public inbox for cygwin-apps@cygwin.com
 help / color / mirror / Atom feed
* [RFC] Minimal Debuginfo by default
@ 2017-11-15  7:38 Yaakov Selkowitz
  2017-11-15  9:09 ` David Stacey
  2017-11-17 14:02 ` Jon Turney
  0 siblings, 2 replies; 4+ messages in thread
From: Yaakov Selkowitz @ 2017-11-15  7:38 UTC (permalink / raw)
  To: cygwin-apps


[-- Attachment #1.1: Type: text/plain, Size: 332 bytes --]

The following concept would allow for sensible backtraces without
installing a -debuginfo, at the expense of a moderate size increase of
binaries (particularly with C++ code):

https://fedoraproject.org/wiki/Changes/MingwMiniDebugInfo

The patch for cygport would be minimal.  Is it worth the size increase?

-- 
Yaakov


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [RFC] Minimal Debuginfo by default
  2017-11-15  7:38 [RFC] Minimal Debuginfo by default Yaakov Selkowitz
@ 2017-11-15  9:09 ` David Stacey
  2017-11-19 21:02   ` Yaakov Selkowitz
  2017-11-17 14:02 ` Jon Turney
  1 sibling, 1 reply; 4+ messages in thread
From: David Stacey @ 2017-11-15  9:09 UTC (permalink / raw)
  To: cygwin-apps

On 15/11/17 07:38, Yaakov Selkowitz wrote:
> The following concept would allow for sensible backtraces without
> installing a -debuginfo, at the expense of a moderate size increase of
> binaries (particularly with C++ code):
>
> https://fedoraproject.org/wiki/Changes/MingwMiniDebugInfo

Did you mean https://fedoraproject.org/wiki/Features/MiniDebugInfo 
Presumably you intend this for Cygwin binaries.

> The patch for cygport would be minimal.  Is it worth the size increase?

The benefit to Fedora is obvious, in that it would increase the quality 
of the automated bug reports. Do we have an equivalent in Cygwin (or is 
one planned)? Right now, we don't have so many backtraces sent in to the 
main list, but who knows - if they were easier to generate then we might.

How would this work for Cygwin? Would we require all the lib.* packages 
to be rebuilt?

Dave.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [RFC] Minimal Debuginfo by default
  2017-11-15  7:38 [RFC] Minimal Debuginfo by default Yaakov Selkowitz
  2017-11-15  9:09 ` David Stacey
@ 2017-11-17 14:02 ` Jon Turney
  1 sibling, 0 replies; 4+ messages in thread
From: Jon Turney @ 2017-11-17 14:02 UTC (permalink / raw)
  To: cygwin-apps

On 15/11/2017 07:38, Yaakov Selkowitz wrote:
> The following concept would allow for sensible backtraces without
> installing a -debuginfo, at the expense of a moderate size increase of
> binaries (particularly with C++ code):
> 
> https://fedoraproject.org/wiki/Changes/MingwMiniDebugInfo
> 
> The patch for cygport would be minimal.  Is it worth the size increase?

Unrelated, but we should probably also turn back on 
--compress-debug-sections for debuginfo packages, since we should now 
have a binutils where this works...

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [RFC] Minimal Debuginfo by default
  2017-11-15  9:09 ` David Stacey
@ 2017-11-19 21:02   ` Yaakov Selkowitz
  0 siblings, 0 replies; 4+ messages in thread
From: Yaakov Selkowitz @ 2017-11-19 21:02 UTC (permalink / raw)
  To: cygwin-apps


[-- Attachment #1.1: Type: text/plain, Size: 1310 bytes --]

On 2017-11-15 03:09, David Stacey wrote:
> On 15/11/17 07:38, Yaakov Selkowitz wrote:
>> The following concept would allow for sensible backtraces without
>> installing a -debuginfo, at the expense of a moderate size increase of
>> binaries (particularly with C++ code):
>>
>> https://fedoraproject.org/wiki/Changes/MingwMiniDebugInfo
> 
> Did you mean https://fedoraproject.org/wiki/Features/MiniDebugInfo
> Presumably you intend this for Cygwin binaries.

Yes, but as we're not an ELF platform, I doubt we support the method
used to implement MiniDebugInfo, so the implementation would be closer
to MingwMiniDebugInfo.

>> The patch for cygport would be minimal.  Is it worth the size increase?
> 
> The benefit to Fedora is obvious, in that it would increase the quality
> of the automated bug reports. Do we have an equivalent in Cygwin (or is
> one planned)? Right now, we don't have so many backtraces sent in to the
> main list, but who knows - if they were easier to generate then we might.
> 
> How would this work for Cygwin? Would we require all the lib.* packages
> to be rebuilt?

Not immediately, but I do think we are getting to the point where we
really need a mass rebuild anyway.  Therefore, I'm trying to queue up
now any changes of such scope.

-- 
Yaakov


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2017-11-19 21:02 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-11-15  7:38 [RFC] Minimal Debuginfo by default Yaakov Selkowitz
2017-11-15  9:09 ` David Stacey
2017-11-19 21:02   ` Yaakov Selkowitz
2017-11-17 14:02 ` Jon Turney

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).