| Summary: | coregrind/m_debuginfo/readelf.c:1977: variable set but not used | ||
|---|---|---|---|
| Product: | [Developer tools] valgrind | Reporter: | dcb314 |
| Component: | general | Assignee: | Julian Seward <jseward> |
| Status: | RESOLVED FIXED | ||
| Severity: | minor | CC: | pjfloyd |
| Priority: | NOR | ||
| Version First Reported In: | unspecified | ||
| Target Milestone: | --- | ||
| Platform: | Compiled Sources | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed/Implemented In: | ||
| Sentry Crash Report: | |||
I saw this on buildbot and fixed it soon after commit 671b1082e33659926d4fff42767cdffb48ec103c Author: Paul Floyd <pjfloyd@wanadoo.fr> Date: Thu Dec 28 12:08:40 2023 +0100 Fix unused variable warning On macOS I see that mach-o binaries can be produced without a RW segment. The code assumed that there was always one (or more). The trigger for reading debuginfo and redirs is that the "right" number of RX and RW segments have been seen. Since the RW condition was wrong the trigger condition was not met. Now that the RW segment count is correct I had to remove all that asserts that there are at least 1 RW segment. In this case I put back the assert but modified it with the extra condition that di->fsm.rw_map_count can be zero. |
A build of today's git head produces this: m_debuginfo/readelf.c:1977:12: warning: variable ‘has_nonempty_rw’ set but not used [-Wunused-but-set-variable] A quick check on the variable: $ grep has_nonempty_rw coregrind/m_debuginfo/readelf.c Bool has_nonempty_rw = False; has_nonempty_rw = True; $ Maybe some development work in progress or a candidate for deletion.