Re: planning to upload binutils 2.35.2
On 1/28/21 8:36 PM, Paul Gevers wrote:
> Hi Matthias,
>
> On 27-01-2021 22:42, Matthias Klose wrote:
>> I have been following the way the linux source package was uploaded. Apparently
>> the package entered unstable with just an announcement like this. And more than
>> one time.
>
> For linux there was alignment, but see below.
>
>>> So, can you please clarify why you think these changes are needed? What
>>> are the risks of including or not including these changes? How are the
>>> risks mitigated?
>>
>> staging in experimental is not possible, unless you remove 2.36, or override it
>> bumping the epoch.
>
> (or a +really version number).
>
>> - PR27218 is an obvious bug fix, avoiding a segfault.
>> - DWARF5 is not enabled by default, the DWARF5 fixes are
>> required for GCC 11 defaulting to use DWARF5. And no,
>> I'm not planning to upload gcc-11 to unstable.
>>
>> I'm very unhappy about the private decision making for some uploads, while
>> showing a pedantic attitude towards others.
>
> I must confess that indeed the alignment with the Release Team on linux
> uploads happened in private. It shouldn't have, or at least the outcome
> should have been public.
>
>> - PR27218 is an obvious bug fix, avoiding a segfault.
>
> Sound OK to have.
>
>> - DWARF5 is not enabled by default, the DWARF5 fixes are
>> required for GCC 11 defaulting to use DWARF5.
>
> https://release.debian.org/britney/pseudo-excuses-experimental.html#binutils
> (for 2.36-1) shows a regression for glibc. Hence we're not totally
> confident. If it's not the default, why do we want this feature now?
the log ends with:
----------------8<----------------8<----------------8<-----------------
WARNING: log file truncated at 20 MB (before compression)
----------------8<----------------8<----------------8<-----------------
autopkgtest [01:21:39]: @@@@@@@@@@@@@@@@@@@@ summary
rebuild FAIL non-zero exit status 2
> We would be happy with either of the following:
> 1) upload to unstable with PR27218 only
> 2) upload to experimental first (with a 2.36+really2.35.2 version) to
> check all is fine.
so I don't see what an upload for 2) would provide you with more information.
Reply to: