-
-
Notifications
You must be signed in to change notification settings - Fork 12.6k
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
liquid-dsp 1.2.0 (new formula) #2664
Conversation
Yup, already saw that, but it seems to be a weird inline assembler (?) related error and I'm curious why this only only happens on Mavericks. I guess this could be a compiler issue... |
Could you report it to the developers? Thanks! |
The test build on OS X Mavericks failed with errors like "no such instruction: `vmovss LC1(%rip), %xmm2'" which might relate to AVX intrinsics - at least that's what I've found in a similar issue: OP2/PyOP2#471 The suggested fix is to add the compiler flag "-Wa,-q" which this commit tries to do.
@BrewTestBot Please test this again. |
@muellermartin Why did you close/reopen this? |
@MikeMcQuaid Sorry, I accidentally clicked the wrong button :( |
I just wanted to note that I found a similar issue with OP2/PyOP2#471 which points out that this could be related to AVX instructions. With the last commit I want to check if the build error on Mavericks can be resolved with the suggested fix (adding |
Build looks good, no errors - even on Mavericks. I'll report that issue upstream. Until then this package could be used. Update: This issue is now reported upstream: jgaeddert/liquid-dsp#48 |
depends_on "fftw" => :recommended | ||
# libfec could be an optional dependency, but it's not available, yet | ||
|
||
fails_with :clang do |
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.
We've been extremely 👎 on accepting new formulae which won't build against Clang, particularly if there's no upstream work in progress on making it Clang compatible.
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.
That's understandable. Maybe I can get it to work under LLVM/Clang by fiddling around...
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.
I've found a way to successfully build it with LLVM/Clang instead of GCC with a simple patch which already has an upstream issue (see jgaeddert/liquid-dsp#50). Should I submit another pull request or add it here?
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.
Submitting a PR & then using that PR in patch do
format if it applies cleanly here would be terrific.
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.
I've already submitted a patch upstream (jgaeddert/liquid-dsp#51) Do I understand it correctly that I can add it here in this PR? Is an embedded patch (patch :DATA
+ __END__
) O.K.?
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.
If it will apply cleanly, we'd prefer the:
patch do
url "https://github.com/jgaeddert/liquid-dsp/pull/51.patch"
sha256 ""
end
Format.
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.
Ah, didn't know you also can get a "raw" patch file out of a PR hosted on GitHub. Neat and thanks for this tip!
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.
Added commit with patch.
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.
Aye, you can get diffs if preferred by simply switching it to diff
as well. Very handy feature.
def install | ||
system "./reconf" | ||
system "./configure", "--prefix=#{prefix}" | ||
# Note: "make install" in one step does fail |
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.
Could you report this issue upstream? Thanks!
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.
I'll do that and report back as soon as it's done!
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.
I've now reported this upstream as jgaeddert/liquid-dsp#53 and included a simple PR. Should I add a "patch" with inreplace in the formula to change the relevant file in the Makefile (makefile.in) or just add/replace the comment to mention the upstream bug?
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.
Just add the comment, thanks 👍
I've now added a comment on another upstream issue as suggested. The reported upstream issues (jgaeddert/liquid-dsp#48 and jgaeddert/liquid-dsp#52) do have patches/workarounds until they get resolved and released which seems to take time due to stalling maintenance, but I've heard that this is beeing worked on. |
I might wait another day or two for upstream to reply to your tickets (thanks for filing them!) |
Closing until we get a new stable release with these patches included. |
brew install <formula>
(where<formula>
is the name of the formula you're submitting)?brew audit --strict --online <formula>
(after doingbrew install <formula>
)?