public inbox for libabigail@sourceware.org
 help / color / mirror / Atom feed
From: "dodji at redhat dot com" <sourceware-bugzilla@sourceware.org>
To: libabigail@sourceware.org
Subject: [Bug default/26769] abixml writer misses some member types in the output
Date: Fri, 20 Nov 2020 15:29:12 +0000	[thread overview]
Message-ID: <bug-26769-9487-koaKnuUkXl@http.sourceware.org/bugzilla/> (raw)
In-Reply-To: <bug-26769-9487@http.sourceware.org/bugzilla/>

https://sourceware.org/bugzilla/show_bug.cgi?id=26769

--- Comment #10 from dodji at redhat dot com ---
(In reply to Ben Woodard from comment #8)
> Rather than approaching this the way that you are, may I suggest doing what
> I'm doing which I believe is much simpler.
> 
> I run debuginfod and have a local repository of all the debuginfo RPMs. The
> current elfutils provides this and it works extremely well.

Yes, that is true that debuginfod works well.

However, I went ahead and added a new --self-check option to abipkgdiff.

That way you can get the RPM packages of the program you want to use to
validate libabigail on and type:

$ time abipkgdiff --self-check --d1 gcc-debuginfo-10.2.1-8.fc34.x86_64.rpm --d1
libstdc++-debuginfo-10.2.1-8.fc34.x86_64.rpm libstdc++-10.2.1-8.fc34.x86_64.rpm 
==== SELF CHECK SUCCEEDED for 'libstdc++.so.6.0.28' ====

real    0m33,288s
user    0m30,189s
sys     0m5,312s
$ 

What it does is that it compares the binaries inside the RPM against their
ABIXML representation, using the debug info provided by the ancillary debuginfo
packages.

One advantage of doing that is that I could make the "fedabipkgdiff" too to use
that new option as well.  Now we can ask "fedabipkgidff" to do the same on any
fedora package and that tool handles the downloading of all ancillary packages
for us, and it performs the test for *all* architectures (not just the current
one you are running your box on).  All that with just one command.

For instance, here is how I can ask "fedabipkgdiff" to perform that check on
all the sub-packags of the "boost" package from fedora 34:

$ fedabipkgdiff -a --debug --abipkgdiff build/tools/abipkgdiff --from fc34
--self-compare boost

Note that the --abipkgdiff option is to tell the tool to use a non-installed
version of the abipkgdiff program that I have just built from source revision
control.

Hopefully, this has some value for us to improve our coarse grain regression
tests like the one at https://pagure.io/libabigail-selfcheck/tree/master

I have now updated abipkgdiff and fedabipkgdiff in the branch
https://sourceware.org/git/?p=libabigail.git;a=shortlog;h=refs/heads/dodji/PR26769

-- 
You are receiving this mail because:
You are on the CC list for the bug.

  parent reply	other threads:[~2020-11-20 15:29 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-10-22  5:10 [Bug default/26769] New: assert when comparing libstdc++ to itself woodard at redhat dot com
2020-10-22  6:00 ` [Bug default/26769] " woodard at redhat dot com
2020-10-23  4:48 ` woodard at redhat dot com
2020-10-23 22:22 ` woodard at redhat dot com
2020-11-10 10:22 ` dodji at redhat dot com
2020-11-12 22:23 ` woodard at redhat dot com
2020-11-16 17:15 ` dodji at redhat dot com
2020-11-16 20:33 ` woodard at redhat dot com
2020-11-17 17:39   ` Dodji Seketeli
2020-11-19 10:13   ` Dodji Seketeli
2020-11-17 17:39 ` dodji at seketeli dot org
2020-11-17 17:50 ` woodard at redhat dot com
2020-11-19 16:56 ` dodji at redhat dot com
2020-11-19 16:58 ` [Bug default/26769] abixml writer misses some member types in the output dodji at redhat dot com
2020-11-20 15:29 ` dodji at redhat dot com [this message]
2020-11-20 15:34 ` dodji at redhat dot com
2020-11-20 17:49 ` dodji at seketeli dot org
2020-11-20 22:29 ` woodard at redhat dot com
2020-11-23 12:26 ` dodji at redhat dot com
2020-11-23 17:42 ` woodard at redhat dot com
2020-11-24 10:39 ` dodji at redhat dot com

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bug-26769-9487-koaKnuUkXl@http.sourceware.org/bugzilla/ \
    --to=sourceware-bugzilla@sourceware.org \
    --cc=libabigail@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).