netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Alexander Lobakin <aleksander.lobakin@intel.com>
To: Florian Fainelli <florian.fainelli@broadcom.com>
Cc: <netdev@vger.kernel.org>, <kuba@kernel.org>, <jgg@nvidia.com>,
	<leonro@nvidia.com>, <horms@kernel.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Tal Gilboa <talgi@nvidia.com>,
	"open list:LIBRARY CODE" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH net-next v2] lib: Allow for the DIM library to be modular
Date: Wed, 8 May 2024 11:17:08 +0200	[thread overview]
Message-ID: <38b22230-01f8-4334-a8c1-dd3dde42fe40@intel.com> (raw)
In-Reply-To: <52bfe069-bff1-4f8e-baed-1c556e43a242@broadcom.com>

From: Florian Fainelli <florian.fainelli@broadcom.com>
Date: Tue, 7 May 2024 10:08:19 -0700

> On 5/7/24 06:13, Alexander Lobakin wrote:
>> From: Florian Fainelli <florian.fainelli@broadcom.com>
>> Date: Mon,  6 May 2024 10:50:40 -0700
>>
>>> Allow the Dynamic Interrupt Moderation (DIM) library to be built as a
>>> module. This is particularly useful in an Android GKI (Google Kernel
>>> Image) configuration where everything is built as a module, including
>>> Ethernet controller drivers. Having to build DIMLIB into the kernel
>>> image with potentially no user is wasteful.
>>
>> Some bloat-o-meter -c vmlinux.{before,after} would be good to have here.
>> The library is small, but I personally would like to see it modular.
> 
> ./scripts/bloat-o-meter vmlinux vmlinux.after
> add/remove: 1/16 grow/shrink: 0/0 up/down: 8/-1980 (-1972)
> Function                                     old     new   delta
> e843419@0740_00005048_4014                     -       8      +8
> e843419@07b3_000055c6_5944                     8       -      -8
> dim_park_on_top                                8       -      -8
> dim_park_tired                                16       -     -16
> net_dim_get_def_tx_moderation                 24       -     -24
> net_dim_get_def_rx_moderation                 24       -     -24
> dim_turn                                      52       -     -52
> net_dim_get_rx_moderation                     60       -     -60
> net_dim_get_tx_moderation                     64       -     -64
> tx_profile                                    80       -     -80
> rx_profile                                    80       -     -80
> dim_on_top                                    84       -     -84
> net_dim_step                                 132       -    -132
> dim_calc_stats                               172       -    -172
> net_dim_stats_compare                        176       -    -176
> net_dim                                      464       -    -464
> rdma_dim                                     536       -    -536
> Total: Before=12668884, After=12666912, chg -0.02%
> 
> This is on arm64 FWIW.

-2 Kb sounds reasonable I'd say. This doesn't even require adding new
exports, nice.

> 
>>
>>>
>>> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
>>
>> Reviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com>
>>
>>> ---
>>> Changes in v2:
>>>
>>> - Added MODULE_DESCRIPTION()
>>>
>>>   lib/Kconfig      | 2 +-
>>>   lib/dim/Makefile | 4 ++--
>>>   lib/dim/dim.c    | 3 +++
>>>   3 files changed, 6 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/lib/Kconfig b/lib/Kconfig
>>> index 4557bb8a5256..d33a268bc256 100644
>>> --- a/lib/Kconfig
>>> +++ b/lib/Kconfig
>>> @@ -628,7 +628,7 @@ config SIGNATURE
>>>         Implementation is done using GnuPG MPI library
>>>     config DIMLIB
>>> -    bool
>>> +    tristate
>>>       help
>>>         Dynamic Interrupt Moderation library.
>>>         Implements an algorithm for dynamically changing CQ
>>> moderation values
>>> diff --git a/lib/dim/Makefile b/lib/dim/Makefile
>>> index 1d6858a108cb..c4cc4026c451 100644
>>> --- a/lib/dim/Makefile
>>> +++ b/lib/dim/Makefile
>>> @@ -2,6 +2,6 @@
>>>   # DIM Dynamic Interrupt Moderation library
>>>   #
>>>   -obj-$(CONFIG_DIMLIB) += dim.o
>>> +obj-$(CONFIG_DIMLIB) += dimlib.o
>>
>> I guess you renamed it due to that there's already a module named 'dim'?
> 
> This is required to avoid the follow recursive dependency:
> 
> make[4]: Circular lib/dim/dim.o <- lib/dim/dim.o dependency dropped.

Aaah I see!

Thanks,
Olek

  reply	other threads:[~2024-05-08  9:17 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-06 17:50 [PATCH net-next v2] lib: Allow for the DIM library to be modular Florian Fainelli
2024-05-07 13:13 ` Alexander Lobakin
2024-05-07 17:08   ` Florian Fainelli
2024-05-08  9:17     ` Alexander Lobakin [this message]
2024-05-07 23:50 ` patchwork-bot+netdevbpf

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=38b22230-01f8-4334-a8c1-dd3dde42fe40@intel.com \
    --to=aleksander.lobakin@intel.com \
    --cc=akpm@linux-foundation.org \
    --cc=florian.fainelli@broadcom.com \
    --cc=horms@kernel.org \
    --cc=jgg@nvidia.com \
    --cc=kuba@kernel.org \
    --cc=leonro@nvidia.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=talgi@nvidia.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).