| 03:31.06 | *** join/#edev purl (~purl@rikers.org) |
| 03:31.06 | *** topic/#edev is Discussion of embedded systems (Hardware/Software/Linux/Commercial Device Repurposing) : Busybox and Buildroot : http://www.elinux.org |
| 03:34.21 | *** join/#edev roxfan2 (dunno@91.180.52.64) |
| 03:54.19 | *** join/#edev purl (~purl@207135154175.dsl.xmission.com) |
| 03:54.19 | *** topic/#edev is Discussion of embedded systems (Hardware/Software/Linux/Commercial Device Repurposing) : Busybox and Buildroot : http://www.elinux.org |
| 03:58.50 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 04:22.49 | *** join/#edev purl (~purl@207135154175.dsl.xmission.com) |
| 04:22.49 | *** topic/#edev is Discussion of embedded systems (Hardware/Software/Linux/Commercial Device Repurposing) : Busybox and Buildroot : http://www.elinux.org |
| 04:27.01 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 04:39.46 | *** join/#edev mbuf (user@nat/redhat/x-kucafizwxcvhcsqj) |
| 05:03.18 | *** join/#edev JoeMoose (~snork@cpe-76-87-206-119.socal.res.rr.com) |
| 05:03.20 | *** join/#edev JoeMoose (~snork@unaffiliated/joellama) |
| 06:27.59 | *** join/#edev JoeLlama (~snork@cpe-76-87-206-119.socal.res.rr.com) |
| 06:27.59 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 07:21.19 | *** join/#edev bizhanMona (~bizhan@nat/cisco/x-jcaoamlsnkptpnzi) |
| 08:00.42 | *** join/#edev slashbeast (~slashbeas@funtoo/core/slashbeast) |
| 08:48.48 | *** join/#edev zhangxaochen (~zhangxaoc@115.195.43.150) |
| 08:57.16 | *** join/#edev liuyq (~liuyq0307@221.123.128.226) |
| 09:02.08 | *** join/#edev florian_kc (~fuchs@Maemo/community/contributor/florian) |
| 10:05.58 | *** join/#edev g1powermac (~g1powerma@unaffiliated/g1powermac) |
| 10:30.09 | *** join/#edev nullpuppy (~dustin@c-50-135-12-17.hsd1.wa.comcast.net) |
| 12:05.32 | *** join/#edev zhangxaochen (zhangxaoch@222.205.99.36) |
| 12:42.00 | *** join/#edev gustavoz (~gustavoz@host107.190-139-151.telecom.net.ar) |
| 12:48.30 | RushPL_ | guys, I have one Hynix nandflash ATR - it works beautifuly even without ECC, straight from the factory but on the other BTR I get lots lots lots of single bit errors. I tried several ATR and BTR chips and the same behaviour .. is it normal behaviour? |
| 12:55.48 | SpeedEvil | Nand flash is supposed to have errors in general |
| 12:55.55 | SpeedEvil | other than on the first block |
| 12:56.19 | SpeedEvil | It's specified capacity is something like x correctable blocks after y writes |
| 12:56.48 | av500 | why should the 1st block not have errors? |
| 12:57.12 | SpeedEvil | Because it's (often) specified not to. |
| 12:57.19 | SpeedEvil | Maybe not always |
| 12:57.40 | av500 | e.g. TI OMAP bootloaders try several blocks |
| 12:57.42 | SpeedEvil | at least sometimes - the several I've looked at have said the first block is guaranteed error free for the first thousand writes |
| 12:57.45 | av500 | up to 4 until one works |
| 12:57.57 | av500 | to counter a broken sector 0 |
| 12:57.58 | SpeedEvil | As bootloaders can often not cope with errors |
| 12:58.06 | av500 | but other nands might do that internally |
| 12:58.12 | SpeedEvil | Sure. I'm talking about NAND, not what others can cope with |
| 13:04.42 | *** join/#edev GPSFan (~kenm@64.92.145.112) |
| 13:04.42 | *** mode/#edev [+o GPSFan] by ChanServ |
| 13:16.33 | RushPL | SpeedEvil: well, actually what I am getting is something like this http://wklej.org/hash/81c9642d2b9/ |
| 13:17.05 | SpeedEvil | If you're getting uncorrectable errors - that's something different |
| 13:17.14 | SpeedEvil | Assuming you're following the appropriate ECC |
| 13:17.58 | RushPL | SpeedEvil: I have only 1 bit hardware correction for each 64 bytes chunk .. |
| 13:19.33 | RushPL | but if I'm getting 600 bit errors that wouldn't be correctable by even very good ECC (datasheet says that 24 bit correction for 1024 bytes chunks would allow for 3000 erase program cycles) |
| 13:20.04 | SpeedEvil | From memory (which may be suffering parity errors) - there is a specified ECC method. |
| 13:20.21 | SpeedEvil | The devices are supposed to work, if you follow this. |
| 13:20.27 | SpeedEvil | Rolling your own may not work well. |
| 13:21.37 | *** join/#edev bizhanMona (~bizhan@nat/cisco/x-jyxfslzmdkfkofif) |
| 13:21.52 | SpeedEvil | It's been a while since I read the specs though |
| 13:21.56 | RushPL | yep that's why I am counting single bit errors. |
| 13:22.06 | RushPL | to see if a better correction would fix it |
| 13:22.20 | RushPL | but I would still lost around 30% of pages |
| 13:22.24 | RushPL | *lose |
| 13:22.54 | RushPL | could there be any power issues I am not aware of that could cause a nand flash to behave in a bad way? |
| 13:46.52 | SpeedEvil | What sort of error correction are you doing? REed solomon, or ...? |
| 14:21.06 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 15:07.27 | sjhill | Anyone see Peter K. recently? |
| 15:07.33 | sjhill | aka Jacmet |
| 15:08.55 | gustavoz | sjhill: try in #uclibc or #buildroot |
| 15:16.13 | *** join/#edev prpplague (~danders@192.94.92.11) |
| 15:27.17 | *** join/#edev nullpuppy (~dustin@freematrix/staff/nullpuppy) |
| 17:44.37 | *** join/#edev Posterdati (~tapioca@host113-237-dynamic.6-87-r.retail.telecomitalia.it) |
| 18:02.57 | *** join/#edev florian (~fuchs@sign-4d091d47.pool.mediaWays.net) |
| 18:02.57 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 18:10.50 | *** mode/#edev [+o prpplague] by ChanServ |
| 18:43.46 | *** join/#edev ccole (~cole@coledd.com) |
| 19:21.39 | *** join/#edev bizhanMona (~bizhan@nat/cisco/x-elhzlpozalkqwgtu) |
| 19:56.49 | *** join/#edev jnd (~jenda@unaffiliated/jnd) |
| 21:05.36 | *** join/#edev devxdev (~devxdev@c-67-171-181-60.hsd1.or.comcast.net) |
| 21:10.35 | *** join/#edev gustavoz (~gustavoz@host128.190-231-142.telecom.net.ar) |
| 21:11.01 | *** join/#edev jnd (~jenda@unaffiliated/jnd) |
| 21:12.52 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 21:52.56 | *** join/#edev Matth_FA (~Matth_FA@78.193.124.109) |
| 23:48.46 | *** join/#edev GPSFan (~kenm@64.92.145.112) |
| 23:48.47 | *** mode/#edev [+o GPSFan] by ChanServ |