[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <c4011962-013d-4a7d-8c6a-40afede101e8@wdc.com>
Date: Fri, 4 Aug 2023 10:25:50 +0000
From: Johannes Thumshirn <Johannes.Thumshirn@....com>
To: "Andreas Hindborg (Samsung)" <nmi@...aspace.dk>
CC: Ming Lei <ming.lei@...hat.com>,
"hch@...radead.org" <hch@...radead.org>,
"gost.dev@...sung.com" <gost.dev@...sung.com>,
"open list:BLOCK LAYER" <linux-block@...r.kernel.org>,
Hans Holmberg <Hans.Holmberg@....com>,
Matias Bjørling <Matias.Bjorling@....com>,
Minwoo Im <minwoo.im.dev@...il.com>,
Jens Axboe <axboe@...nel.dk>,
Johannes Thumshirn <jth@...nel.org>,
Aravind Ramesh <Aravind.Ramesh@....com>,
Damien Le Moal <dlemoal@...nel.org>,
open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v10 3/3] ublk: enable zoned storage support
On 04.08.23 12:12, Andreas Hindborg (Samsung) wrote:
>
> Johannes Thumshirn <Johannes.Thumshirn@....com> writes:
>
>> On 03.08.23 16:09, Andreas Hindborg (Samsung) wrote:
>>> + buf = __vmalloc(bufsize, GFP_KERNEL | __GFP_NORETRY);
>>
>> Missing #include <linux/vmalloc.h> so the bot doesn't complain.
>
> Thanks, gotta add that so code compiles for Sega Dreamcast!
>
>> But while we're at it, why can't you just use kvmalloc() here?
>
> I don't see why not. It should be better for small reports I guess. I will change it.
Yep, that way you won't need vmalloc.h either. The double underscore
versions of a function are always a red flag when reviewing code to me.
Powered by blists - more mailing lists