test suite reviews and discussions
 help / color / mirror / Atom feed
From: ohilyard@iol.unh.edu
To: dts@dpdk.org
Cc: lijuan.tu@intel.com, Owen Hilyard <ohilyard@iol.unh.edu>
Subject: [dts] [PATCH] meson compile: Improve compile times
Date: Fri, 22 Jan 2021 12:05:28 -0500	[thread overview]
Message-ID: <20210122170529.672048-1-ohilyard@iol.unh.edu> (raw)

From: Owen Hilyard <ohilyard@iol.unh.edu>

Increased compile timeout
Removed workers parameter from ninja compilation command

After a review of CI metrics, we noticed that dts runtimes had increased
greatly over the last few months. A review of recent logs showed
timeouts for compiling dpdk. Upon review, we discovered that, if the
"dut_cores" parameter is not set in conf/crbs.cfg, then dts uses the
default value, which is a single core, as the core list. The number of cores
in this list was then used the number of workers for ninja. This causes DTS
to attempt to compile DPDK serially, resulting in greatly increased compile
times. The IOL's development machine saw a 30x compile speed increase from
the included changes.

Signed-off-by: Owen Hilyard <ohilyard@iol.unh.edu>
---
 framework/project_dpdk.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/framework/project_dpdk.py b/framework/project_dpdk.py
index 915c648f..af542a91 100644
--- a/framework/project_dpdk.py
+++ b/framework/project_dpdk.py
@@ -281,7 +281,7 @@ class DPDKdut(Dut):
         """
         Build DPDK source code on linux use meson
         """
-        build_time = 900
+        build_time = 1200
         target_info = target.split('-')
         arch = target_info[0]
         machine = target_info[1]
@@ -315,7 +315,7 @@ class DPDKdut(Dut):
                         toolchain, extra_options, options_config, default_library, target), "# ", build_time)
         assert ("FAILED" not in out), "meson setup failed ..."
 
-        out = self.send_expect("ninja -C %s -j %d" % (target, self.number_of_cores), "# ", build_time)
+        out = self.send_expect("ninja -C %s" % target, "# ", build_time)
         assert ("FAILED" not in out), "ninja complie failed ...\r\n %s" % out
 
         # copy kmod file to the folder same as make
-- 
2.27.0


             reply	other threads:[~2021-01-22 17:05 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-01-22 17:05 ohilyard [this message]
2021-02-02  5:27 ` Tu, Lijuan

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=20210122170529.672048-1-ohilyard@iol.unh.edu \
    --to=ohilyard@iol.unh.edu \
    --cc=dts@dpdk.org \
    --cc=lijuan.tu@intel.com \
    /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;
as well as URLs for NNTP newsgroup(s).