Re: [PATCH 6/6] MIPS: net: BPF: Introduce BPF ASM helpers
- To: Markos Chandras <markos.chandras@imgtec.com>
- Cc: linux-mips@linux-mips.org, netdev@vger.kernel.org, "David S. Miller" <davem@davemloft.net>, Alexei Starovoitov <ast@plumgrid.com>, Daniel Borkmann <dborkman@redhat.com>, Hannes Frederic Sowa <hannes@stressinduktion.org>, linux-kernel@vger.kernel.org, debian-kernel@lists.debian.org
- Subject: Re: [PATCH 6/6] MIPS: net: BPF: Introduce BPF ASM helpers
- From: Aurelien Jarno <aurelien@aurel32.net>
- Date: Thu, 13 Aug 2015 22:42:46 +0200
- Message-id: <[🔎] 20150813204246.GA24857@aurel32.net>
- Mail-followup-to: Markos Chandras <markos.chandras@imgtec.com>, linux-mips@linux-mips.org, netdev@vger.kernel.org, "David S. Miller" <davem@davemloft.net>, Alexei Starovoitov <ast@plumgrid.com>, Daniel Borkmann <dborkman@redhat.com>, Hannes Frederic Sowa <hannes@stressinduktion.org>, linux-kernel@vger.kernel.org, debian-kernel@lists.debian.org
- In-reply-to: <1433415376-20952-7-git-send-email-markos.chandras@imgtec.com>
- References: <1433415376-20952-1-git-send-email-markos.chandras@imgtec.com> <1433415376-20952-7-git-send-email-markos.chandras@imgtec.com>
On 2015-06-04 11:56, Markos Chandras wrote:
> This commit introduces BPF ASM helpers for MIPS and MIPS64 kernels.
> The purpose of this patch is to twofold:
>
> 1) We are now able to handle negative offsets instead of either
> falling back to the interpreter or to simply not do anything and
> bail out.
>
> 2) Optimize reads from the packet header instead of calling the C
> helpers
>
> Because of this patch, we are now able to get rid of quite a bit of
> code in the JIT generation process by using MIPS optimized assembly
> code. The new assembly code makes the test_bpf testsuite happy with
> all 60 test passing successfully compared to the previous
> implementation where 2 tests were failing.
> Doing some basic analysis in the results between the old
> implementation and the new one we can obtain the following
> summary running current mainline on an ER8 board (+/- 30us delta is
> ignored to prevent noise from kernel scheduling or IRQ latencies):
>
> Summary: 22 tests are faster, 7 are slower and 47 saw no improvement
>
> with the most notable improvement being the tcpdump tests. The 7 tests
> that seem to be a bit slower is because they all follow the slow path
> (bpf_internal_load_pointer_neg_helper) which is meant to be slow so
> that's not a problem.
>
> Cc: netdev@vger.kernel.org
> Cc: "David S. Miller" <davem@davemloft.net>
> Cc: Alexei Starovoitov <ast@plumgrid.com>
> Cc: Daniel Borkmann <dborkman@redhat.com>
> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org>
> Cc: linux-kernel@vger.kernel.org
> Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
> ---
> I have uploaded the script and the bpf result files in my LMO webspace
> in case you want to have a look. I didn't paste them in here because they
> are nearly 200 lines. Simply download all 3 files and run './bpf_analysis.py'
This patch relies on R2 instructions, and thus the Linux kernel fails to
build when targetting non-R2 CPUs. See for example:
https://buildd.debian.org/status/fetch.php?pkg=linux&arch=mipsel&ver=4.2%7Erc6-1%7Eexp1&stamp=1439480000
--
Aurelien Jarno GPG: 4096R/1DDD8C9B
aurelien@aurel32.net http://www.aurel32.net
Reply to: