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
FreeBSD issues #1420
Comments
Reverting both files to the previous version fixes the build on FreeBSD. I've not yet tested on OpenBSD but would expect the same result. |
Reverting is easy, but what exactly are we going to lose by reverting? Ping @Kaiepi, what are your thoughts on this issue? (will simply revert relatively soon unless there's new information) |
This reverts commit 28eaf97. See rakudo/rakudo#1420
I did a rough revert of these commits, and bumped nqp & moar. @stmuk can you please check that everything is working now? If yes, this is no longer a blocker, and we can close the two referenced tickets. This ticket will remain open so that we can coordinate our efforts on rereverting the commits in a way that works. |
I was enable to build both on OpenBSD and FreeBSD now to create
The original commits were intended to support building with "egcc" which is already possible. egcc is a OpenBSDism for an externally (via ports) installed gcc package Standard base gcc is 4.2.1 and I installed 4.9.4 via "pkg_add" I was enable to use egcc without any code changes but purely using configuration options on the command line.
confirmed with
|
Good! Thanks. Blocker label removed. Even though you say that it is unneeded, I'l leave it open for a bit (e.g. if @Kaiepi has something to say). |
Ran into the build issues myself earlier... embarrassing I'm alright with reverting the first commit, but I think MoarVM/MoarVM@28eaf97 is likely to blame for the problems here. Fixing the syntax error here fixes the OpenBSD build errors on my machine. Does reverting this line fix the ones on FreeBSD?
True, I wasn't aware of that when I made the commit. Would trying to use egcc if no compiler flag's given be a good idea? |
@Kaiepi I'm no expert on FreeBSD/OpenBSD, so can't answer the questions, but just wanted to let you know that new pull requests are welcome. |
I don't see any reason for any of the patches. It's not clear why it was
attempted to patch FreeBSD when the objective was to patch OpenBSD.
It's easy to build with another compiler without the patches.
I don't think its a good idea to use a non-standard egcc by default if it
is installed.
The default compiler should always be used unless the user explicitly tells
the config otherwise.
We probably do need to advertise/document how to use other compilers more
clearly so I'd favour document pull requests for documentation about the
build rather than making the build config more complex.
The whole build config is a bit fragile and I've messed it up myself before!
S
…On 22 January 2018 at 06:16, Kaiepi ***@***.***> wrote:
Ran into the build issues myself earlier... embarrassing
I'm alright with reverting the first commit, but I think MoarVM/MoarVM@
28eaf97
<MoarVM/MoarVM@28eaf97>
is likely to blame here. Fixing the syntax error here
<MoarVM/MoarVM@28eaf97#diff-2a9806f1c161480015223272ecd41308R462>
fixes the OpenBSD build errors on my machine. Does reverting this line
<MoarVM/MoarVM@28eaf97#diff-2a9806f1c161480015223272ecd41308R483>
fix the ones on FreeBSD?
The original commits were intended to support building with "egcc" which
is already possible.
True, I wasn't aware of that when I made the commit. Would trying to use
egcc if no compiler flag's given be a good idea?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#1420 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AEUuY4imDZbe4JxsJ0feuOMfx7ErINWTks5tNCesgaJpZM4RkqNk>
.
--
4096R/EA75174B Steve Mynott <steve.mynott@gmail.com>
|
Ah, alright |
One thing which would be very useful is for more people to set up "smokers"
to automatically build *BSDs.
Maybe a shell script run out of cron talking to an IRC bot or something
which just warns if a platform is broken.
Mostly people build on macOS and linux and the *BSDs do need more love.
S
…On 22 January 2018 at 08:34, Kaiepi ***@***.***> wrote:
Ah, alright
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#1420 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AEUuY9BIFTP2vV2EMAj80L02FDOguCSeks5tNEgZgaJpZM4RkqNk>
.
--
4096R/EA75174B Steve Mynott <steve.mynott@gmail.com>
|
Late, but the documentation is something I'll get a start on. IRC bots are something I can do, but there's a couple issues related to that'll need to get fixed before tests will be able to pass on OpenBSD. One of them also fails on FreeBSD, but won't prevent me from being able to write the bot entirely on that platform for now, just as long as it can connect to FreeNode without having to resolve any domain names. I'll refer to this issue when I post them Edit: spoke a little too soon about the DNS issue. It seems to only happen for me on FreeBSD |
Wait, isn't it fixed now? |
Yes, I'm pretty sure the build issue has been fixed. |
See these tickets:
Marking it as a blocker, at least for now.
The text was updated successfully, but these errors were encountered: