This is the mail archive of the archer@sourceware.org mailing list for the Archer project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Problem with merge


Thank you very much Jan, I applied this patch and it works perfectly.

I appreciate you taking the time to take a look at this and fix it.

Rick


Jan Kratochvil wrote:
Hi Rick,

the problem IMO comes from archer-sergio-catch-syscall and some of its
interaction with the other patches.

The problem can be reproduced just on the archer-sergio-catch-syscall branch
by:
$ ../gdb -nx -ex start -ex 'set debug infrun 1' -ex 'p marker2(99)' -ex bt gdb.base/break
which ends on the FSF GDB by:
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x400530
infrun: BPSTAT_WHAT_STOP_SILENT
infrun: stop_stepping
$1 = 1
#0  main (argc=1, argv=0x7fffffffd438, envp=0x7fffffffd448) at ../.././gdb/testsuite/gdb.base/break.c:86
(gdb) q

but on the archer-sergio-catch-syscall branch it ends up with:
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x400530
infrun: BPSTAT_WHAT_CHECK_SHLIBS
infrun: stop_stepping
$1 = 1
#0  main (argc=1, argv=0x7fffffffd438, envp=0x7fffffffd448) at ../.././gdb/testsuite/gdb.base/break.c:86
(gdb) q

This causes the crash-loop on the archer-rmoseley-demo-merge branch due to
some other correlations I did not analyse.

The attached patch IMO makes the `entrybp' state less intrusive (=not
corrupting the current state if not required) while still
`gdb.base/catch-syscall.exp' fully PASSes there.  It also fixed the
`gdb.base/break.exp' testcase for archer-rmoseley-demo-merge.

I did not compare the full testsuite runs on the both branches.


Regards,
Jan


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]