public inbox for libabigail@sourceware.org
 help / color / mirror / Atom feed
From: "woodard at redhat dot com" <sourceware-bugzilla@sourceware.org>
To: libabigail@sourceware.org
Subject: [Bug default/27514] New: recursively compare libraries in abicompat
Date: Thu, 04 Mar 2021 18:02:20 +0000	[thread overview]
Message-ID: <bug-27514-9487@http.sourceware.org/bugzilla/> (raw)

https://sourceware.org/bugzilla/show_bug.cgi?id=27514

            Bug ID: 27514
           Summary: recursively compare libraries in abicompat
           Product: libabigail
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: default
          Assignee: dodji at redhat dot com
          Reporter: woodard at redhat dot com
                CC: libabigail at sourceware dot org
  Target Milestone: ---

This is similar and kind of related to 27208 but goes a bit deeper.

abicompat currently only compares one library at a time when evaluating library
compatibility. For example:

abicompat app libfoo-v1 libfoo-v2 

The app is also likely to have other dependencies which require ABI artifacts
out of libfoo that the app itself may not use. 

app requires libfoo.so and libbar.so
   app only uses foo_init()
   libbar uses additional functions out of libfoo.so

Currently abicompat would say that app is compatible with libfoo so long as
foo_init() didn't change. However, if libfoo-v2 had changes that impacted
libbar.so which is needed by the app then libfoo-v2 is not compatible with the
app.

Therefore the comparison currently done by abicompat where it simply looks at
the undefined symbols in the app is not sufficient to guarantee that libfoo-v1
and libfoo-v2 are in fact compatible with the application. It must also look at
the dependencies needed by app and include any symbols provided by them.

To be able to address this problem while not breaking current behavior a new
mode of operation is needed for abicompat one that loads the corpora not just
for the app but also for all its dependencies and then does the abi comparison.

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

             reply	other threads:[~2021-03-04 18:02 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-04 18:02 woodard at redhat dot com [this message]
2021-03-04 18:02 ` [Bug default/27514] " woodard at redhat dot com
2021-03-04 18:10 ` woodard at redhat dot com
2021-03-23 13:34 ` woodard at redhat dot com
2021-04-13 20:02 ` woodard at redhat dot com
2021-06-23  5:23 ` marksmithh644 at gmail 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-27514-9487@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).