-
Notifications
You must be signed in to change notification settings - Fork 165
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
Fix GNU readline detection on OpenBSD, and make the configure test for it more robust #3428
Conversation
Codecov Report
@@ Coverage Diff @@
## master #3428 +/- ##
==========================================
+ Coverage 85.17% 85.19% +0.01%
==========================================
Files 699 697 -2
Lines 346081 343975 -2106
==========================================
- Hits 294784 293037 -1747
+ Misses 51297 50938 -359
|
I adjusted the title of this PR to be more suitable for release notes; please re-edit if you disagree; you may also want to change the commit message of your first commit, as it still reference OSX; and squash the second commit into the first. |
@dimpase Ping. |
yep, I should get back to this. |
Ready for review |
@dimpase thanks for the update, I'll take a look. But the commit message still is odd; it says "also, do not wrongly accuse OSX of wrongdoing" but the only other reference to OSX is in the first line of the commit message? So maybe just remove both mentions -- or does OSX have any relevance for this PR after all? |
It probably referred to some already removed comment in configure.ac. I'll edit the commit message, no problem. Needless to say, this change works on OSX if It also should be mentioned that this fixes a bug: testing for |
imroved commit message, no other changes |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
OK, this looks pretty good, but autoconf's horrible syntax with too many braces, brackets, parens and no keywords makes it rather hard to see what is nested where, so I want to check the code out locally now. In the meantime, a few comments/questions. And thanks for working on this!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good now up to some very minor nitpicks, thanks!
@kaashif - perhaps this is now ready for making an OpenBSD GAP package. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks, that LGTM. Would you mind squashing the second commit?
rl_ding test does not correctly identify the needed readline version, so this is changed to rl_bind_keyseq. OpenBSD installs modern readline as libereadline in /usr/local, and its headers in /usr/local/include/ereadline/ Thus we provide custom -I in CFLAGS and custom -L in LDFLAGS even if we use --with-readline=yes without an explicit path in case of libereadline detected. Address PR 3428 reviewer comments
squashed. Thanks for taking care of this. |
@ChrisJefferson @alex-konovalov I'd like to backport to this to stable-4.11, OK? |
That's fine with me |
@dimpase Maybe. I'll make another pass on the weekend at making a package and submitting it. |
For packaging, I recommend talking with/to us -- we can often provide hints how to overcome certain problems. Even if you get stuck and decide to abandon the effort, we are grateful for any constructive feedback, to help improve things (In the end, lack of time/resources/contributors sadly is always a limiting factor) |
Backported to stable-4.11 as f67d98a |
Description
this fixes #3426.
Text for release notes
Better handling of incompatible with GAP readline libraries.
Enable building of GAP on OpenBSD (with ereadline fixed as in
https://marc.info/?l=openbsd-ports&m=155646111509306&w=2)
Further details
Do not forget to run ./autogen + ./configure before testing this.