From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Bavier Subject: RE: 01/01: gnu: fftw: Build SIMD codelets. Date: Tue, 8 May 2018 21:01:29 +0000 Message-ID: References: <20180417212549.1283.62731@vcs0.savannah.gnu.org> <20180417212551.BB270208E4@vcs0.savannah.gnu.org> <87efisjp91.fsf@netris.org> <87r2mlkgru.fsf@netris.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:57327) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fG9jv-00050W-H2 for guix-devel@gnu.org; Tue, 08 May 2018 17:01:40 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fG9js-0008U3-Ee for guix-devel@gnu.org; Tue, 08 May 2018 17:01:39 -0400 Received: from esa1.cray.iphmx.com ([68.232.142.33]:19686) by eggs.gnu.org with esmtps (TLS1.0:RSA_ARCFOUR_SHA1:16) (Exim 4.71) (envelope-from ) id 1fG9jr-0008TQ-TK for guix-devel@gnu.org; Tue, 08 May 2018 17:01:36 -0400 In-Reply-To: <87r2mlkgru.fsf@netris.org> Content-Language: en-US List-Id: "Development of GNU Guix and the GNU System distribution." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org Sender: "Guix-devel" To: 'Mark H Weaver' Cc: "'guix-devel@gnu.org'" Ok, if you feel confident, that's fine with me. Otherwise, I have a patch that I've tested on x86-64, and in the middle of = testing for armhf that I can push in a few hours. I could easily add a com= mit that remove the "--enable-armv7a-cntvct" flag. Eric Bavier, Scientific Libraries, Cray Inc. > -----Original Message----- > From: Mark H Weaver [mailto:mhw@netris.org] > Sent: Tuesday, May 08, 2018 3:57 PM > To: Eric Bavier > Cc: guix-devel@gnu.org > Subject: Re: 01/01: gnu: fftw: Build SIMD codelets. >=20 > Hi Eric, >=20 > Mark H Weaver writes: >=20 > > ericbavier@centurylink.net (Eric Bavier) writes: > > > >> bavier pushed a commit to branch core-updates in repository guix. > >> > >> commit 65bb22796f854cbc3eae053a80b1d64365dad376 > >> Author: Eric Bavier > >> Date: Fri Apr 6 10:53:06 2018 -0500 > >> > >> gnu: fftw: Build SIMD codelets. > >> > >> * gnu/packages/algebra.scm (fftw)[arguments]: Remove 'no-native > phase; use > >> configure cache value instead. Add configure flags for SIMD codel= ets. > >> (fftwf)[arguments]: Add neon configuration flag for 32-bit arm. > >> (fftw-avx): Remove variable. > > > > [...] > > > >> @@ -560,7 +569,10 @@ cosine/ sine transforms or DCT/DST).") > >> (arguments > >> (substitute-keyword-arguments (package-arguments fftw) > >> ((#:configure-flags cf) > >> - `(cons "--enable-float" ,cf)))) > >> + (if (string-prefix? "arm" (or (%current-target-system) > >> + (%current-system))) > >> + `(cons "--enable-neon" ,cf) > >> + cf)))) > > > > Did you intend to remove the "--enable-float" configure flag in fftwf? > > You didn't mention this change in the commit log. > > > > The description of fftwf appends "Single-precision version" to the > > description from fftw, but since your commit above, the fftw and fftwf > > packages are identical except on armhf, as far as I can tell. On > > armhf, the build now fails with "configure: error: NEON requires > > single precision". > > > > https://hydra.gnu.org/build/2674813/nixlog/1/tail-reload > > > > Can you take a look? >=20 > Another problem with the above commit has now become apparent. In > addition to the 'fftwf' failure at configuration time on armhf, the 'fftw= ' > package also now fails its test suite on our armhf build slaves. > While running the test suite, a test process is killed with SIGILL (Illeg= al > instruction). >=20 > I looked into it, and it appears that the "--enable-armv7a-cntvct" > configure flag that you added should be removed. Adding it was a good > guess, but apparently the ARMv7-a CNTVCT instruction is not available fro= m > user mode unless you use a specially patched kernel to enable it. >=20 > For details, see: >=20 > https://github.com/FFTW/fftw3/blob/master/README-perfcnt.md > https://github.com/thoughtpolice/enable_arm_pmu > https://neocontra.blogspot.com/2013/05/user-mode-performance- > counters-for.html >=20 > Since these issues are delaying the build-out of 'core-updates' on armhf > with ~1K dependency failures, and I'm reasonably confident what needs to > be done here, I'll go ahead and apply untested fixes for these issues to > 'core-updates' soon. >=20 > Regards, > Mark