[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200126123014.0983991d@cakuba>
Date: Sun, 26 Jan 2020 12:30:14 -0800
From: Jakub Kicinski <kuba@...nel.org>
To: Michal Kalderon <mkalderon@...vell.com>
Cc: Ariel Elior <aelior@...vell.com>,
"davem@...emloft.net" <davem@...emloft.net>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"linux-rdma@...r.kernel.org" <linux-rdma@...r.kernel.org>,
"linux-scsi@...r.kernel.org" <linux-scsi@...r.kernel.org>,
David Miller <davem@...emloft.net>
Subject: Re: [PATCH v2 net-next 00/13] qed*: Utilize FW 8.42.2.0
On Sun, 26 Jan 2020 11:29:50 +0000, Michal Kalderon wrote:
> > On Thu, 23 Jan 2020 12:58:23 +0200, Michal Kalderon wrote:
> > > Changes from V1
> > > ---------------
> > > - Remove epoch + kernel version from device debug dump
> > > - don't bump driver version
> >
> > But you haven't fixed the fact that in patch 1 you already strat changing
> > defines for the new FW version, even though the version is only enforced
> > (reportedly) in patch 9?
>
> Right, I'll move the version change to patch #1 in V3.
>
> However, the entire series is required (except a few patches not prefixed with FW 8.42.2.0 ) to be
> taken to work correctly with the FW.
Right, exactly.
> Our FW is not backward/forward compatible.
Well, the driver can also be backward/forward compatible
(even if only for the short period of a series of patches)
as FW is usually more resource constrained.
> I have mentioned this in the cover letter, the split into smaller patches and prefix with
> FW 8.42.2.0 is to ease review and was done due to previous feedback that it is very difficult to review the FW patches:
>
> https://www.spinics.net/lists/linux-rdma/msg58810.html
>
> I am fine with squashing all the patches that are required to working with FW8.42.2.0 into one single patch if that is required and acceptable,
> But I believe that would make reviewing the changes more difficult.
The choice is not either one giant unreviewable patch or a driver
broken between commits. There are tens of Ethernet drivers in tree
and none of them seems to be having this issue.
Maybe Dave will let it fly this time around but you need to take
a hard look at your process for the future.
Powered by blists - more mailing lists