[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Reproduce-devel] [bug #56588] FFTW not building single-precision float
From: |
Mohammad Akhlaghi |
Subject: |
[Reproduce-devel] [bug #56588] FFTW not building single-precision float |
Date: |
Sun, 7 Jul 2019 12:02:38 -0400 (EDT) |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:67.0) Gecko/20100101 Firefox/67.0 |
URL:
<https://savannah.nongnu.org/bugs/?56588>
Summary: FFTW not building single-precision float
Project: Reproducible paper template
Submitted by: makhlaghi
Submitted on: Sun 07 Jul 2019 05:02:37 PM BST
Category: Software
Severity: 3 - Normal
Item Group: Output not reasonable
Status: In Progress
Privacy: Public
Assigned to: makhlaghi
Open/Closed: Open
Discussion Lock: Any
_______________________________________________________
Details:
On the computers of Zahra Sharbaf and Hamed Altafi, we noticed that
SExtractor's build was complaining about not having a single-precision FFTW
library! This was confusing because it was being installed with no problem on
other systems so far.
After checking on my own computer (with `ldd .local/bin/sex'), I noticed that
it was actually building with the single precision FFTW library (libfftw3f) of
my host system (in `/usr/lib'), not the pipeline's build.
After inspecting ArchLinux's build of FFTW
<https://git.archlinux.org/svntogit/packages.git/tree/trunk/PKGBUILD?h=packages/fftw>,
I noticed that they are using the `--enable-float' configure option, not
`--enable-single' that we are using. In the output of `./configure --help', it
says these two are the same, so we were using the latter. However, after using
the former, we now have `libfftw3f' in the pipeline and so SExtractor links
with it.
I am now busy finalizing the fix now.
_______________________________________________________
Reply to this item at:
<https://savannah.nongnu.org/bugs/?56588>
_______________________________________________
Message sent via Savannah
https://savannah.nongnu.org/
- [Reproduce-devel] [bug #56588] FFTW not building single-precision float,
Mohammad Akhlaghi <=