This is the mail archive of the
mailing list for the binutils project.
Re: PATCH: PR binutils/18209: objdump -h is not very helpful printing decompressed section names
- From: Doug Evans <dje at google dot com>
- To: "H.J. Lu" <hjl dot tools at gmail dot com>
- Cc: Nicholas Clifton <nickc at redhat dot com>, Binutils <binutils at sourceware dot org>
- Date: Tue, 21 Apr 2015 16:35:16 -0700
- Subject: Re: PATCH: PR binutils/18209: objdump -h is not very helpful printing decompressed section names
- Authentication-results: sourceware.org; auth=none
- References: <20150419141344 dot GA18454 at gmail dot com> <55367835 dot 3070608 at redhat dot com> <CAMe9rOp=UArmwx3_jjK3J-wDvWU5Z1PY3ifP4HnEdcFiVk9qyQ at mail dot gmail dot com>
On Tue, Apr 21, 2015 at 9:25 AM, H.J. Lu <email@example.com> wrote:
> On Tue, Apr 21, 2015 at 9:17 AM, Nicholas Clifton <firstname.lastname@example.org> wrote:
>> Hi H.J.
>>> OK for master?
>> This patch appears to introduce a lot of regressions. For example:
>> Checking Binutils in: mips-elf ... BIN: 4 done
>> BIN REGRESSION: objdump compress debug sections
>> BIN REGRESSION: objdump compress debug sections 3
>> BIN REGRESSION: objdump -W
>> and so on - there are lots more.
> It is because some compressed debug sections aren't smaller
> for big endian targets. Please try this patch.
fwiw I ran into something like this as well.
The cleverness of not actually compressing the section when being asked to
is annoying. For myself, IWBN if there was a
"Yes, I said compress. So please do as I say!" flag.
I'm curious if the current cleverness is demonstrably a win.
I'm guessing it is, otherwise it wouldn't have gone in,
but I'd still be curious to see the data.