[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <234bab6c-6d31-4c93-8a69-5b3687ba9b85@linux.ibm.com>
Date: Tue, 25 Nov 2025 16:58:32 +0530
From: Nilay Shroff <nilay@...ux.ibm.com>
To: "hch@....de" <hch@....de>
Cc: Chaitanya Kulkarni <chaitanyak@...dia.com>,
Christoph Hellwig <hch@...radead.org>,
Chaitanya Kulkarni <ckulkarnilinux@...il.com>,
"kbusch@...nel.org" <kbusch@...nel.org>, "hare@...e.de" <hare@...e.de>,
"sagi@...mberg.me" <sagi@...mberg.me>,
"axboe@...nel.dk" <axboe@...nel.dk>,
"dlemoal@...nel.org" <dlemoal@...nel.org>,
"wagi@...nel.org"
<wagi@...nel.org>,
"mpatocka@...hat.com" <mpatocka@...hat.com>,
"yukuai3@...wei.com" <yukuai3@...wei.com>,
"xni@...hat.com"
<xni@...hat.com>,
"linan122@...wei.com" <linan122@...wei.com>,
"bmarzins@...hat.com" <bmarzins@...hat.com>,
"john.g.garry@...cle.com" <john.g.garry@...cle.com>,
"edumazet@...gle.com" <edumazet@...gle.com>,
"ncardwell@...gle.com" <ncardwell@...gle.com>,
"kuniyu@...gle.com" <kuniyu@...gle.com>,
"davem@...emloft.net" <davem@...emloft.net>,
"dsahern@...nel.org" <dsahern@...nel.org>,
"kuba@...nel.org"
<kuba@...nel.org>,
"pabeni@...hat.com" <pabeni@...hat.com>,
"horms@...nel.org" <horms@...nel.org>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"linux-nvme@...ts.infradead.org" <linux-nvme@...ts.infradead.org>,
"linux-block@...r.kernel.org" <linux-block@...r.kernel.org>
Subject: Re: [RFC blktests fix PATCH] tcp: use GFP_ATOMIC in tcp_disconnect
On 11/25/25 4:51 PM, hch@....de wrote:
> On Tue, Nov 25, 2025 at 04:43:25PM +0530, Nilay Shroff wrote:
>> The memalloc_noreclaim_save() above shall already prevent filesystem
>> reclaim,
>
> memalloc_noreclaim_save is oddly misnamed, as it sets the
> PF_MEMALLOC, which does not cause any gfp_t flag adjustments, but
> instead avoid direct reclaim. Thinking of it I have no idea why
> it is even used here.
>
>From git history, I see that was added to avoid memory reclaim to avoid
possible circular locking dependency. This commit 83e1226b0ee2 ("nvme-tcp:
fix possible circular locking when deleting a controller under memory
pressure") adds it.
Thanks,
--Nilay
Powered by blists - more mailing lists