[PATCH] D38295: Do not use posix_fallocate because it's slow.

[Rui Ueyama via llvm-commits]

On Thu, Sep 28, 2017 at 8:35 PM, Rui Ueyama <ruiu at google.com> wrote:

> On Thu, Sep 28, 2017 at 8:26 PM, Sean Silva <chisophugis at gmail.com> wrote:
>
>>
>>
>> On Sep 28, 2017 8:45 PM, "Rui Ueyama via Phabricator via llvm-commits" <
>> llvm-commits at lists.llvm.org> wrote:
>>
>> ruiu abandoned this revision.
>> ruiu added a comment.
>>
>> Thank you for reviewing. But I think my test environment was not set up
>> correctly. While it is an ext4, fallocate(2) returned an error. After
>> mkfs'ing the partition, fallocate(2) started working, and the original code
>> worked as fast as with this patch. So it seems there's no point to submit
>> this patch. I'll withdraw it.
>>
>>
>> Does disk encryption affect it?
>>
>
> I can't verify as I've already reinitialized the filesystem, but I doubt
> it, because that was a new SSD drive, and I formatted it myself.
>

I think I found what was wrong. If you create a new filesystem using mke2fs
(which creates an ext2 filesystem), you can still mount it as ext4, but you
can't call fallocate(2) on it. Looks like you are supposed to use mkfs.ext4
instead.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170929/cb64f3ff/attachment.html>