linux-kernel - Re: [PATCH 1/2] fs: Do not dispatch FITRIM through separate super

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: <4CE5F21F.7010806@teksavvy.com>
Date:	Thu, 18 Nov 2010 22:42:23 -0500
From:	Mark Lord <kernel@...savvy.com>
To:	"Martin K. Petersen" <martin.petersen@...cle.com>
CC:	Greg Freemyer <greg.freemyer@...il.com>,
	James Bottomley <James.Bottomley@...e.de>,
	Jeff Moyer <jmoyer@...hat.com>,
	Christoph Hellwig <hch@...radead.org>,
	Matthew Wilcox <matthew@....cx>,
	Josef Bacik <josef@...hat.com>,
	Lukas Czerner <lczerner@...hat.com>, tytso@....edu,
	linux-ext4@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-fsdevel@...r.kernel.org, sandeen@...hat.com
Subject: Re: [PATCH 1/2] fs: Do not dispatch FITRIM through separate super_operation

On 10-11-18 08:49 PM, Martin K. Petersen wrote:
> So assuming we walk the filesystem to reclaim space on ATA SSDs on a
> weekly basis (since that's the only sane approach):
>
>         What is the performance impact of not coalescing discontiguous
>         block ranges when cron scrubs your /home at 4am Sunday morning?

In the case of FITRIM, you're right: the performance impact
probably doesn't matter much for a 4am cronjob.

But a lot of people currently (at least) prefer to run it manually,
and they don't want it to take forever.

Though that's still not the primary worry:  each TRIM seems to trigger
a flash erase cycle (or cycles) on the most common SSDs on the market,
including anything Indilinx-based and as far as I can tell also for
anything SandForce based.  That's probably 70% of the SSDs out there today.

And I'm very concerned about premature wear -- MLC is only for 10000 cycles (avg).

Also, the current one-range-at-a-time interface is just not correct
for the majority of devices out there:  they are SATA, not some obscure
enterprise-only one-range-at-a-time thing.  We need an implementation
that reflects real-life for uses other than data centres.

If nobody else does it, I'll probably implement it correctly this winter.
But it would really be better for a real filesystem/DM person to do it.

Thanks for hanging in there this far, though!

Cheers
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/