diff options
author | Lancelot SIX <lancelot.six@amd.com> | 2022-08-11 15:09:55 +0100 |
---|---|---|
committer | Lancelot SIX <lancelot.six@amd.com> | 2022-08-11 15:09:55 +0100 |
commit | 739be95178196df4babdcb47de856a12ba06253f (patch) | |
tree | 6599960bd0fee5a00c7debca1cca6f9c2cc6480b /gdb/testsuite/gdb.mi/mi-var-invalidate-shlib.exp | |
parent | 537710a69cf440250c27a59b7d01b5a5a1a9b462 (diff) | |
download | fsf-binutils-gdb-739be95178196df4babdcb47de856a12ba06253f.zip fsf-binutils-gdb-739be95178196df4babdcb47de856a12ba06253f.tar.gz fsf-binutils-gdb-739be95178196df4babdcb47de856a12ba06253f.tar.bz2 |
gdb/varobj: Do not invalidate locals in varobj_invalidate_iter
The varobj_invalidate_iter function has logic to invalidate any local
varobj it can find. However since bc20e562ec0 "gdb/varobj: Fix use
after free in varobj" all varobj containing references to an objfile are
cleared when the objfile goes out of scope. This means that at this
point any local varobj seen by varobj_invalidate_iter either has
already been invalidated by varobj_invalidate_if_uses_objfile or only
contains valid references and there is no reason to invalidate it.
This patch proposes to remove this unnecessary invalidation and adds a
testcase which exercises a scenario where a local varobj can legitimately
survive a call to varobj_invalidate_iter.
At this point the varobj_invalidate and varobj_invalidate_iter seem
misnamed since they deal with re-creating invalid objects and do not do
invalidation, but this will be fixed in a following patch.
Tested on x86_64-linux.
Diffstat (limited to 'gdb/testsuite/gdb.mi/mi-var-invalidate-shlib.exp')
-rw-r--r-- | gdb/testsuite/gdb.mi/mi-var-invalidate-shlib.exp | 40 |
1 files changed, 36 insertions, 4 deletions
diff --git a/gdb/testsuite/gdb.mi/mi-var-invalidate-shlib.exp b/gdb/testsuite/gdb.mi/mi-var-invalidate-shlib.exp index 87d1d4a..c8b9b3e 100644 --- a/gdb/testsuite/gdb.mi/mi-var-invalidate-shlib.exp +++ b/gdb/testsuite/gdb.mi/mi-var-invalidate-shlib.exp @@ -40,10 +40,6 @@ if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable $opt } proc do_test { separate_debuginfo } { - if { $separate_debuginfo } { - gdb_gnu_strip_debug $::shlib_path - } - if { [mi_clean_restart] } { unsupported "failed to start GDB" return @@ -132,6 +128,42 @@ proc do_test { separate_debuginfo } { } } +proc_with_prefix local_not_invalidated { separate_debuginfo } { + if { [mi_clean_restart] } { + unsupported "failed to start GDB" + return + } + + # Start the process once and create varobjs referencing the loaded objfiles. + with_test_prefix "setup" { + mi_load_shlibs $::shlib_path + if { $separate_debuginfo } { + mi_load_shlibs ${::shlib_path}.debug + } + + mi_gdb_reinitialize_dir $::srcdir/$::subdir + mi_gdb_load $::binfile + + mi_runto foo -pending + mi_next "next" + mi_create_varobj local_var local_var "create local varobj" + } + + # At this point we are stopped in the shared library. If we reload symbols + # for the main binary, symbols for the shared library remain valid. A + # varobj tracking variables in the scope of the shared library only should + # not be invalidated. + mi_gdb_load ${::binfile} + mi_gdb_test "-var-update local_var" \ + "\\^done,changelist=\\\[\\\]" \ + "local_var preserved" +} + foreach_with_prefix separate_debuginfo {0 1} { + if { $separate_debuginfo } { + gdb_gnu_strip_debug $::shlib_path + } + do_test $separate_debuginfo + local_not_invalidated $separate_debuginfo } |