microdump processor unable to process microdump due to whitespaces in input
Opened this issue · 3 comments
GoogleCodeExporter commented
What steps will reproduce the problem?
The microdump processor parsing has a bug where it does not trim trailing
spaces in the cpu context and exits with the following error.
"Malformed CPU context. Got 369 bytes instead of 368.
After applying the below patch in microdump.cc, it is able to process the dump.
@@ -256,7 +258,12 @@
stack_content.insert(stack_content.end(), chunk.begin(), chunk.end());
} else if ((pos = line.find(kCpuKey)) != string::npos) {
- string cpu_state_str(line, pos + strlen(kCpuKey));
+ string cpu_state_str_raw(line, pos + strlen(kCpuKey));
+std::istringstream cpu_state_tokens(cpu_state_str_raw);
+string cpu_state_str;
+cpu_state_tokens >> cpu_state_str;
Can somebody take this patch to the main code.
Original issue reported on code.google.com by atishayj...@gmail.com
on 9 Jun 2015 at 6:56
GoogleCodeExporter commented
We don't take patches here. If you would like to contribute a patch, upload it
to codereview.chromium.org
Original comment by thestig@chromium.org
on 9 Jun 2015 at 6:14
GoogleCodeExporter commented
This is not a patch submission. This is just to root cause the problem and
explain it clearly.
Original comment by atishayj...@gmail.com
on 19 Jun 2015 at 8:08
GoogleCodeExporter commented
There is no need for a patch, this is being addressed in
https://codereview.chromium.org/1181113004/
Original comment by primi...@chromium.org
on 19 Jun 2015 at 8:42