Message ID | 20200213055921.218338-1-tamur@google.com |
---|---|
State | New |
Headers | show |
Thanks. On 2/13/20 2:59 AM, Ali Tamur via gdb-patches wrote: > Process a die with DW_TAG_skeleton_unit tag similar to > DW_TAG_partial_unit and DW_TAG_compile_unit. > > gdb/ChangeLog: > > * dwarf2/read.c (process_die): Add DW_TAG_skeleton_unit. > --- > gdb/dwarf2/read.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c > index 7edbd9d7df..bd19a41eef 100644 > --- a/gdb/dwarf2/read.c > +++ b/gdb/dwarf2/read.c > @@ -9784,6 +9784,7 @@ process_die (struct die_info *die, struct dwarf2_cu *cu) > break; > case DW_TAG_compile_unit: > case DW_TAG_partial_unit: > + case DW_TAG_skeleton_unit: > read_file_scope (die, cu); > break; > case DW_TAG_type_unit: > Should we at least have some small smoke test to check we do support reading DW_TAG_skeleton_unit now? Is there something functionally different compared to the partial and compile unit? I'm wondering if mentioning this in the NEWS would also be nice, maybe with a reference of when compilers started generating this. What do you think?
diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c index 7edbd9d7df..bd19a41eef 100644 --- a/gdb/dwarf2/read.c +++ b/gdb/dwarf2/read.c @@ -9784,6 +9784,7 @@ process_die (struct die_info *die, struct dwarf2_cu *cu) break; case DW_TAG_compile_unit: case DW_TAG_partial_unit: + case DW_TAG_skeleton_unit: read_file_scope (die, cu); break; case DW_TAG_type_unit: