This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [patch v1] nptl: smarter not-parallel-ing
- From: Joseph Myers <joseph at codesourcery dot com>
- To: Florian Weimer <fweimer at redhat dot com>
- Cc: Carlos O'Donell <carlos at redhat dot com>, DJ Delorie <dj at redhat dot com>, <libc-alpha at sourceware dot org>
- Date: Wed, 14 Aug 2019 16:47:27 +0000
- Subject: Re: [patch v1] nptl: smarter not-parallel-ing
- Ironport-sdr: 5S5Ji4pjQQoo4L+xZJc02/CEK7PQGPOy46tkh13nvsLabusTI/2Y61dhkYj+2lTMbOZTR/QTr5 IpDOgllcufvd2Kr7SGkfmPi+T5yabnY0RFOW4vW58mnRHVC2mhFcIUcHv9zUsvcqgd/cfCECG6 Hyx+uhvl7YldR6lAy81dHc+Hy3Gm8x3TJrGfMlkSqRxIZ0ClaacNH9x7lnw+pgngKAiRvtGPyW nkFYgzo4nEuUBAoyJjh97wzeqTRVLLrxZn1c87Oxx14IpqvzziBcflUtQtuCK66l28CKKpALRi +58=
- Ironport-sdr: 8r5ADS2aS4iGAiaUJLdJoE1sIkTdiw7m5kyEOGTHn+3rLR4cJF2uiIpTbATix+2f9dg1H2CH9T ICgNQu8swcy+bcysqOs1h4pQ7XZllcoWy1RGK525pqlKjzQ1yjorb4wV8Ku7gJr7AE/91WFLNv NA7UuPYq8g0eQ8dZ6iQ2OlqmEzD2GnKALZbymahLd6YXGVSG31czfzQ30pqjS3z5B2ObK1Dm7L 1z9qp6Ob4p4/hpd67qZHSGb8HZR96BtqNI4Go9YFPum8ei2RfJNhjg5T2d0qLwyKs3mP0NwzXF wjQ=
- References: <xnk1bgoi3e.fsf@greed.delorie.com> <87h86jdidq.fsf@oldenburg2.str.redhat.com> <f472b71a-9553-3ba7-406b-636be09a588e@redhat.com> <871rxndez9.fsf@oldenburg2.str.redhat.com>
On Wed, 14 Aug 2019, Florian Weimer wrote:
> * Carlos O'Donell:
>
> > On 8/14/19 8:46 AM, Florian Weimer wrote:
> >> * DJ Delorie:
> >>
> >>> +$(foreach pair,$(TESTPAIRS),$(eval $(pair)))
> >>
> >> One more comment: In the past, there was a sustained objection against
> >> using $(eval …) in makefiles. If I recall correctly, the person who
> >> voiced the objection no longer participates in the project.
> >
> > Do we remember why?
>
> Which part? No deep technical reason was given at the time for the
> $(eval …) objection, and I did not make an attempt do understand this
> position.
>
> For target-generating loops, the workaround has always been to use
> o-iterator.mk, which I personally do not consider an improvement.
We could revisit Andreas's patch
<https://sourceware.org/ml/libc-alpha/2012-07/msg00250.html>, updated for
subsequently added uses of o-iterator.mk, and see if it still gives the
reported speedup.
--
Joseph S. Myers
joseph@codesourcery.com