netdev - About improving the qlge Ethernet driver by following drivers/staging/qlge/TODO

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-ID: <20210504131421.mijffwcruql2fupn@Rk>
Date:   Tue, 4 May 2021 21:14:21 +0800
From:   Coiby Xu <coiby.xu@...il.com>
To:     Benjamin Poirier <benjamin.poirier@...il.com>
Cc:     linux-staging@...ts.linux.dev, netdev@...r.kernel.org
Subject: About improving the qlge Ethernet driver by following
 drivers/staging/qlge/TODO

Hi Benjamin,

As you have known, I'm working on improving drivers/staging/qlge. I'm
not sure if I correctly understand some TODO items. Since you wrote 
the TODO list, could you explain some of the items or comment on the
corresponding fix for me?

> * while in that area, using two 8k buffers to store one 9k frame is a poor
>   choice of buffer size.

Currently, LARGE_BUFFER_MAX_SIZE is defined as 8192. How about we simply
changing LARGE_BUFFER_MAX_SIZE to 4096? This is what 
drivers/net/ethernet/intel/e1000 does for jumbo frame right now.

> * in the "chain of large buffers" case, the driver uses an skb allocated with
>   head room but only puts data in the frags.

Do you suggest implementing the copybreak feature which exists for e1000 for 
this driver, i.e., allocing a sk_buff and coping the header buffer into it?

> * fix weird line wrapping (all over, ex. the ql_set_routing_reg() calls in
>   qlge_set_multicast_list()).

This issue of weird line wrapping is supposed to be all over. But I can
only find the ql_set_routing_reg() calls in qlge_set_multicast_list have
this problem,

			if (qlge_set_routing_reg
			    (qdev, RT_IDX_PROMISCUOUS_SLOT, RT_IDX_VALID, 1)) {

I can't find other places where functions calls put square and arguments
in the new line. Could you give more hints?

Thanks!

-- 
Best regards,
Coiby