| *** xes_ has joined #maemo-ssu | 00:09 | |
| *** xes has quit IRC | 00:12 | |
| *** trx has quit IRC | 00:19 | |
| *** delphi has joined #maemo-ssu | 00:19 | |
| *** delphi has quit IRC | 00:28 | |
| *** trx has joined #maemo-ssu | 00:33 | |
| *** trx has joined #maemo-ssu | 00:33 | |
| *** xes_ is now known as xes | 01:08 | |
| *** LauRoman|Alt has quit IRC | 01:11 | |
| *** LauRoman|Alt has joined #maemo-ssu | 02:31 | |
| *** LauRoman has joined #maemo-ssu | 02:32 | |
| *** LauRoman|Alt has quit IRC | 02:35 | |
| *** LauRoman has quit IRC | 03:16 | |
| *** drrz has joined #maemo-ssu | 03:17 | |
| *** drrrz has quit IRC | 03:20 | |
| *** drrrz has joined #maemo-ssu | 03:20 | |
| *** drrz has quit IRC | 03:25 | |
| *** drrrz has quit IRC | 03:26 | |
| *** freemangordon has quit IRC | 04:16 | |
| *** freemangordon has joined #maemo-ssu | 04:18 | |
| *** APic has quit IRC | 05:04 | |
| *** DocScrutinizer05 has quit IRC | 07:00 | |
| *** DocScrutinizer05 has joined #maemo-ssu | 07:00 | |
| *** chainsawbike has quit IRC | 08:12 | |
| *** chainsawbike has joined #maemo-ssu | 08:13 | |
| *** peetah has quit IRC | 11:10 | |
| *** Pali has joined #maemo-ssu | 12:51 | |
| *** peetah has joined #maemo-ssu | 13:30 | |
| *** dafox has joined #maemo-ssu | 14:30 | |
| *** macmaN has joined #maemo-ssu | 15:19 | |
| *** APic has joined #maemo-ssu | 18:17 | |
| *** APic has quit IRC | 18:21 | |
| *** APic has joined #maemo-ssu | 18:21 | |
| *** APic has quit IRC | 18:22 | |
| *** APic has joined #maemo-ssu | 18:23 | |
| *** handaxe has joined #maemo-ssu | 22:19 | |
| *** handaxe has quit IRC | 22:24 | |
| *** DrCode has quit IRC | 22:26 | |
| DocScrutinizer05 | RFC: backport TRIM ioctl cmd support to MMC driver | 22:32 |
|---|---|---|
| DocScrutinizer05 | rationale: particularly for swap volume a TRIM will significantly improve performance | 22:33 |
| kerio | i don't think TRIM is a thing, for microSDs | 22:33 |
| DocScrutinizer05 | I'm talking about mmc | 22:34 |
| kerio | isn't it basically just a big SD card anyway | 22:34 |
| DocScrutinizer05 | though TRIM even exists in ATA | 22:34 |
| kerio | the issue is that the controller needs to understand it | 22:34 |
| DocScrutinizer05 | well, that's why TRIM got standardized | 22:34 |
| kerio | yeah but | 22:34 |
| kerio | i don't think the one in the N900 understands it | 22:35 |
| DocScrutinizer05 | subject to evaluation: can *N900* eMMC understand TRIM | 22:35 |
| kerio | can the neo900 eMMC understand TRIM? | 22:35 |
| DocScrutinizer05 | https://en.wikipedia.org/wiki/Trim_(computing)#SD.2FMMC | 22:37 |
| kerio | ok so it's still a thing | 22:37 |
| DocScrutinizer05 | http://datasheet.octopart.com/THGBM1G8D8EBAI2-Toshiba-datasheet-11973698.pdf | 22:48 |
| DocScrutinizer05 | Full compliance w ith JEDEC/MMCA Ver. 4.3 | 22:48 |
| DocScrutinizer05 | https://en.wikipedia.org/wiki/MultiMediaCard#Open_standard | 22:54 |
| kerio | aww, 4.5 is the one with the proper TRIM | 22:55 |
| *** handaxe has joined #maemo-ssu | 22:56 | |
| DocScrutinizer05 | yes, with even better TRIM, but 4.4 and prolly 4.3 already support something similar | 23:00 |
| *** handaxe has quit IRC | 23:01 | |
