[PATCH] [PR corefiles/32441] Fix segfault if target_fileio_read_alloc fails

[Andrew Burgess]

Brandon Belew <brandon.belew@zetier.com> writes:

> Andrew Burgess <aburgess@redhat.com> writes:
>> I think you should reconsider here.  The function returns LONGEST, so
>> that's what should be used.  GDB's general policy is to fix little
>> bugs like this as the code gets touched for other reasons.
>
> Thanks Andrew! I'll go back and look at the LONGEST return value and get
> back to you on that with an updated patch. To be clear, you want the
> other function that was already using ssize_t for the return fixed as
> well? 
>
>> Otherwise, I agree with Luis, this looks great.  If you repost with the
>> description in the commit message we can get this merged.
>
> I replied to Luis' message with some further questions about
> formatting. For reference, my original reply has a Message-ID of
> hwjgya7c6ulo46.fsf@brandonb.zetier.com. I'll redo those questions here:
>
> It was my understanding of git-format-patch / git-send-email (and in
> reverse, git-am) that the commit message would be taken from everything
> up to the '---'. Then everything between the '---' and the actual diff
> was considered "timely commentary" and would not be present in the
> commit message. This typically contains the diffstat output but also is
> used for commentary on the patch that shouldn't go into the commit
> itself.
>
> As described in https://git-scm.com/docs/git-format-patch:
>> Typically it will be placed in a MUA’s drafts folder, edited to add
>> timely commentary that should not go in the changelog ***after the
>> three dashes*** [emphasis added] 
>
> And also as described at
> https://spacekookie.de/blog/collaborating-with-git-send-email/:
>> Another often overlooked feature here is "timely commentary", are
>> comments in the patch e-mail that won't be part of the patch or the
>> commit message itself. They can be made after the --- marker in a
>> patch mail, but before the actual patch starts. This section is
>> usually used for the diff-stat of that particular patch.
>
> I tried sending the patch email to myself before I sent it to the list,
> piping it to `git-am`, and it correctly applied the commit without my
> "timely commentary", so I was pretty convinced I had formatted it
> correctly.
>
> Do you get something different when you `git am` my message? If so, can
> you be more specific about how I need to reformat the email?

I guess if that works, then it's fine.  It's not a layout I've seen
before, and it doesn't make much sense to me.  That is, placing the
email reply between the commit message and the patch seems an odd
ordering to me, I'd usually reply to the various points, and then just
place the patch, including commit message, and the tail of the email.

But thanks for referencing the actual docs.

Thanks,
Andrew