RE: Howto: implement AMD SB9xx RAID5 support in dmraid
Mark-Willem Jansen <[email protected]> Sun, 1 Apr 2012 22:21:17 +0200
| Newsgroups | gmane.linux.ataraid |
|---|---|
| Message-ID | <[email protected]> |
--_35021a01-00cc-45f7-a59e-2c5180dda599_ Content-Type: multipart/alternative; boundary="_88e4e76f-98fb-453c-b077-f593011f648a_" --_88e4e76f-98fb-453c-b077-f593011f648a_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hello Phillip (and other people in this list)=2C Here are some outputs of dmraid that may help you understand the situation. $ dmraid -ay -vvv -d WARN: locking /var/lock/dmraid/.lock NOTICE: /dev/sdd: asr discovering NOTICE: /dev/sdd: ddf1 discovering NOTICE: /dev/sdd: hpt37x discovering NOTICE: /dev/sdd: hpt45x discovering NOTICE: /dev/sdd: isw discovering DEBUG: not isw at 500107860992 DEBUG: isw trying hard coded -2115 offset. DEBUG: not isw at 500106779136 NOTICE: /dev/sdd: jmicron discovering NOTICE: /dev/sdd: lsi discovering NOTICE: /dev/sdd: nvidia discovering NOTICE: /dev/sdd: pdc discovering ERROR: pdc: zero sectors on /dev/sdd ERROR: pdc: setting up RAID device /dev/sdd NOTICE: /dev/sdd: sil discovering NOTICE: /dev/sdd: via discovering NOTICE: /dev/sdc: asr discovering NOTICE: /dev/sdc: ddf1 discovering NOTICE: /dev/sdc: hpt37x discovering NOTICE: /dev/sdc: hpt45x discovering NOTICE: /dev/sdc: isw discovering DEBUG: not isw at 500107860992 DEBUG: isw trying hard coded -2115 offset. DEBUG: not isw at 500106779136 NOTICE: /dev/sdc: jmicron discovering NOTICE: /dev/sdc: lsi discovering NOTICE: /dev/sdc: nvidia discovering NOTICE: /dev/sdc: pdc discovering ERROR: pdc: zero sectors on /dev/sdc ERROR: pdc: setting up RAID device /dev/sdc NOTICE: /dev/sdc: sil discovering NOTICE: /dev/sdc: via discovering NOTICE: /dev/sdb: asr discovering NOTICE: /dev/sdb: ddf1 discovering NOTICE: /dev/sdb: hpt37x discovering NOTICE: /dev/sdb: hpt45x discovering NOTICE: /dev/sdb: isw discovering DEBUG: not isw at 500107860992 DEBUG: isw trying hard coded -2115 offset. DEBUG: not isw at 500106779136 NOTICE: /dev/sdb: jmicron discovering NOTICE: /dev/sdb: lsi discovering NOTICE: /dev/sdb: nvidia discovering NOTICE: /dev/sdb: pdc discovering ERROR: pdc: zero sectors on /dev/sdb ERROR: pdc: setting up RAID device /dev/sdb NOTICE: /dev/sdb: sil discovering NOTICE: /dev/sdb: via discovering NOTICE: /dev/sda: asr discovering NOTICE: /dev/sda: ddf1 discovering NOTICE: /dev/sda: hpt37x discovering NOTICE: /dev/sda: hpt45x discovering NOTICE: /dev/sda: isw discovering DEBUG: not isw at 64023256064 DEBUG: isw trying hard coded -2115 offset. DEBUG: not isw at 64022174208 NOTICE: /dev/sda: jmicron discovering NOTICE: /dev/sda: lsi discovering NOTICE: /dev/sda: nvidia discovering NOTICE: /dev/sda: pdc discovering NOTICE: /dev/sda: sil discovering NOTICE: /dev/sda: via discovering no raid disks WARN: unlocking /var/lock/dmraid/.lock I have also attached the output of "dmraid -rD" Hopefully you can point me to the correct offset for the metadata. Kind regards=2C Mark-Willem > Date: Fri=2C 23 Mar 2012 13:23:46 -0400 > From: [email protected] > To: [email protected] > CC: [email protected]=3B [email protected]=3B [email protected] > Subject: Re: Howto: implement AMD SB9xx RAID5 support in dmraid >=20 > On 3/23/2012 12:52 PM=2C Mark-Willem Jansen wrote: > > > > Dear dmraid developer(s)/maintainer(s)=2C > > > > Since a few weeks I am trying to get my new rig up and running under > > linux. I run a dual-boot system with Windows 7 and Debian/wheezy. I > > have installed the OS files on a small SSD and have setup a RAID-5 > > for the rest of the files. It is a FakeRAID/SATA-RAID which is > > controlled by the AMD SB950 chip. Unfortunately dmraid does not seem > > to support this chip in combination with a RAID-5 setup. I would like > > to implement this=2C but to do that I need some help and information. > > The plan is as follows >=20 > I would advise using mdadm instead since it is much better supported and= =20 > more reliable and feature complete. >=20 > > - Get dmraid working with the dm-raid module. Dmraid needs a > > dm-target that support RAID-5. In the past there was a patch for the > > prior 3.x linux-kernel which added the dm-raid45 module. Now a days > > there is a module called dm-raid which will provide the same > > aliases(dm-raid4=2C dm-raid5) as the dm-raid45 module. How does on > > check if dmraid can be made compatible with this new module? My first > > idea was to look at the symbols dm-raid and dm-raid45 export to the > > user space and which of these symbols are used by dmriad. >=20 > I believe it should just work unless they changed the format of the=20 > dmsetup table. >=20 > > - Find the metadata offset for the SB950 chipset. For this I need not > > know with which characters the header of the metadata starts? The > > SB950 seems to contain a Promise controller. >=20 > dmraid -n doesn't already recognize it? You can find the existing known= =20 > offsets in pdc.h. >=20 > > - Interpret the metadata provided on the disks. From the setup screen > > provided by the UEFI driver I can find the stripe size=2C but in which > > sequence(ls=2Crs=2Cla=2Cra) these stripes are order I do not know=2C ye= t. > > What is the best way to find this sequence? The RAID-5 disk hold a > > GPT table=2C maybe one can look at how the header of this table is > > distributed over the disk to find the stripe sequence. >=20 > dmraid already understands pdc metadata. >=20 > > - [optionally] Implement GPT support. >=20 > Ubtunu 12.04 has switched to letting kpartx activate the partitions=20 > instead of dmraid=2C which supports GPT. >=20 = --_88e4e76f-98fb-453c-b077-f593011f648a_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable <html> <head> <style><!