Skip site navigation (1)Skip section navigation (2)
Date:      10 Nov 2007 02:18:47 +0100
From:      "Arno J. Klaassen" <arno@heho.snv.jussieu.fr>
To:        delphij@delphij.net
Cc:        Nathan Butcher <n-butcher@fusiongol.com>, Alexander Sabourenkov <screwdriver@lxnt.info>, Soren Schmidt <sos@deepcore.dk>, freebsd-current@freebsd.org
Subject:   Re: Remaining SATA (and other) issues
Message-ID:  <wpy7d6ew3s.fsf@heho.snv.jussieu.fr>
In-Reply-To: <4734E837.9000804@delphij.net>
References:  <47326FB8.50602@fusiongol.com> <4732CEE3.3070003@lxnt.info> <4732DA32.3090601@deepcore.dk> <4732E18A.6040802@lxnt.info> <47336FE1.6010109@deepcore.dk> <wphcjvf47e.fsf@heho.snv.jussieu.fr> <4734E16B.3000909@delphij.net> <wpsl3ff2lk.fsf@heho.snv.jussieu.fr> <4734E837.9000804@delphij.net>

next in thread | previous in thread | raw e-mail | index | archive | help

Hello,

I tried again with kernel sources (releng_6) without
latest MFC but with patch applied (minus the marvel_setprd-chunk).

No difference, but ... something (else???) is really weird :

1) # dd if=/dev/zero of=/dev/ad6 count=2 ; sleep 1; ls -l /dev/ad6*
     2+0 records in
     2+0 records out
     1024 bytes transferred in 0.000486 secs (2107442 bytes/sec)
     crw-r-----  1 root  operator    0,  68 Nov 10 01:21 /dev/ad6


2) # fdisk -BI /dev/ad6; sleep 1; ls -l /dev/ad6s1*
   ******* Working on device /dev/ad6 *******
   fdisk: invalid fdisk partition table found
   fdisk: Geom not found
   crw-r-----  1 root  operator    0,  94 Nov 10 00:43 /dev/ad6s1

   Note : no /dev/ad6s1c showing up ... 


3) # disklabel -re /dev/ad6s1

     # /dev/ad6s1:
     8 partitions:
     #        size   offset    fstype   [fsize bsize bps/cpg]
       c: 976773105       63    unused        0     0         # "raw" part, don't edit
       d: 976773105       63    4.2BSD        0     0     0

       partition c: partition extends past end of unit
     disklabel: partition c doesn't start at 0!
     disklabel: An incorrect partition c may cause problems for standard system utilities
     partition d: partition extends past end of unit
     re-edit the label? [y]: 

   (hard to reproduce, but after several repetitions, it sometimes
    shows "0" for the offsets and I can write the label without
    complaint)


4a) I edit the label, set offset to "0" by hand, and then :

    # ls -l /dev/ad6*
    crw-r-----  1 root  operator    0,  68 Nov 10 01:21 /dev/ad6
    crw-r-----  1 root  operator    0,  94 Nov 10 01:33 /dev/ad6s1

4b) Idem, but I reduce size of partition 'd' by "63" :

    # ls -l /dev/ad6*
    crw-r-----  1 root  operator    0,  68 Nov 10 01:21 /dev/ad6
    crw-r-----  1 root  operator    0,  94 Nov 10 01:33 /dev/ad6s1

   No way /dev/ad6s1c and /dev/ad6s1d will show up (though they
   do sometimes show up for a short while ... hard to repeat )

4c) Take a beer from the fridge and change radio station :

   # ls -l /dev/ad6*          
   crw-r-----  1 root  operator    0,  68 Nov 10 01:44 /dev/ad6

   Even /dev/ad6s1 quietly disappeared ....