| DocScrutinizer05 | >>The MultiMediaCard and SD ERASE (CMD38) command provides similar functionality to the ATA TRIM command, although it requires that erased blocks be overwritten with either zeroes or ones<< https://en.wikipedia.org/wiki/Trim_(computing)#SD.2FMMC | 23:02 |
| kerio | yeah but in addition to supporting the command, the controller should also know how to use it in a smart way, shouldn't it | 23:05 |
| kerio | and since it's likely not very used, we'd need some testing done on it | 23:05 |
| DocScrutinizer05 | ohmy, you realy think a command is implemented in a chip in a way so it returns no error and does nothing else? | 23:13 |
| DocScrutinizer05 | I doubt JEDEC would agree on such implementation | 23:13 |
| *** DrCode has joined #maemo-ssu | 23:16 | |
| kerio | DocScrutinizer05: turn ERASE into "write 0" | 23:17 |
| kerio | through the normal channels | 23:17 |
| DocScrutinizer05 | aha | 23:17 |
| DocScrutinizer05 | and then? | 23:17 |
| kerio | and then you have an implementation of ERASE | 23:17 |
| kerio | that doesn't help at all | 23:17 |
| DocScrutinizer05 | nonsense | 23:17 |
| kerio | but it's valid | 23:17 |
| kerio | you're underestimating the hardware manufacturers' cheapness | 23:18 |
| DocScrutinizer05 | there's clearly a difference on all levels of device management between a released unused (erased/TRIMed) block and a block that holds valid and valuable data even when that data consists of all zeroes | 23:18 |
| DocScrutinizer05 | >>There are different types of TRIM defined by SATA ...<< Non-deterministic TRIM, Deterministic TRIM (DRAT), Deterministic Read Zero after TRIM (RZAT). I guess >> MultiMediaCard and SD ERASE (CMD38) command provides similar functionality to the ATA TRIM<< RZAT, that's why | 23:26 |
| DocScrutinizer05 | prolly the controler is even smarter than you'd hope, by actually sensing the all 0 write and transforming it into a special shortcut that doesn't need any real write | 23:28 |
| DocScrutinizer05 | my SSD does for sure, writing all zero files I get write speeds at 600MB/s | 23:29 |
| DocScrutinizer05 | sustainable | 23:29 |
| kerio | DocScrutinizer05: writing zeroes on my ssd yields 650MB/s | 23:33 |
| kerio | which is the normal write speed | 23:33 |
| DocScrutinizer05 | sure, please tell me which SSD you got | 23:35 |
| DocScrutinizer05 | a SSD that can do a sustained write speed of 650MB/s on random data - that thing must be a monster | 23:36 |
| DocScrutinizer05 | It also clearly proves that we need better interfaces since 650MB/s is the theoretical max of a 6G SATA | 23:37 |
| DocScrutinizer05 | well, ok, fastest SSDs seem to be near 512MB/s | 23:45 |
| DocScrutinizer05 | http://ssd.userbenchmark.com/Patriot-Ignite-960GB/Rating/3575 | 23:46 |
| DocScrutinizer05 | nah, we already got better interfaces: SATA Express 8 Gbit/s and 16 Gbit/s | 23:54 |
| kerio | this is a PCIe SSD, yeah | 23:57 |
| kerio | what are you talking about | 23:58 |
| DocScrutinizer05 | TRIM was the topic | 23:59 |
| kerio | the intel 750 gets up to 1200MB/s of sequential writes | 23:59 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!