[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+55aFxZM9JeOeMkksZR93LojhvPZC38QnyvDEdazYg2SG2qHQ@mail.gmail.com>
Date: Mon, 2 Jul 2018 16:03:59 -0700
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: James Bottomley <James.Bottomley@...senpartnership.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>,
Waiman Long <longman@...hat.com>,
Al Viro <viro@...iv.linux.org.uk>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
linux-fsdevel <linux-fsdevel@...r.kernel.org>,
Jan Kara <jack@...e.cz>,
Paul McKenney <paulmck@...ux.vnet.ibm.com>,
Ingo Molnar <mingo@...nel.org>,
Miklos Szeredi <mszeredi@...hat.com>,
Matthew Wilcox <willy@...radead.org>,
Larry Woodman <lwoodman@...hat.com>,
"Wangkai (Kevin,C)" <wangkai86@...wei.com>,
linux-mm <linux-mm@...ck.org>, Michal Hocko <mhocko@...nel.org>
Subject: Re: [PATCH v5 0/6] fs/dcache: Track & limit # of negative dentries
On Mon, Jul 2, 2018 at 3:54 PM Linus Torvalds
<torvalds@...ux-foundation.org> wrote:
>
> Lookie here, for example:
>
> [torvalds@i7 linux]$ strace -e trace=%file -c git status
So in the name of honestly, that's slightly misleading.
"git" will happily thread the actual index file up-to-date testing.
And that's hidden in the above numbers (because I didn't use "-f" to
follow threads), and they are all successful (because git will go an
'lstat()' on every single entry in the index file, and the index file
obviously is all valid filenames).
So the numbers quoted are closer to the
git ls-files -o --exclude-standard
command (which doesn't check the index state, it only checks "what
non-tracked files do I have that aren't the ones I explicitly
exclude").
Linus
Powered by blists - more mailing lists