-- .hmmessage P { margin:0px=3B padding:0px } body.hmmessage { font-size: 10pt=3B font-family:Tahoma } --></style></head> <body class=3D'hmmessage'><div dir=3D'ltr'> Hello Phillip (and other people in this list)=2C<br><br>Here are some outpu= ts of dmraid that may help you understand the situation.<br><br>$ dmraid -a= y -vvv -d<br>WARN: locking /var/lock/dmraid/.lock<br>NOTICE: /dev/sdd: asr&= nbsp=3B =3B =3B =3B discovering<br>NOTICE: /dev/sdd: ddf1 = =3B =3B =3B discovering<br>NOTICE: /dev/sdd: hpt37x =3B discove= ring<br>NOTICE: /dev/sdd: hpt45x =3B discovering<br>NOTICE: /dev/sdd: i= sw =3B =3B =3B =3B discovering<br>DEBUG: not isw at 5001078= 60992<br>DEBUG: isw trying hard coded -2115 offset.<br>DEBUG: not isw at 50= 0106779136<br>NOTICE: /dev/sdd: jmicron discovering<br>NOTICE: /dev/sdd: ls= i =3B =3B =3B =3B discovering<br>NOTICE: /dev/sdd: nvidia&n= bsp=3B discovering<br>NOTICE: /dev/sdd: pdc =3B =3B =3B =3B= discovering<br>ERROR: pdc: zero sectors on /dev/sdd<br>ERROR: pdc: setting= up RAID device /dev/sdd<br>NOTICE: /dev/sdd: sil =3B =3B =3B&n= bsp=3B discovering<br>NOTICE: /dev/sdd: via =3B =3B =3B =3B= discovering<br>NOTICE: /dev/sdc: asr =3B =3B =3B =3B disco= vering<br>NOTICE: /dev/sdc: ddf1 =3B =3B =3B discovering<br>NOT= ICE: /dev/sdc: hpt37x =3B discovering<br>NOTICE: /dev/sdc: hpt45x = =3B discovering<br>NOTICE: /dev/sdc: isw =3B =3B =3B =3B di= scovering<br>DEBUG: not isw at 500107860992<br>DEBUG: isw trying hard coded= -2115 offset.<br>DEBUG: not isw at 500106779136<br>NOTICE: /dev/sdc: jmicr= on discovering<br>NOTICE: /dev/sdc: lsi =3B =3B =3B =3B dis= covering<br>NOTICE: /dev/sdc: nvidia =3B discovering<br>NOTICE: /dev/sd= c: pdc =3B =3B =3B =3B discovering<br>ERROR: pdc: zero sect= ors on /dev/sdc<br>ERROR: pdc: setting up RAID device /dev/sdc<br>NOTICE: /= dev/sdc: sil =3B =3B =3B =3B discovering<br>NOTICE: /dev/sd= c: via =3B =3B =3B =3B discovering<br>NOTICE: /dev/sdb: asr=  =3B =3B =3B =3B discovering<br>NOTICE: /dev/sdb: ddf1 = =3B =3B =3B discovering<br>NOTICE: /dev/sdb: hpt37x =3B discove= ring<br>NOTICE: /dev/sdb: hpt45x =3B discovering<br>NOTICE: /dev/sdb: i= sw =3B =3B =3B =3B discovering<br>DEBUG: not isw at 5001078= 60992<br>DEBUG: isw trying hard coded -2115 offset.<br>DEBUG: not isw at 50= 0106779136<br>NOTICE: /dev/sdb: jmicron discovering<br>NOTICE: /dev/sdb: ls= i =3B =3B =3B =3B discovering<br>NOTICE: /dev/sdb: nvidia&n= bsp=3B discovering<br>NOTICE: /dev/sdb: pdc =3B =3B =3B =3B= discovering<br>ERROR: pdc: zero sectors on /dev/sdb<br>ERROR: pdc: setting= up RAID device /dev/sdb<br>NOTICE: /dev/sdb: sil =3B =3B =3B&n= bsp=3B discovering<br>NOTICE: /dev/sdb: via =3B =3B =3B =3B= discovering<br>NOTICE: /dev/sda: asr =3B =3B =3B =3B disco= vering<br>NOTICE: /dev/sda: ddf1 =3B =3B =3B discovering<br>NOT= ICE: /dev/sda: hpt37x =3B discovering<br>NOTICE: /dev/sda: hpt45x = =3B discovering<br>NOTICE: /dev/sda: isw =3B =3B =3B =3B di= scovering<br>DEBUG: not isw at 64023256064<br>DEBUG: isw trying hard coded = -2115 offset.<br>DEBUG: not isw at 64022174208<br>NOTICE: /dev/sda: jmicron= discovering<br>NOTICE: /dev/sda: lsi =3B =3B =3B =3B disco= vering<br>NOTICE: /dev/sda: nvidia =3B discovering<br>NOTICE: /dev/sda:= pdc =3B =3B =3B =3B discovering<br>NOTICE: /dev/sda: sil&n= bsp=3B =3B =3B =3B discovering<br>NOTICE: /dev/sda: via =3B=  =3B =3B =3B discovering<br>no raid disks<br>WARN: unlocking /v= ar/lock/dmraid/.lock<br><br>I have also attached the output of "dmraid -rD"= <br><br>Hopefully you can point me to the correct offset for the metadata.<= br><br>Kind regards=2C<br><br>Mark-Willem<br><br><div><div id=3D"SkyDrivePl= aceholder"></div>>=3B Date: Fri=2C 23 Mar 2012 13:23:46 -0400<br>>=3B F= rom: [email protected]<br>>=3B To: [email protected]<br>>=3B CC: at= [email protected]=3B [email protected]=3B [email protected]<br>>= =3B Subject: Re: Howto: implement AMD SB9xx RAID5 support in dmraid<br>>= =3B <br>>=3B On 3/23/2012 12:52 PM=2C Mark-Willem Jansen wrote:<br>>=3B= >=3B<br>>=3B >=3B Dear dmraid developer(s)/maintainer(s)=2C<br>>= =3B >=3B<br>>=3B >=3B Since a few weeks I am trying to get my new rig= up and running under<br>>=3B >=3B linux. I run a dual-boot system with= Windows 7 and Debian/wheezy. I<br>>=3B >=3B have installed the OS file= s on a small SSD and have setup a RAID-5<br>>=3B >=3B for the rest of t= he files. It is a FakeRAID/SATA-RAID which is<br>>=3B >=3B controlled b= y the AMD SB950 chip. Unfortunately dmraid does not seem<br>>=3B >=3B t= o support this chip in combination with a RAID-5 setup. I would like<br>>= =3B >=3B to implement this=2C but to do that I need some help and informa= tion.<br>>=3B >=3B The plan is as follows<br>>=3B <br>>=3B I would = advise using mdadm instead since it is much better supported and <br>>=3B= more reliable and feature complete.<br>>=3B <br>>=3B >=3B - Get dmra= id working with the dm-raid module. Dmraid needs a<br>>=3B >=3B dm-targ= et that support RAID-5. In the past there was a patch for the<br>>=3B >= =3B prior 3.x linux-kernel which added the dm-raid45 module. Now a days<br>= >=3B >=3B there is a module called dm-raid which will provide the same<= br>>=3B >=3B aliases(dm-raid4=2C dm-raid5) as the dm-raid45 module. How= does on<br>>=3B >=3B check if dmraid can be made compatible with this = new module? My first<br>>=3B >=3B idea was to look at the symbols dm-ra= id and dm-raid45 export to the<br>>=3B >=3B user space and which of the= se symbols are used by dmriad.<br>>=3B <br>>=3B I believe it should jus= t work unless they changed the format of the <br>>=3B dmsetup table.<br>&= gt=3B <br>>=3B >=3B - Find the metadata offset for the SB950 chipset. F= or this I need not<br>>=3B >=3B know with which characters the header o= f the metadata starts? The<br>>=3B >=3B SB950 seems to contain a Promis= e controller.<br>>=3B <br>>=3B dmraid -n doesn't already recognize it? = You can find the existing known <br>>=3B offsets in pdc.h.<br>>=3B <br= >>=3B >=3B - Interpret the metadata provided on the disks. From the set= up screen<br>>=3B >=3B provided by the UEFI driver I can find the strip= e size=2C but in which<br>>=3B >=3B sequence(ls=2Crs=2Cla=2Cra) these s= tripes are order I do not know=2C yet.<br>>=3B >=3B What is the best wa= y to find this sequence? The RAID-5 disk hold a<br>>=3B >=3B GPT table= =2C maybe one can look at how the header of this table is<br>>=3B >=3B = distributed over the disk to find the stripe sequence.<br>>=3B <br>>=3B= dmraid already understands pdc metadata.<br>>=3B <br>>=3B >=3B - [op= tionally] Implement GPT support.<br>>=3B <br>>=3B Ubtunu 12.04 has swit= ched to letting kpartx activate the partitions <br>>=3B instead of dmraid= =2C which supports GPT.<br>>=3B <br></div> </div></body> </html>= --_88e4e76f-98fb-453c-b077-f593011f648a_-- --_35021a01-00cc-45f7-a59e-2c5180dda599_ Content-Type: application/x-bzip Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="SB950-raid5_dmraid.tar.bz2" QlpoOTFBWSZTWSnj8lQABkF/////////////////////////////////////////////0AUe8pAo BIgQBJgAJgACYAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAJMABMAATAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAABJgAJgACYAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAADeqkgKT9TKA BoPUBoNAAAaDQAAAAB6mQBoAaNBoAaMgBoAAaAANA0AAAAAADQAAAApVEIJT0TT9UaA8kPUZpGgY EbU0Gj9NFGgBpkGmjTRp5IA2UaGQ2o0AAAAAAaYgAGgNAGRoaDIAANGQBk2pubHJlxMt023Tb2ax tpZbG4DgY2GjZUzGNpWKpwNSpWXJCNHUYhQkisyepFTI1TyFCaVLq6wMwNUrRKZNIlhWpSik02XK pVJbVRKC4pcqpM/DEyJiZDsyZLqtgma4gxqMY6LG3GgsV+V18YWvYWlElZdY5YV2AkqsUlm+Kltb UkUsyXciWrd4sLcSWFtdq68F/X1bFpRasolZ/tMNGw3SzLXI68jeEaloW8ozreka9G9o3ZGW7Ajs KN8R2JG17IjcaxGh5Mb6jAb8jf0cAjMmuuBRwSOzo7QjjlsEdpR2MrhEcmnUE2d2q7YjtqOFRtO3 o00bFGojZI2aO4I1aOUuGRqI5C4ZG222THHHGmOLfkYR1eNPZ5Y1LCk56xSVklYiaupJEwjCMG3i 8g2JSJZuUlJqikuGaxRhWXMmysVq0ws7NYU2WDB3TJYMEZLBOHsbqKMiYyyJnalnI1VpZbJn2qRn 6GZmz4zkzSZrbWNimXTRxForBG73EozXfEZrvqO/I7+jikeARubwKPBI4xHg0cajdrwiPCo45Hhk bRGtt5Ru9tUa9GNvaNhiuNRtTj0d649eHR128QjxKNG4nxSN/RpWXxaO5IxWxR8fxqPHI5BG2Rxa Nht0aJ49HBI4O8gjKnCo7xLJX6JgowUScQqDYVJwsSmrM1jboliJKl7+2G4SqqUzy/8ctM2YeJLt WcHEnFK7Hz0/hpTERJYW4kyplhbhHK2rvMI325VGVGitYjWJwlkRulrbPs663bwjAmcTzY1wz080 jMjqSaduqcHZ6eMTqaOv2helOKjrFrbdbXo3pNJNFkiqJvMX2JwyyMwZfp2exORGWjGyTEi9Gcs7 GUsMNF5McoMSLzFDFDBsWOVHLDlxzAzcahJzI5oc2Lqc4OdGMHPDnx0A6EdEOjHSDpR0w6cdQOpH VDqx1gxowAxkk60dcOvHYDsR2Q7MdoO1HbDtx3A7kd0McO7Guu8HejvhesLI0zv5GPi+aXLF0C+e W0Lub6BfRL6Rc2X0y+oX1S+sX1y+wXur7JfaLnC50udLTvtl9wueL7pfeLky6s++XPlyhapPwF+E vxF+Mst0Ea6Oy3vboY/JH5Y/NH5489Gk/RH6Y/VH64/ZH7Y/dH74/hH8Y/lH846KP6R/WOjjo41h 0kf2j+8f4j/Mf6j28Z9/uP+R/2Oljpo/9HTxy0aBG5jpY5wunq9lYRjbGzoyI8q4fn7ob/4u5Ipw oSBTx+So --_35021a01-00cc-45f7-a59e-2c5180dda599_ Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Ataraid-list mailing list [email protected] https://www.redhat.com/mailman/listinfo/ataraid-list --_35021a01-00cc-45f7-a59e-2c5180dda599_--