From: "Brandes, Shai" <shaibran@amazon.com>
To: Stephen Hemminger <stephen@networkplumber.org>
Cc: "ferruh.yigit@amd.com" <ferruh.yigit@amd.com>,
"dev@dpdk.org" <dev@dpdk.org>, Michal Krawczyk <mk@semihalf.com>,
"Schmeilin, Evgeny" <evgenys@amazon.com>,
"Chauskin, Igor" <igorch@amazon.com>,
"Beider, Ron" <rbeider@amazon.com>,
"Brandes, Shai" <shaibran@amazon.com>,
"Atrash, Wajeeh" <atrwajee@amazon.com>
Subject: RE: [PATCH 4/5] net/ena: add support for ena-express metrics
Date: Tue, 24 Oct 2023 15:55:21 +0000 [thread overview]
Message-ID: <3e4c84831b074d92897576fab8894865@amazon.com> (raw)
In-Reply-To: <20231024082639.1dbb7638@hermes.local>
On Tue, 24 Oct 2023 13:21:27 +0300
<shaibran@amazon.com> wrote:
>> struct ena_offloads {
>> uint32_t tx_offloads;
>> uint32_t rx_offloads;
>> @@ -329,6 +346,7 @@ struct ena_adapter {
>> */
>> uint64_t metrics_stats[ENA_MAX_CUSTOMER_METRICS] __rte_cache_aligned;
>> uint16_t metrics_num;
>> + struct ena_stats_srd srd_stats __rte_cache_aligned;
>> };
> If metrics_num was before the metrics_stats[] you would save some space.
Thank you for the review! My reasoning was that the metrics_num should reside in the same cache line as the metrics_stats (metrics_stats can contain up to 6 metrics so its max length is 48B).
In addition, the struct ena_stats_srd was added in a later patch, so I will run pahole and verify which order of these three fields gives the best results.
BTW, we are starting a cache optimization task now that should address all these issues, this is indeed one of our main structures, so it will definitely get modified later.
next prev parent reply other threads:[~2023-10-24 15:55 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-24 10:21 [PATCH 0/5] net/ena: v2.8.0 driver release shaibran
2023-10-24 10:21 ` [PATCH 1/5] net/ena: hal upgrade shaibran
2023-10-24 10:21 ` [PATCH 2/5] net/ena: add support for connection tracking metric shaibran
2023-10-24 10:21 ` [PATCH 3/5] net/ena: report Rx overrun errors in xstats shaibran
2023-10-24 10:21 ` [PATCH 4/5] net/ena: add support for ena-express metrics shaibran
2023-10-24 15:26 ` Stephen Hemminger
2023-10-24 15:55 ` Brandes, Shai [this message]
2023-10-25 12:34 ` Brandes, Shai
2023-10-24 10:21 ` [PATCH 5/5] net/ena: update ena version to 2.8.0 shaibran
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=3e4c84831b074d92897576fab8894865@amazon.com \
--to=shaibran@amazon.com \
--cc=atrwajee@amazon.com \
--cc=dev@dpdk.org \
--cc=evgenys@amazon.com \
--cc=ferruh.yigit@amd.com \
--cc=igorch@amazon.com \
--cc=mk@semihalf.com \
--cc=rbeider@amazon.com \
--cc=stephen@networkplumber.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;
as well as URLs for NNTP newsgroup(s).