Bug 173468 - [ata] lost WDC hard drive on ICH5 after upgrade [regression]
Summary: [ata] lost WDC hard drive on ICH5 after upgrade [regression]
Status: Open
Alias: None
Product: Base System
Classification: Unclassified
Component: kern (show other bugs)
Version: 8.3-STABLE
Hardware: Any Any
: Normal Affects Only Me
Assignee: freebsd-bugs (Nobody)
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-11-08 08:40 UTC by Alexander M. Pravkin
Modified: 2017-12-31 22:29 UTC (History)
0 users

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Alexander M. Pravkin 2012-11-08 08:40:00 UTC
After an upgrade from 8.2-STABLE Sat Feb 26 14:47:49 MSK 2011, kernel failed to locate one of my hard disks (ad6) on an Intel ICH5 controller. Other attached disks are fine.

diff between `egrep 'atapci1|ata3|ad6'` on dmesg's before and after upgrade:
--- dmesg-verbose-8.2-ata3.log  2012-11-07 11:33:43.000000000 +0400
+++ dmesg-verbose-8.3-ata3.log  2012-11-07 11:33:50.000000000 +0400
@@ -2,28 +2,20 @@ atapci1: <Intel ICH5 SATA150 controller>
 atapci1: Reserved 0x10 bytes for rid 0x20 type 4 at 0xdc00
 atapci1: [MPSAFE]
 atapci1: [ITHREAD]
-ata2: <ATA channel 0> on atapci1
+ata2: <ATA channel> at channel 0 on atapci1
 atapci1: Reserved 0x8 bytes for rid 0x10 type 4 at 0xec00
 atapci1: Reserved 0x4 bytes for rid 0x14 type 4 at 0xe800
-ata3: <ATA channel 1> on atapci1
+ata3: <ATA channel> at channel 1 on atapci1
 atapci1: Reserved 0x8 bytes for rid 0x18 type 4 at 0xe400
 atapci1: Reserved 0x4 bytes for rid 0x1c type 4 at 0xe000
+ata3: SATA reset: ports status=0x02
 ata3: p0: SATA connect time=0ms status=00000113
 ata3: reset tp1 mask=01 ostat0=80 ostat1=00
 ata3: stat0=0x50 err=0x01 lsb=0x3c msb=0xc3
-ata3: reset tp2 stat0=50 stat1=00 devices=0x1
+ata3: reset tp2 stat0=50 stat1=00 devices=0x0
 ata3: [MPSAFE]
 ata3: [ITHREAD]
-ata3: Identifying devices: 00000001
-ata3: New devices: 00000001
-ata3-master: pio=PIO4 wdma=WDMA2 udma=UDMA133 cable=40 wire
-ad6: setting UDMA100
-ad6: 238475MB <WDC WD2500AAJS-62B4A0 03.03A01> at ata3-master UDMA100 SATA 1.5Gb/s
-ad6: 488397168 sectors [484521C/16H/63S] 16 sectors/interrupt 1 depth queue
-GEOM: new disk ad6
-ad6: Intel check1 failed
-ad6: Adaptec check1 failed
-ad6: LSI (v3) check1 failed
-ad6: LSI (v2) check1 failed
-ad6: FreeBSD check1 failed
+ata3: Identifying devices: 00000000
+ata3: New devices: 00000000
 Trying to mount root from ufs:/dev/ad6s1a
+vfs.root.mountfrom=ufs:/dev/ad6s1a

Probably it's the same problem as in http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/159511

Fix: 

Looks like problem appeared after this commit:
http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/ata/ata-lowlevel.c.diff?r1=1.84.2.8;r2=1.84.2.9

After rolling src/sys/dev/ata back to 2011-05-25 and rebuilding kernel, ad6 is fine again.
Comment 1 Xin LI freebsd_committer freebsd_triage 2012-11-08 23:21:57 UTC
Responsible Changed
From-To: freebsd-bugs->mav

Over to maintainer.
Comment 2 Alexander Motin freebsd_committer freebsd_triage 2012-11-26 20:50:16 UTC
Hi.

Your guess about commit causing this looks realistic. But I don't see 
problem in it. As I understand, the problem is either with the device or 
the controller reporting strange values instead of device signature. 
Respecting that this commit has its purpose and this is the only issue 
for almost a year I am not sure we should just revert it.

Could you try to boot from FreeBSD 9.1-RC3 install image with enabled 
verbose messages to see whether disk will be detected there and how?

-- 
Alexander Motin
Comment 3 Eitan Adler freebsd_committer freebsd_triage 2017-12-31 08:00:18 UTC
For bugs matching the following criteria:

Status: In Progress Changed: (is less than) 2014-06-01

Reset to default assignee and clear in-progress tags.

Mail being skipped