AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:
CFG:
PT:0
S:andy.sharp@lsi.com
RQ:
SSV:mhbs.lsil.com
NSV:
SSH:
R:<Patrick.Haverty@lsi.com>,<Brian.Stark@lsi.com>
MAID:2
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
RMID:#imap/LSI/INBOX	0	08A1E1BF683B9844B320294C380BF34C04256D1A@cosmail03.lsi.com
X-Sylpheed-End-Special-Headers: 1
Date: Thu, 3 Dec 2009 12:01:46 -0800
From: Andrew Sharp <andy.sharp@lsi.com>
To: "Haverty, Patrick" <Patrick.Haverty@lsi.com>
Cc: "Stark, Brian" <Brian.Stark@lsi.com>
Subject: Re: Compact Flash
Message-ID: <20091203120146.51f227df@ripper.onstor.net>
In-Reply-To: <08A1E1BF683B9844B320294C380BF34C04256D1A@cosmail03.lsi.com>
References: <E1EC65251D4B3D46BBC0AAA3C0629222AFC68702@cosmail02.lsi.com>
	<08A1E1BF683B9844B320294C380BF34C04256D1A@cosmail03.lsi.com>
Organization: LSI
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

You actually don't want to mount it.

Instead use dd but use

dd if=/dev/sdb1 of=/dev/null    for read testing

dd of=/dev/sdb1 if=/dev/zero    for write testing

You will of course have to reformat the flash after doing write testing.

If you mount it, the filesystem and caching semantics will get in the
way of getting reliable numbers.

Cheers,

a

On Thu, 3 Dec 2009 12:30:26 -0700 "Haverty, Patrick"
<Patrick.Haverty@lsi.com> wrote:

