Re: Ghostscript/GhostPDL 9.27 Release Candidate 1

Chris Liddell <[email protected]> Thu, 28 Mar 2019 09:51:32 +0000
Newsgroups gmane.comp.printing.ghostscript.devel
Message-ID <[email protected]>
On 28/03/2019 01:15, Jonas Smedegaard wrote:
> Quoting Chris Liddell (2019-03-27 13:41:34)
>> I have created archives (and Ghostscript Windows installers) for the
>> first public release candidates for Ghostscript and GhostPDL 9.27, and
>> they are available here:
>>
>> https://github.com/ArtifexSoftware/ghostpdl-downloads/releases/tag/9.27rc1
>>
>> Please give them a try on your system if you're interested in helping
>> test the release-in-progress. Your feedback is appreciated.
>>
>> This release involves making a large number of non-standard operators
>> only intended for internal use inaccessible to "user" Postscript jobs.
>> Tools that use heavily non-standard Postscript may be affected by this.
>> If, in your testing, you find such tools, we'd appreciate it if you
>> could encourage the tools author to contact us to the discuss the issue
>> (the best places are #ghostscript on freenode or the gs-devel list).
> 
> Unfortunately Debian builds of the new prerelease segfaults when simply 
> executing "gs --version".
> 
> Here is a build log from a git snapshot few days ago: 
> https://buildd.debian.org/status/fetch.php?pkg=ghostscript&arch=amd64&ver=9.27%7E%7Edc1%7Edfsg-1&stamp=1553515391&raw=0
> 
> Build logs for the actual prerelease will emerge within an hour at 
> https://buildd.debian.org/status/package.php?p=ghostscript&suite=experimental
> 
> I can make the coredump for amd64 build available if needed, or examine 
> it if you help tell me how.
> 
> Debian builds link against system shared libraries, and apply a set of 
> patches: 
> https://salsa.debian.org/printing-team/ghostscript/tree/master/debian/patches

It's a change in the front end API, and we missed it for some of the
callers of that API:

https://tinyurl.com/y4uwbowc

is a commit that fixes the specific problem, and removes the reliance on
the compiler initialising globals in other places (just makes it clearer
what's required).


Thanks!
signature.asc (application/pgp-signature, 488 B)
-----BEGIN PGP SIGNATURE-----

iQEzBAEBCgAdFiEEvaAaDXMS0efwz2frUpDUqObqF2oFAlycmSQACgkQUpDUqObq
F2oaBQf8C8+VzUKwoBpH5LnCwet+jrw/zxkZR7KAnGl6ofM5b4kGvaavHWkoN3Ay
P/pzd9IM77SFVyIloqgieTj4AOmZYMmSw9I5xWvTC/FixHb26RcE7kBenNwbPGb2
xr7UmEJRaFGf48PVrbmXWN9aCaz8N9lMr6+CD29iQdc8I5xfIRo3wIhAsXdXSTSN
uI7EpYxyqGw1QX3NpthUPG8+NN6NQVgd8aooHDTAmpuuKHNfHdO8YL9/U5QDavAj
iRkpewO1eg1/YmlHIctnhFJRnztGFR95DOnPygzXBDyYNgcER7o86+jiV2gZZQPZ
mHmoAqOuDmNyjISyUDl90Bqh0hulEg==
=IqM0
-----END PGP SIGNATURE-----