This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] A few comment cleanups
- From: Doug Evans <dje at google dot com>
- To: Simon Marchi <simon dot marchi at ericsson dot com>
- Cc: gdb-patches <gdb-patches at sourceware dot org>
- Date: Wed, 17 Dec 2014 14:30:18 -0800
- Subject: Re: [PATCH] A few comment cleanups
- Authentication-results: sourceware.org; auth=none
- References: <1418842893-21536-1-git-send-email-simon dot marchi at ericsson dot com> <CADPb22Th5Skd2Ja+6X9UCdWZeRR8BpqSFMP_X_JOZaHm25aY5g at mail dot gmail dot com> <5491F976 dot 50804 at ericsson dot com>
On Wed, Dec 17, 2014 at 1:45 PM, Simon Marchi <simon.marchi@ericsson.com> wrote:
> On 2014-12-17 02:37 PM, Doug Evans wrote:
>> On Wed, Dec 17, 2014 at 11:01 AM, Simon Marchi
>> <simon.marchi@ericsson.com> wrote:
>>> I stumbled upon a few comments that I think are outdated.
>>>
>>> Comment for elfread.c (elf_symfile_init): As far as history goes in git,
>>> I don't see anything related to that.
>>>
>>> Comment for elfread.c (elf_symfile_read): References a parameter that was
>>> removed in 1999.
>>>
>>> Comment for struct sym_fns/sym_offsets: References a parameter that was
>>> changed in 1999.
>>>
>>> gdb/ChangeLog:
>>>
>>> * elfread.c (elf_symfile_init): Remove stale comment.
>>> (elf_symfile_read): Same.
>>> * symfile.h (struct sym_fns): Same.
>>
>> Hi.
>> Ok with one nit, below.
>>
>>> ---
>>> gdb/elfread.c | 13 +------------
>>> gdb/symfile.h | 6 +-----
>>> 2 files changed, 2 insertions(+), 17 deletions(-)
>>>
>>> diff --git a/gdb/elfread.c b/gdb/elfread.c
>>> index b4ec067..93e5045 100644
>>> --- a/gdb/elfread.c
>>> +++ b/gdb/elfread.c
>>> @@ -1236,10 +1236,6 @@ elf_read_minimal_symbols (struct objfile *objfile, int symfile_flags,
>>> We have been initialized by a call to elf_symfile_init, which
>>> currently does nothing.
>>>
>>> - SECTION_OFFSETS is a set of offsets to apply to relocate the symbols
>>> - in each section. We simplify it down to a single offset for all
>>> - symbols. FIXME.
>>> -
>>> This function only does the minimum work necessary for letting the
>>> user "name" things symbolically; it does not read the entire symtab.
>>> Instead, it reads the external and static symbols and puts them in partial
>>> @@ -1418,14 +1414,7 @@ elf_symfile_finish (struct objfile *objfile)
>>> dwarf2_free_objfile (objfile);
>>> }
>>>
>>> -/* ELF specific initialization routine for reading symbols.
>>> -
>>> - It is passed a pointer to a struct sym_fns which contains, among other
>>> - things, the BFD for the file whose symbols are being read, and a slot for
>>> - a pointer to "private data" which we can fill with goodies.
>>> -
>>> - For now at least, we have nothing in particular to do, so this function is
>>> - just a stub. */
>>> +/* ELF specific initialization routine for reading symbols. */
>>>
>>> static void
>>> elf_symfile_init (struct objfile *objfile)
>>> diff --git a/gdb/symfile.h b/gdb/symfile.h
>>> index 1e8c230..b2cb089 100644
>>> --- a/gdb/symfile.h
>>> +++ b/gdb/symfile.h
>>> @@ -354,11 +354,7 @@ struct sym_fns
>>> void (*sym_finish) (struct objfile *);
>>>
>>> /* This function produces a file-dependent section_offsets
>>> - structure, allocated in the objfile's storage, and based on the
>>> - parameter. The parameter is currently a CORE_ADDR (FIXME!) for
>>> - backward compatibility with the higher levels of GDB. It should
>>> - probably be changed to a string, where NULL means the default,
>>> - and others are parsed in a file dependent way. */
>>> + structure, allocated in the objfile's storage. */
>>>
>>> void (*sym_offsets) (struct objfile *, const struct section_addr_info *);
>>
>> Can I ask you to document what the struct section_addr_info * argument is for?
>>
>
> Hmm I am not sure my knowledge allows me to produce a useful comment. Don't hesitate to
> correct/enhance. Here is my first shot, after a bit of code reading:
>
> /* This function produces a file-dependent section_offsets
> structure, allocated in the objfile's storage.
>
> The section_addr_info structure contains the offset of loadable and
> allocated sections, relative to the absolute offsets found in the BFD. */
>
> void (*sym_offsets) (struct objfile *, const struct section_addr_info *);
"works for me"
Thanks!