This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB 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: [3/10] introduce psymtab users


On Mon, Apr 30, 2012 at 11:44 AM, Doug Evans <dje@google.com> wrote:
> On Wed, Apr 25, 2012 at 11:21 AM, Tom Tromey <tromey@redhat.com> wrote:
>> @@ -165,6 +169,11 @@ partial_map_symtabs_matching_filename (struct objfile *objfile,
>>
>> ? ALL_OBJFILE_PSYMTABS_REQUIRED (objfile, pst)
>> ? {
>> + ? ?/* We can skip shared psymtabs here, because any file name will be
>> + ? ? ? attached to the unshared psymtab. ?*/
>> + ? ?if (pst->users != NULL)
>> + ? ? ?continue;
>> +
>> ? ? if (FILENAME_CMP (name, pst->filename) == 0
>> ? ? ? ?|| (!is_abs && compare_filenames_for_search (pst->filename,
>> ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? name, name_len)))
>
> The abstraction feels broken if ALL_OBJFILE_PSYMTABS* includes these
> special shared psymtabs.
>
> [This is akin to the objfile list including separate debug file objfiles.]

Actually, I take that back.
Looking into something unrelated, I can understand implementing them
that way (given the source as it is today).
Still, IWBN to have an iterator that did just iterate over the "real" psymtabs.
e.g., have a version of ALL_OBJFILE_PSYMTABS_REQUIRED that skipped
shared psymtabs so the caller didn't have to do that.


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