5) Bon, I repeat the same things for ad12 :


   # fdisk -BI /dev/ad12; sleep 1; ls -l /dev/ad12s1*
   ******* Working on device /dev/ad12 *******
   fdisk: Geom not found
   crw-r-----  1 root  operator    0,  76 Nov 10 00:43 /dev/ad12s1
   crw-r-----  1 root  operator    0,  77 Nov 10 00:43 /dev/ad12s1c
   [root@charlotte ~]# disklabel -re /dev/ad12s1:
   8 partitions:
   #        size   offset    fstype   [fsize bsize bps/cpg]
     a: 625142432       16    unused        0     0
     c: 625142385        0    unused        0     0         # "raw" part, don't edit


  (Note, 'a' is not correct here, but that might be 'normal' due
   to my very limited 'geom'-knowledge (it does not show up
   in 'disklabel /dev/ad12s1' anyway))
  I add a partition 'd' equal to 'c' (and remove'a').

  # ls -l /dev/ad12*  
  crw-r-----  1 root  operator    0,  69 Nov 10 01:44 /dev/ad12
  crw-r-----  1 root  operator    0,  76 Nov 10 01:48 /dev/ad12s1
  crw-r-----  1 root  operator    0,  94 Nov 10 00:43 /dev/ad12s1c
  crw-r-----  1 root  operator    0,  95 Nov 10 00:43 /dev/ad12s1d

  (Note, weird atimes for /dev/ad12s1c and /dev/ad12s1d ..., I just
   created them ...  enfin)

  # newfs /dev/ad12s1d
  # mount /dev/ad12s1d /mnt
  # cd /mnt/
  # /usr/temp/test_raid.sh 10 20  (creates 20 random files of 10M,
                                   copies them and does md5-cksum
                                   on original and copy)
  # /usr/temp/test_raid.sh 100 4
  # /usr/temp/test_raid.sh 1000 2
  # /usr/temp/test_raid.sh 4000 1


  No crash, all cksum-comparisons OK ...


To finish, the hardware differences between ad6 and ad12 :

# atacontrol cap ad6 

  Protocol              Serial ATA II
  device model          WDC WD5000AAKS-00TMA0
  serial number         WD-WCAPW3059906
  firmware revision     12.01C01
  cylinders             16383
  heads                 16
  sectors/track         63
  lba supported         268435455 sectors
  lba48 supported       976773168 sectors
  dma supported
  overlap not supported

  Feature                      Support  Enable    Value           Vendor
  write cache                    yes      yes
  read ahead                     yes      yes
  Native Command Queuing (NCQ)   yes       -      31/0x1F
  Tagged Command Queuing (TCQ)   no       no      31/0x1F
  SMART                          yes      yes
  microcode download             yes      yes
  security                       yes      no
  power management               yes      yes
  advanced power management      no       no      0/0x00
  automatic acoustic management  yes      no      254/0xFE        128/0x80
# atacontrol cap ad12

  Protocol              Serial ATA v1.0
  device model          WDC WD3200JD-22KLB0
  serial number         WD-WMAMR1280328
  firmware revision     08.05J08
  cylinders             16383
  heads                 16
  sectors/track         63
  lba supported         268435455 sectors
  lba48 supported       625142448 sectors
  dma supported
  overlap not supported

  Feature                      Support  Enable    Value           Vendor
  write cache                    yes      yes
  read ahead                     yes      yes
  Native Command Queuing (NCQ)   no        -      0/0x00
  Tagged Command Queuing (TCQ)   no       no      0/0x00
  SMART                          yes      yes
  microcode download             yes      yes
  security                       yes      no
  power management               yes      yes
  advanced power management      no       no      0/0x00
  automatic acoustic management  yes      no      254/0xFE        128/0x80


and ad6 is only device at atapci0: <Promise PDC40718 SATA300 controller>
and ad12 only device at atapci1: <VIA 6420 SATA150 controller>


Hope I didn't do something stupidly wrong and that
this might give someone a clue.

NB, this test-box is not under my desk but a university in Paris. Next
planned time I need to go there is thuesday. I.e. I can swap disks/cables
but some patience needed

Best,

Arno



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?wpy7d6ew3s.fsf>