public inbox for gentoo-dev@lists.gentoo.org
 help / color / mirror / Atom feed
From: "Michał Górny" <mgorny@gentoo.org>
To: gentoo-dev@lists.gentoo.org
Cc: "Michał Górny" <mgorny@gentoo.org>
Subject: [gentoo-dev] [PATCH 14/15] scons-utils.eclass: Use nproc when --jobs is used without an argument
Date: Fri,  1 Jan 2016 17:41:20 +0100	[thread overview]
Message-ID: <1451666481-22145-15-git-send-email-mgorny@gentoo.org> (raw)
In-Reply-To: <1451666481-22145-1-git-send-email-mgorny@gentoo.org>

Try to guess the number of processors when --jobs is passed without
an argument. We can't use a high number equivalent to GNU make behavior
(no limit) since SCons does not have an equivalent of --load-avg option.
Still, this is better than assuming some random, fixed number.
---
 eclass/scons-utils.eclass   | 30 ++++++++++++++++++++++++++++--
 eclass/tests/scons-utils.sh |  2 +-
 2 files changed, 29 insertions(+), 3 deletions(-)

diff --git a/eclass/scons-utils.eclass b/eclass/scons-utils.eclass
index 3185282..89618f9 100644
--- a/eclass/scons-utils.eclass
+++ b/eclass/scons-utils.eclass
@@ -149,6 +149,32 @@ escons() {
 	return ${ret}
 }
 
+# @FUNCTION: _scons_get_default_jobs
+# @INTERNAL
+# @DESCRIPTION:
+# Output the default number of jobs, used if -j is used without
+# argument. Tries to figure out the number of logical CPUs, falling
+# back to hardcoded constant.
+_scons_get_default_jobs() {
+	local nproc
+
+	if type -P nproc &>/dev/null; then
+		# GNU
+		nproc=$(nproc)
+	elif type -P python &>/dev/null; then
+		# fallback to python2.6+
+		# note: this may fail (raise NotImplementedError)
+		nproc=$(python -c 'import multiprocessing; print(multiprocessing.cpu_count());' 2>/dev/null)
+	fi
+
+	if [[ ${nproc} ]]; then
+		echo $(( nproc + 1 ))
+	else
+		# random default
+		echo 5
+	fi
+}
+
 # @FUNCTION: _scons_clean_makeopts
 # @INTERNAL
 # @USAGE: [makeflags] [...]
@@ -192,7 +218,7 @@ _scons_clean_makeopts() {
 					shift
 				else
 					# no value means no limit, let's pass a random int
-					new_makeopts+=( ${1}=5 )
+					new_makeopts+=( ${1}=$(_scons_get_default_jobs) )
 				fi
 				;;
 			# strip other long options
@@ -215,7 +241,7 @@ _scons_clean_makeopts() {
 								new_optstr+="j ${2}"
 								shift
 							else
-								new_optstr+="j 5"
+								new_optstr+="j $(_scons_get_default_jobs)"
 							fi
 							;;
 						# otherwise, everything after -j is treated as an arg
diff --git a/eclass/tests/scons-utils.sh b/eclass/tests/scons-utils.sh
index 6355c54..fcb5125 100755
--- a/eclass/tests/scons-utils.sh
+++ b/eclass/tests/scons-utils.sh
@@ -28,7 +28,7 @@ test-scons_clean_makeopts() {
 }
 
 # jobcount expected for non-specified state
-jc=5
+jc=$(_scons_get_default_jobs)
 # failed test counter
 failed=0
 
-- 
2.6.4



  parent reply	other threads:[~2016-01-01 16:49 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-01-01 16:41 [gentoo-dev] [PATCH 00/15] scons-utils.eclass: EAPI 6, better docs and cleanup Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 01/15] scons-utils.eclass: Modernize the example not to rely on myescons Michał Górny
2016-01-01 16:57   ` [gentoo-dev] " Michael Palimaka
2016-01-01 17:15     ` Michał Górny
2016-01-01 17:42       ` Michael Palimaka
2016-01-02 13:38         ` Michael Orlitzky
2016-01-01 16:41 ` [gentoo-dev] [PATCH 02/15] scons-utils.eclass: Modernize the example to use usex Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 03/15] scons-utils.eclass: Describe common issues with scons Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 04/15] scons-utils.eclass: escons doc, 'die' does not respect nonfatal Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 05/15] scons-utils.eclass: escons, invert EAPI check to cover future EAPIs Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 06/15] scons-utils.eclass: escons, respect nonfatal in EAPI 6 Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 07/15] scons-utils.eclass: Deprecate myesconsargs, and kill it " Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 08/15] scons-utils.eclass: Deprecate use_scons, ban " Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 09/15] scons-utils.eclass: tests, be more verbose on tests being performed Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 10/15] scons-utils.eclass: scons_clean_makeopts, mark internal Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 11/15] scons-utils.eclass: _scons_clean_makeopts, fix result caching Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 12/15] scons-utils.eclass: _scons_clean_makeopts, stop exporting cache vars Michał Górny
2016-01-01 16:41 ` [gentoo-dev] [PATCH 13/15] scons-utils.eclass: _scons_clean_makeopts, clean up and simplify Michał Górny
2016-01-01 16:41 ` Michał Górny [this message]
2016-01-01 16:41 ` [gentoo-dev] [PATCH 15/15] scons-utils.eclass: Enable EAPI 6 Michał Górny
2016-01-08  5:15 ` [gentoo-dev] [PATCH 00/15] scons-utils.eclass: EAPI 6, better docs and cleanup Michał Górny

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1451666481-22145-15-git-send-email-mgorny@gentoo.org \
    --to=mgorny@gentoo.org \
    --cc=gentoo-dev@lists.gentoo.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox