<p>I think it has some slow write performance since two page write is lower than one page. It means cache program is not work as expected.<br>
That's reason no warning message. In my case I enabled the sync write and get 10MiB or more performance. Please see my speedtest result.</p>
<p>Thank you,<br>
Kyungmin Park</p>
<p>Sent from Android </p>
<div class="gmail_quote">On Nov 4, 2010 9:40 PM, "Roman Tereshonkov" <<a href="mailto:roman.tereshonkov@nokia.com">roman.tereshonkov@nokia.com</a>> wrote:<br type="attribution">> <br>> Hi Park,<br>> <br>
> I tried several times<br>> fsstress -p 3 -n 10000 -d /mnt/ubifs <br>> for 28 MB onenand partition. No warnings observed.<br>> The logs are below.<br>> <br>> <br>> # modprobe ubi mtd=4<br>> [ 101.013549] UBI: attaching mtd4 to ubi0<br>
> [ 101.017486] UBI: physical eraseblock size: 262144 bytes (256 KiB)<br>> [ 101.023895] UBI: logical eraseblock size: 253952 bytes<br>> [ 101.029510] UBI: smallest flash I/O unit: 4096<br>> [ 101.034271] UBI: VID header offset: 4096 (aligned 4096)<br>
> [ 101.040405] UBI: data offset: 8192<br>> [ 101.054565] UBI: empty MTD device detected<br>> [ 101.058868] UBI: create volume table (copy #1)<br>> [ 101.067504] UBI: create volume table (copy #2)<br>
> [ 101.077239] UBI: attached mtd4 to ubi0<br>> [ 101.081054] UBI: MTD device name: "swap"<br>> [ 101.086090] UBI: MTD device size: 32 MiB<br>> [ 101.091033] UBI: number of good PEBs: 123<br>
> [ 101.095733] UBI: number of bad PEBs: 5<br>> [ 101.100280] UBI: max. allowed volumes: 128<br>> [ 101.104949] UBI: wear-leveling threshold: 8192<br>> [ 101.109771] UBI: number of internal volumes: 1<br>
> [ 101.114257] UBI: number of user volumes: 0<br>> [ 101.118804] UBI: available PEBs: 117<br>> [ 101.123504] UBI: total number of reserved PEBs: 6<br>> [ 101.128295] UBI: number of PEBs reserved for bad PEB handling: 2<br>
> [ 101.134399] UBI: max/mean erase counter: 0/0<br>> [ 101.138763] UBI: image sequence number: 0<br>> [ 101.142852] UBI: background thread "ubi_bgt0d" started, PID 104<br>> [ 101.150085] modprobe used greatest stack depth: 6068 bytes left<br>
> # modprobe ubifs<br>> # ubimkvol /dev/ubi0 -N test -m<br>> Set volume size to 29712384<br>> Volume ID 0, size 117 LEBs (29712384 bytes, 28.3 MiB), LEB size [ 111.348449] ubimkvol used greatest stack depth: 5820 bytes left<br>
> 253952 bytes (248.0 KiB), dynamic, name "test", alignment 1<br>> #mount -t ubifs ubi0_0 /mnt/ubifs<br>> [ 133.515991] UBIFS: default file-system created<br>> [ 133.565490] UBIFS: mounted UBI device 0, volume 0, name "test"<br>
> [ 133.571594] UBIFS: file system size: 27426816 bytes (26784 KiB, 26 MiB, 108 LEBs)<br>> [ 133.579406] UBIFS: journal size: 2031617 bytes (1984 KiB, 1 MiB, 6 LEBs)<br>> [ 133.586791] UBIFS: media format: w4/r0 (latest is w4/r0)<br>
> [ 133.592712] UBIFS: default compressor: lzo<br>> [ 133.596893] UBIFS: reserved for root: 1295435 bytes (1265 KiB)<br>> [ 133.604675] mount used greatest stack depth: 5768 bytes left<br>> # fsstress -p 3 -n 10 -d /mnt/ubifs<br>
> seed = 397323<br>> # fsstress -p 3 -n 100 -d /mnt/ubifs<br>> seed = 826848<br>> # fsstress -p 3 -n 1000 -d /mnt/ubifs<br>> seed = 209658<br>> # fsstress -p 3 -n 10000 -d /mnt/ubifs<br>> seed = 748381<br>
> #<br>> <br>> <br>> mtd_speedtest gives no warnings also. Only bad block related errors:<br>> <br>> # insmod mtd_speedtest.ko dev=4<br>> [ 23.759948] <br>> [ 23.761505] =================================================<br>
> [ 23.767639] mtd_speedtest: MTD device: 4<br>> [ 23.771728] mtd_speedtest: MTD device size 33554432, eraseblock size 262144, page size 4096, count of eraseblocks 128, pages per eraseblock 64, OOB size 128<br>> [ 23.800933] mtd_speedtest: scanning for bad eraseblocks<br>
> [ 23.808441] Bad eraseblock 42 at 0x00a80000<br>> [ 23.812713] mtd_speedtest: block 22 is bad<br>> [ 23.818542] Bad eraseblock 60 at 0x00f00000<br>> [ 23.822814] mtd_speedtest: block 40 is bad<br>> [ 23.827423] Bad eraseblock 65 at 0x01040000<br>
> [ 23.831665] mtd_speedtest: block 45 is bad<br>> [ 23.837860] Bad eraseblock 87 at 0x015c0000<br>> [ 23.842163] mtd_speedtest: block 67 is bad<br>> [ 23.850891] Bad eraseblock 136 at 0x02200000<br>> [ 23.855224] mtd_speedtest: block 116 is bad<br>
> [ 23.860534] mtd_speedtest: scanned 128 eraseblocks, 5 are bad<br>> [ 24.010498] mtd_speedtest: testing eraseblock write speed<br>> [ 26.267669] mtd_speedtest: eraseblock write speed is 13988 KiB/s<br>> [ 26.273803] mtd_speedtest: testing eraseblock read speed<br>
> [ 27.286437] mtd_speedtest: eraseblock read speed is 31269 KiB/s<br>> [ 27.430114] mtd_speedtest: testing page write speed<br>> [ 36.082855] mtd_speedtest: page write speed is 3641 KiB/s<br>> [ 36.088409] mtd_speedtest: testing page read speed<br>
> [ 37.086517] mtd_speedtest: page read speed is 31678 KiB/s<br>> [ 37.444854] mtd_speedtest: testing 2 page write speed<br>> [ 60.709533] mtd_speedtest: 2 page write speed is 1353 KiB/s<br>> [ 60.715148] mtd_speedtest: testing 2 page read speed<br>
> [ 61.624176] mtd_speedtest: 2 page read speed is 34870 KiB/s<br>> [ 61.629791] mtd_speedtest: Testing erase speed<br>> [ 61.868347] mtd_speedtest: erase speed is 133991 KiB/s<br>> [ 61.873565] mtd_speedtest: finished<br>
> [ 61.877105] =================================================<br>> <br>> <br>> <br>> Regards<br>> Roman Tereshonkov<br>> <br>> On Thu, Nov 04, 2010 at 10:10:36AM +0100, ext Kyungmin Park wrote:<br>
>> On Thu, Nov 4, 2010 at 5:58 PM, <<a href="mailto:roman.tereshonkov@nokia.com">roman.tereshonkov@nokia.com</a>> wrote:<br>>> ><br>>> > Hi,<br>>> ><br>>> ><br>>> > I used driver/mtd/tests and did not observe any warnings except a couple of bad blocks for some boards.<br>
>> Yes in case of mtd_test* there's no messages.<br>>> ><br>>> > Can you discribe your test case for me to reproduce it. And show your kernel log when it happens if possible?<br>>> > Is it for 2kb or 4kb page onenand?<br>
>> <br>>> Mount ubifs on OneNAND which has 4KiB pagesize and run fsstress -p 3<br>>> -n 1000000000 -d /ubifs<br>>> <br>>> then you can or maybe see the UBI scrubbing message.<br>>> <br>
>> Thank you,<br>>> Kyungmin Park<br>>> ><br>>> ><br>>> > Regards<br>>> > Roman Tereshonkov<br>>> ><br>>> ><br>>> >>-----Original Message-----<br>
>> >>From: <a href="mailto:kyungmin78@gmail.com">kyungmin78@gmail.com</a> [mailto:<a href="mailto:kyungmin78@gmail.com">kyungmin78@gmail.com</a>] On<br>>> >>Behalf Of ext Kyungmin Park<br>>> >>Sent: 04 November, 2010 03:15<br>
>> >>To: Tereshonkov Roman (Nokia-MS/Helsinki)<br>>> >>Cc: <a href="mailto:linux-mtd@lists.infradead.org">linux-mtd@lists.infradead.org</a><br>>> >>Subject: Re: [PATCHv2 0/3] mtd: onenand: add cache program<br>
>> >>feature for 4kb page onenand<br>>> >><br>>> >>Hi,<br>>> >><br>>> >>One concern is that after apply patch, there are some 1-bit ECC<br>>> >>warnings so UBI scrubbing messages are displayed.<br>
>> >><br>>> >>Are there these message at your board?<br>>> >><br>>> >>Thank you,<br>>> >>Kyungmin Park<br>>> >><br>>> >>On Wed, Nov 3, 2010 at 7:55 PM, Roman Tereshonkov<br>
>> >><<a href="mailto:roman.tereshonkov@nokia.com">roman.tereshonkov@nokia.com</a>> wrote:<br>>> >>> Implement cache program feature for 4KB page onenand.<br>>> >>> This feature improves the write data performance.<br>
>> >>> The observed 128KB data program speed change is<br>>> >>> from 8827KB/s to 14156 KB/s when the feature is enabled.<br>>> >>><br>>> >>> Roman Tereshonkov (3):<br>
>> >>> mtd: onenand: add option and variable for cache program feature<br>>> >>> mtd: onenand: fix omap2 code to handle cache program feature<br>>> >>> mtd: onenand: implement cache program feature for 4kb page onenand<br>
>> >>><br>>> >>> drivers/mtd/onenand/omap2.c | 12 +++++++++---<br>>> >>> drivers/mtd/onenand/onenand_base.c | 22 +++++++++++++++++++---<br>>> >>> include/linux/mtd/onenand.h | 12 ++++++++++++<br>
>> >>> 3 files changed, 40 insertions(+), 6 deletions(-)<br>>> >>><br>>> >>><br>>> >>> ______________________________________________________<br>>> >>> Linux MTD discussion mailing list<br>
>> >>> <a href="http://lists.infradead.org/mailman/listinfo/linux-mtd/">http://lists.infradead.org/mailman/listinfo/linux-mtd/</a><br>>> >>><br>>> >><br>>> > ______________________________________________________<br>
>> > Linux MTD discussion mailing list<br>>> > <a href="http://lists.infradead.org/mailman/listinfo/linux-mtd/">http://lists.infradead.org/mailman/listinfo/linux-mtd/</a><br>>> ><br></div>