS::MemMoveFunction CreateMemMoveFunction();
// Copy memory area to disjoint memory area.
void OS::MemMove(void* dest, const void* src, size_t size) {
+ if (size == 0) return;
// Note: here we rely on dependent reads being ordered. This is true
// on all architectures we currently support.
(*memmove_function)(dest, src, size);
--
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
