oohal
commented
4 years ago Given that subscribing to you rmailing list does not work I'll send the pull request anyway, despite the warning that I shouldn't.
Weird. I'll ask our sysadmin to have a look.
You can post directly to the list without being subscribed. The message will go into a moderation queue at first, but I always whitelist people sending actual patches.
The check here in the code is wrong. mmap never returns zero, on failure it returns MAP_FAILED (or -1).
There are 2 more cases in the code, however it's less clear to me how to fix them. In libstb/create-container.c there is an mmap call that doesn't seem to check the return value at all.
In extract-gcov.c there's this:
addr = mmap(NULL, sb.st_size, PROT_READ, MAP_PRIVATE, fd, 0); assert(addr != NULL);This is doubly wrong, as asserts shouldn't be used for error checking.
Suppose. extract-gcov and print-container are random helper applications which is probably why no one bothered fixing them. Thanks for looking at this stuff.
Given that subscribing to you rmailing list does not work I'll send the pull request anyway, despite the warning that I shouldn't.
The check here in the code is wrong. mmap never returns zero, on failure it returns MAP_FAILED (or -1).
There are 2 more cases in the code, however it's less clear to me how to fix them. In libstb/create-container.c there is an mmap call that doesn't seem to check the return value at all.
In extract-gcov.c there's this:
This is doubly wrong, as asserts shouldn't be used for error checking.