> Cool, I'll give it a go.  Thanks. 
> 
> -----Original Message-----
> From: Stark, Brian 
> Sent: Thursday, December 03, 2009 11:14 AM
> To: Haverty, Patrick; Sharp, Andy
> Subject: Re: Compact Flash
> 
> To mount the secondary, do the following from the Linux prompt:
> 
> mount /dev/sdb1 /mnt
> 
> That assumes root is mounted on sda1, e.g. that's the flash you
> booted.  If you're actually booted from sdb1, then change the above
> to /dev/sda1.
> 
> You can then cd to /mnt and see everything on the secondary.
> 
> 
> 
> ----- Original Message -----
> From: Haverty, Patrick
> To: Stark, Brian; Sharp, Andy
> Sent: Thu Dec 03 11:58:07 2009
> Subject: RE: Compact Flash
> 
> I tried to get an idea of throughput using dd.  First I make sure the
> file is erased.  Then I write it.  Then I read it.  Then I do it all
> over again.  Results below from one of the new Unigen CF modules.
> I'm running it from / so I'm assuming this is writing and reading
> from the primary CF module.
> 
> If I knew how to mount the secondary flash I could probably run a
> larger file size and might end up with a little more accurate
> number.  But maybe not.
> 
> - Pat
> 
> pleasanton:/# rm temppjh
> pleasanton:/# time -p dd if=/dev/zero of=/temppjh bs=128k count=256
> 256+0 records in
> 256+0 records out
> 33554432 bytes (34 MB) copied, 2.83429 seconds, 11.8 MB/s real 2.84
> user 0.00 sys 1.01 pleasanton:/# time -p dd if=/temppjh of=/dev/null
> bs=128k count=256 256+0 records in 256+0 records out
> 33554432 bytes (34 MB) copied, 4.89549 seconds, 6.9 MB/s real 5.50
> user 0.00 sys 1.53 pleasanton:/# rm temppjh pleasanton:/# time -p dd
> if=/dev/zero of=/temppjh bs=128k count=256 256+0 records in 256+0
> records out 33554432 bytes (34 MB) copied, 2.66415 seconds, 12.6 MB/s
> real 2.67 user 0.00 sys 0.83 pleasanton:/# time -p dd if=/temppjh
> of=/dev/null bs=128k count=256 256+0 records in 256+0 records out
> 33554432 bytes (34 MB) copied, 4.88966 seconds, 6.9 MB/s real 5.25
> user 0.00 sys 1.82 pleasanton:/# rm temppjh pleasanton:/# time -p dd
> if=/dev/zero of=/temppjh bs=128k count=256 256+0 records in 256+0
> records out 33554432 bytes (34 MB) copied, 4.36816 seconds, 7.7 MB/s
> real 4.45 user 0.00 sys 1.58 pleasanton:/# time -p dd if=/temppjh
> of=/dev/null bs=128k count=256 256+0 records in 256+0 records out
> 33554432 bytes (34 MB) copied, 3.98721 seconds, 8.4 MB/s real 4.41
> user 0.00 sys 1.32 pleasanton:/# rm temppjh pleasanton:/# time -p dd
> if=/dev/zero of=/temppjh bs=128k count=256 256+0 records in 256+0
> records out 33554432 bytes (34 MB) copied, 4.69891 seconds, 7.1 MB/s
> real 4.76 user 0.00 sys 1.57 pleasanton:/# time -p dd if=/temppjh
> of=/dev/null bs=128k count=256 256+0 records in 256+0 records out
> 33554432 bytes (34 MB) copied, 6.46484 seconds, 5.2 MB/s real 6.87
> user 0.00 sys 1.46 pleasanton:/# rm temppjh pleasanton:/# time -p dd
> if=/dev/zero of=/temppjh bs=128k count=256 256+0 records in 256+0
> records out 33554432 bytes (34 MB) copied, 4.17214 seconds, 8.0 MB/s
> real 4.17 user 0.00 sys 1.45 pleasanton:/# time -p dd if=/temppjh
> of=/dev/null bs=128k count=256 256+0 records in 256+0 records out
> 33554432 bytes (34 MB) copied, 2.25435 seconds, 14.9 MB/s real 2.56
> user 0.00 sys 0.75 
> 
> -----Original Message-----
> From: Stark, Brian
> Sent: Wednesday, December 02, 2009 5:34 PM
> To: Sharp, Andy; Haverty, Patrick
> Subject: RE: Compact Flash
> 
> One more thing...
> 
> The code that checks the minimum size cannot be overridden.  There's
> a strict check against CF_MIN_SIZE which is defined as 980MB for the
> Coog.
> 
> So, if we want to use any of these flashes in the field, there must
> be a code change...
> 
> 
> 
> -----Original Message-----
> From: Stark, Brian
> Sent: Wednesday, December 02, 2009 5:30 PM
> To: Sharp, Andy; Haverty, Patrick
> Subject: RE: Compact Flash
> 
> We didn't pop out any secondary flashes that were mounted.  Surely, I
> checked for that before doing anything.
> 
> In looking at the code in ssc-nfxsh/cmd_flash.c, it looks like
> there's a minimum size requirement of 980MB.  The first question, of
> course, is why?
> 
> Depending on the answer to that, should we consider changing it?  We
> really need to get at least 2 more CF vendors qualified since we only
> have 1 right now.
> 
> And what about the throughput question?
> 
> 
> -----Original Message-----
> From: Andrew Sharp [mailto:andy.sharp@lsi.com]
> Sent: Wednesday, December 02, 2009 4:33 PM
> To: Haverty, Patrick
> Cc: Stark, Brian
> Subject: Re: Compact Flash
> 
> I can answer all your questions in one sentence.  Oops, no I can't.
> 
> Anyway, the first series of errors is because you popped out/in CF
> card(s) in a bad way.  Bad Pattie.
> <slap-don't-slap-do-slap-that-slap-again-slap>
> 
> Be nice to your kernel, and it will be nice to you ~:^)  Always make
> sure you unmount any mounted partitions *before* pulling a CF card.
> Ignore the LEDs, I've seen the code that controls those ~:^)
> 
> As for the second message, yes, "system copy init" has a minimum size
> requirement.  That card is too small.  I think there's a secret
> hidden option to ignore that, but I don't remember.  There's also
> flash-install.sh, which you should remember.  That's documented in
> the wiki somewhere.  And that doc is probably not more than 3 years
> out of date.  flash-install very likely has a way to override the min
> size limit, because I wrote it and that's the way I roll; it's a
> shell script, so it might be "easy" to look at it and see if you can
> figure out what the secret option is for that.
> 
> I think that's it.  Good!  Time for my nap.
> 
> 
> 
> On Wed, 2 Dec 2009 16:13:04 -0700 "Haverty, Patrick"
> <Patrick.Haverty@lsi.com> wrote:
> 
> > And I meant to ask if you had a method to test the read/write 
> > throughput of CF modules which I might use?
> > 
> > Thanks,
> > 
> > Pat
> > 
> > ________________________________
> > From: Haverty, Patrick
> > Sent: Wednesday, December 02, 2009 2:51 PM
> > To: Sharp, Andy
> > Cc: Stark, Brian
> > Subject: Compact Flash
> > 
> > Hello Andy,
> > 
> > Brian and I are trying out some new Compact Flash modules to try
> > and find an alternate part.  Below I'll paste some of the console
> > messages seen when trying the new modules.  Do you know if the size
> > error could be caused by a mismatch between the CF module
> > configurations (different vendors in each slot), and if so can you
> > tell me how I can create a new CF rather than copy an existing one?
> > 
> > Brian could probably speak more intelligently about what we are 
> > seeing, but he's tied up in a phone meeting right now.
> > 
> > Thanks,
> > 
> > Pat
> > 
> > Stec CF Module
> > 
> > (Insert module)
> > 
> > ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in Buffer I/O error on device sdb, logical block 0
> > ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in Buffer I/O error on device sdb, logical block 0
> > 
> > 
> > (Try copy init)
> > 
> > pleasanton diag> system copy init
> > ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:20:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 16384 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:20:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 16384 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:20:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 16384 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:20:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 16384 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:20:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 16384 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:20:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 16384 in Buffer I/O error on device sdb, logical block 0 Buffer I/O 
> > error on device sdb, logical block 1 Buffer I/O error on device
> > sdb, logical block 2 Buffer I/O error on device sdb, logical block 3
> > ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
> > ata3.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 cdb 0x0 data
> > 4096 in Buffer I/O error on device sdb, logical block 0 Secondary
> > CF card is 977 MiB Disk size 1024966656 bytes is not supported
> > Problem initializing CF card /dev/sdb % Command failure.
> > ******************
> > 
> > Unmarked CF Module
> > 
> > (No console messages upon insertion)
> > 
> > (Try copy init)
> > 
> > pleasanton diag> system copy init
> > Secondary CF card is 977 MiB
> > Disk size 1024966656 bytes is not supported Problem initializing CF 
> > card /dev/sdb % Command failure.
> > 
> > ******************
> > 
> > Viking CF Module
> > 
> > (No console messages upon insertion)
> > 
> > (Try copy init)
> > 
> > 
> > pleasanton diag> system copy init
> > Secondary CF card is 955 MiB
> > Disk size 1002258432 bytes is not supported Problem initializing CF 
> > card /dev/sdb % Command failure.
> > 
