misc/147227: SSD preformance issues

[Garrett Cooper]

The following reply was made to PR misc/147227; it has been noted by GNATS.

From: Garrett Cooper <yanefbsd at gmail.com>
To: Brandon Falk <falkman at gamozo.org>
Cc: freebsd-gnats-submit at freebsd.org
Subject: Re: misc/147227: SSD preformance issues
Date: Sun, 30 May 2010 16:25:59 -0700

 On Sun, May 30, 2010 at 11:04 AM, Brandon Falk <falkman at gamozo.org> wrote:
 >
 >>Number: =A0 =A0 =A0 =A0 147227
 >>Category: =A0 =A0 =A0 misc
 >>Synopsis: =A0 =A0 =A0 SSD preformance issues
 >>Confidential: =A0 no
 >>Severity: =A0 =A0 =A0 serious
 >>Priority: =A0 =A0 =A0 medium
 >>Responsible: =A0 =A0freebsd-bugs
 >>State: =A0 =A0 =A0 =A0 =A0open
 >>Quarter:
 >>Keywords:
 >>Date-Required:
 >>Class: =A0 =A0 =A0 =A0 =A0sw-bug
 >>Submitter-Id: =A0 current-users
 >>Arrival-Date: =A0 Sun May 30 18:10:06 UTC 2010
 >>Closed-Date:
 >>Last-Modified:
 >>Originator: =A0 =A0 Brandon Falk
 >>Release: =A0 =A0 =A0 =A08.0-RELEASE
 >>Organization:
 > Gamozo
 >>Environment:
 > FreeBSD localhost 8.0-RELEASE FreeBSD 8.0-RELEASE #0: Sat Nov 21 15:02:08=
  UTC 2009 =A0 =A0 root at mason.cse.buffalo.edu:/usr/obj/usr/src/sys/GENERIC =
 =A0amd64
 >>Description:
 > I have an SSD that is rated at 285MB/s read, 275MB/s write, and 250MB/s s=
 ustained write. On Linux, I easily sustain all these speeds, but on FreeBSD=
  I happen to only get about 108 MB/s write, and 116MB/s read. Another thing=
  I noticed is that my drive is only running at SATA 150, even though my dri=
 ve is SATA 300 and my controller is also detected as SATA 300.
 >
 > *snip from dmesg*
 >
 > atapci1: <Intel ICH7 SATA300 controller> port 0xc400-0xc407,0xc080-0xc083=
 ,0xc000-0xc007,0xbc00-0xbc03,0xb880-0xb88f irq 19 at device 31.2 on pci0
 > ad4: 47703MB <OCZ VERTEX2 1.00> at ata2-master SATA150
 >
 > *snip from dmesg*
 >
 > I do not have an AHCI setting in my BIOS, so that is not something that I=
  can change, but still Linux happens to preform perfectly fine with the dri=
 ve, so it has to be something to do with FreeBSD.
 >>How-To-Repeat:
 > Always occurs.
 >>Fix:
 > Unknown.
 >
 >>Release-Note:
 >>Audit-Trail:
 >>Unformatted:
 
 This has been discussed at least once on the lists and in the forums.
 
 You might want to read through some really quickly google results I
 came up with: http://old.nabble.com/ATA-4K-sector-issues-td27929909i20.html
 , http://forums.freebsd.org/showthread.php?t=3D12594 .
 
 HTH,
 -Garrett