Ignore double faults as start of a second report.
Double fault can happen during handling of paging faults
if memory is badly corrupted. Also it usually happens
synchronously, which means that maybe the report is not corrupted.
But of course it can come from another CPU as well.
Add more interesting test cases.
Builds in one distro, but another says:
In file included from <stdin>:39:0:
/usr/powerpc64le-linux-gnu/include/linux/if.h:143:8: error: redefinition of ‘struct ifmap’
/usr/powerpc64le-linux-gnu/include/net/if.h:111:8: note: originally defined here
Mess. Try to fix it.
Not sure what's the right solution and it it even exists.
ip command caused several problems:
1. It is installed in different locations or
not installed at all in different distros.
2. It does not support latest kernel devices,
e.g. setup of hsr currently fails because
our ip does not understand its custom prose.
3. ip command is slow, unbearably slow in emulator
(full setup takes tens of seconds). This change
reduces setup from ~2s to ~400ms.
4. ip is not present in gvisor, but it will support netlink.
Use netlink directly to solve all these problems.
FreeBSD sets the kcov buffer as number of bytes instead of number
of entries. This also fixes the mmap() call, which was failing
due to inconsistent sizes. The failing was hidden due to wrong
error handling.
trace2syz used to always add a null byte to strings.
This isn't correct behavior since we may end up writing null bytes to files.
The extra byte can affect system calls like ioctl FS_IOC_ENABLE_VERITY.
We now only add the byte for filenames.
We don't properly parse octal numbers which is an issue because that's how strace decodes file permissions even under -Xraw.
We used to do so but we didn't have a test for it so as when we added support for negative numbers we broke how we parse octals.
This commit fixes how we parse octals and includes a test to catch future regressions.
We don't specify trailing unused args for some syscalls
(e.g. ioctl that does not use its arg).
Executor always filled tailing unsed args with 0's
but pkg/csource didn't. Some such syscalls actually
check that the unsed arg is 0 and as the result failed with C repro.
We could statically check and eliminate all such cases,
but it turns out the warning fires in 1500+ cases:
a3ace5a63f/gistfile1.txt
So instead fill such args with 0's in pkg/csource too.
Strace decodes bitsets like fd_set or sa_mask into the following structure: [a b c d e].
Before we couldn't parse these structures, but now parse them into the Group IR Type
Currently we only support canExit flag.
However there are actually 3 separate conditions:
- program can exit normally
- program can timeout (e.g. fuzzer test or runtest can't)
- program can exit with error (e.g. C test can)
Allow to specify these 3 conditions separately.
