linux-kernel - Re: [PATCH] nvme: Check for readiness more quickly, to speed up boot time

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

Message-ID: <20200301090239.GC216567@localhost>
Date:   Sun, 1 Mar 2020 01:02:39 -0800
From:   Josh Triplett <josh@...htriplett.org>
To:     Chaitanya Kulkarni <Chaitanya.Kulkarni@....com>
Cc:     Keith Busch <kbusch@...nel.org>, Jens Axboe <axboe@...com>,
        Christoph Hellwig <hch@....de>,
        Sagi Grimberg <sagi@...mberg.me>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-nvme@...ts.infradead.org" <linux-nvme@...ts.infradead.org>
Subject: Re: [PATCH] nvme: Check for readiness more quickly, to speed up boot
 time

On Sun, Mar 01, 2020 at 02:01:05AM +0000, Chaitanya Kulkarni wrote:
> Nit:- please have a look at the patch subject line and make
> sure it is not exceeding the required length.

Documentation/process/submitting-patches.rst says "no more than 70-75
characters,", and the summary here is 61. Checkpatch similarly says 75.
Is there somewhere I missed that gives a different number?

> One question though, have you seen similar kind of performance 
> improvements when system is booted ?

I tested with nvme compiled in, both with one NVMe device and two NVMe
devices, and in both cases it provided a *substantial* speedup. I didn't
test nvme compiled as a module, but in general I'd expect that if you're
trying to optimize initialization time you'd want to build it in.

> I took some numbers and couldn't see similar benefit. See [1] :-
> 
> Without :-
> 
> 714.532560-714.456099 = .076461
> 721.189886-721.110845 = .079041
> 727.836938-727.765572 = .071366
> 734.589886-734.519779 = .070107
> 741.244296-741.173503 = .070793

With numbers in this range, I don't see how you could be hitting the
100ms msleep at all, even once, which means this patch shouldn't have
any effect on the timing you're measuring.

- Josh Triplett