linux ipv4 forward performance issue

[Liu Ting]

Hello all,

When upgrading meta-linaro/meta-linaro-toolchain from daisy to fido, we faced drastic performance downgrade (-34.06%) on linux ipv4 forwarding. The board is ls1021atwr (ARM Cortex-A7 MPCore compliant with ARMv7-A architecture).

We did some investigation, and suspected that it is caused by the toolchain upgrade. Did someone meet similar issue?

The detailed toolchain info:
https://git.linaro.org/openembedded/meta-linaro.git/shortlog/refs/heads/fido

gcc-linaro-4.9.3-2015.03, glibc-linaro-2.20-2014.11, binutils-linaro-2.25-2015.01

https://git.linaro.org/openembedded/meta-linaro.git/shortlog/refs/heads/daisy
gcc-linaro-4.8.3-2014.04, eglibc-linaro-2.19-r2014.04, binutils-linaro-2.24-r2014.03

thanks.

-Ting

[Jim Wilson]

Post by Liu Ting
When upgrading meta-linaro/meta-linaro-toolchain from daisy to fido, we
faced drastic performance downgrade (-34.06%) on linux ipv4 forwarding. The
board is ls1021atwr (ARM Cortex-A7 MPCore compliant with ARMv7-A
architecture).
We did some investigation, and suspected that it is caused by the toolchain
upgrade. Did someone meet similar issue?

I haven't heard of any such issue, but that doesn't mean much, as I
don't read kernel mailing lists, and don't have full knowledge of all
gcc bugs that have been privately and publicly reported. Basically,
without a testcase, that we can use to reproduce the problem, you are
unlikely to get much useful info from a toolchain developer. And if
the testcase requires building a linux system from scratch, then that
may be too much trouble for us to follow up on. You might try asking
some linaro kernel developers about this if you haven't already. They
would know more about kernel performance issues.

Jim

[Adhemerval Zanella]

Post by Jim Wilson

Post by Liu Ting
When upgrading meta-linaro/meta-linaro-toolchain from daisy to fido, we
faced drastic performance downgrade (-34.06%) on linux ipv4 forwarding. The
board is ls1021atwr (ARM Cortex-A7 MPCore compliant with ARMv7-A
architecture).
We did some investigation, and suspected that it is caused by the toolchain
upgrade. Did someone meet similar issue?

I haven't heard of any such issue, but that doesn't mean much, as I
don't read kernel mailing lists, and don't have full knowledge of all
gcc bugs that have been privately and publicly reported. Basically,
without a testcase, that we can use to reproduce the problem, you are
unlikely to get much useful info from a toolchain developer. And if
the testcase requires building a linux system from scratch, then that
may be too much trouble for us to follow up on. You might try asking
some linaro kernel developers about this if you haven't already. They
would know more about kernel performance issues.

Also if you could provide any annotate profile which outline some code
generation issue it will help narrow down if this is really a toolchain
issue.

Post by Jim Wilson
Jim
_______________________________________________
linaro-toolchain mailing list
https://lists.linaro.org/mailman/listinfo/linaro-toolchain

[Liu Ting]

Hello all,

Thanks for your help.

The root cause was found. It had nothing to do with linaro toolchain.
A backend process is abnormal and consumes a lot of CPU.

Sorry for the noise.

-Ting

-----Original Message-----
Behalf Of Adhemerval Zanella
Sent: Thursday, October 15, 2015 8:01 PM
Subject: Re: linux ipv4 forward performance issue

Post by Jim Wilson

Post by Liu Ting
When upgrading meta-linaro/meta-linaro-toolchain from daisy to fido,
we faced drastic performance downgrade (-34.06%) on linux ipv4
forwarding. The board is ls1021atwr (ARM Cortex-A7 MPCore compliant
with ARMv7-A architecture).
We did some investigation, and suspected that it is caused by the
toolchain upgrade. Did someone meet similar issue?

I haven't heard of any such issue, but that doesn't mean much, as I
don't read kernel mailing lists, and don't have full knowledge of all
gcc bugs that have been privately and publicly reported. Basically,
without a testcase, that we can use to reproduce the problem, you are
unlikely to get much useful info from a toolchain developer. And if
the testcase requires building a linux system from scratch, then that
may be too much trouble for us to follow up on. You might try asking
some linaro kernel developers about this if you haven't already. They
would know more about kernel performance issues.

Also if you could provide any annotate profile which outline some code
generation issue it will help narrow down if this is really a toolchain issue.

Post by Jim Wilson
Jim
_______________________________________________
linaro-toolchain mailing list
https://lists.linaro.org/mailman/listinfo/linaro-toolchain

_______________________________________________
linaro-toolchain mailing list
https://lists.linaro.org/mailman/listinfo/linaro-toolchain