New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ExtUtils::Command::MM not installing files in vendor arch-specific directories #16014
Comments
From al.dunsmuir@sympatico.caCreated by al.dunsmuir@sympatico.caPerl Info
|
From al.dunsmuir@sympatico.caJust noticed that ExtUtils::MakeMaker 7.30 was reeleased today. -- |
From [Unknown Contact. See original ticket]Just noticed that ExtUtils::MakeMaker 7.30 was reeleased today. -- |
From al.dunsmuir@sympatico.caOn Mon, 12 Jun 2017 14:56:30 -0700, al.dunsmuir@sympatico.ca wrote:
Same problem persists after updating embedded EU:MM to 7.30 release. |
From al.dunsmuir@sympatico.caOn Mon, 12 Jun 2017 18:03:59 -0700, al.dunsmuir@sympatico.ca wrote:
Attached are the build logs from AIX 5.3 for 5.26.0 (with updated |
From @tonycozOn Wed, 14 Jun 2017 15:29:10 -0700, al.dunsmuir@sympatico.ca wrote:
Could I see the generated Makefile from the 5.26 build please? Tony |
The RT System itself - Status changed from 'new' to 'open' |
From al.dunsmuir@sympatico.caOn Tue, 20 Jun 2017 22:28:52 -0700, tonyc wrote:
Tony, Attached are the Makefiles generated for the builds on both 5.22 and 5.26. Please let me apologize for not noticing something important - the arch-specific builds with 5.22 are also failing the same way! All of my 5.22 builds on my main system were completed by August of last year. I normally only use one system as a build environment, as my other 2 AIX boxes are tiny B50s (32-bit only). I rebuilt the 5.22 after getting the 5.26 failure (to get logs and verify the tests passed), and then used that 5.22 and a 32-bit subset of my main systems build setup to create the perl build environment on a B50. I guess in the rush to generate a log, I missed the glob failure messages on 5.22 entirely 8^( The use of the wrong directory for arch builds is likely a result of an unintended change in my main perl build environment. Hopefully we can quickly track that down and correct it so that both the 5.22 and 5.26 arch builds work. Thanks again for your assistance, -- |
From al.dunsmuir@sympatico.ca-On Thu, 22 Jun 2017 05:01:28 -0700, al.dunsmuir@sympatico.ca wrote:
Makefile.526 at line 93 has: Makefile.526 at line 97 has: Looks like the problem is caused during Makefile generation. Attached is the build log for perl-Variable-Magic 5.26, with verbose testing turned on (in case it adds something useful). |
From al.dunsmuir@sympatico.caOn Thu, 22 Jun 2017 09:33:39 -0700, al.dunsmuir@sympatico.ca wrote:
As I understand it, EU:MM gets the configuration from Config.pm [bob:/apps/rpmbuild/BUILD/perl-5.26.0/32bit]$grep -in vendorarch config.sh Somehow the value from config.sh is being altered. |
From [Unknown Contact. See original ticket]On Thu, 22 Jun 2017 09:33:39 -0700, al.dunsmuir@sympatico.ca wrote:
As I understand it, EU:MM gets the configuration from Config.pm [bob:/apps/rpmbuild/BUILD/perl-5.26.0/32bit]$grep -in vendorarch config.sh Somehow the value from config.sh is being altered. |
From al.dunsmuir@sympatico.caOn Thu, 22 Jun 2017 13:20:05 -0700, al.dunsmuir@sympatico.ca wrote:
My original RPM .spec file recipe was: I changed this to: It appears that overriding INSTALLVENDORLIB without overriding INSTALLVENDORARCH caused the INSTALLVENDORARCH value to be corrupted. I tried the minimalistic version: So it appears that the EU:MM handling of just INSTALLVENDORLIB results in an invalid INSTALLVENDORARCH value. These overrides were from perl RPM builds back to 5.8.8 days, and appear to no longer be required. Since I'm doing a full rebase to 5.26.0 anyway, I will remove them from all my .spec file recipes. That solves my immediate problem, but EU:MM should be corrected in this area. |
From [Unknown Contact. See original ticket]On Thu, 22 Jun 2017 13:20:05 -0700, al.dunsmuir@sympatico.ca wrote:
My original RPM .spec file recipe was: I changed this to: It appears that overriding INSTALLVENDORLIB without overriding INSTALLVENDORARCH caused the INSTALLVENDORARCH value to be corrupted. I tried the minimalistic version: So it appears that the EU:MM handling of just INSTALLVENDORLIB results in an invalid INSTALLVENDORARCH value. These overrides were from perl RPM builds back to 5.8.8 days, and appear to no longer be required. Since I'm doing a full rebase to 5.26.0 anyway, I will remove them from all my .spec file recipes. That solves my immediate problem, but EU:MM should be corrected in this area. |
From @tonycozOn Fri, 23 Jun 2017 05:51:10 -0700, al.dunsmuir@sympatico.ca wrote:
Ok, it sounds like you've tracked this down yourself. EU::MM is maintained on CPAN, bug in it should be reported via the Extutils::MakeMaker queue at: https://rt.cpan.org/Public/Dist/Display.html?Name=ExtUtils-MakeMaker Tony |
From al.dunsmuir@sympatico.caOn Sun, 25 Jun 2017 17:04:53 -0700, tonyc wrote:
Tony, Created: -- |
@iabyn - Status changed from 'open' to 'resolved' |
Migrated from rt.perl.org#131559 (status was 'resolved')
Searchable as RT131559$
The text was updated successfully, but these errors were encountered: