IRC log of #maemo-ssu for Saturday, 2016-07-09

*** xes_ has joined #maemo-ssu00:09
*** xes has quit IRC00:12
*** trx has quit IRC00:19
*** delphi has joined #maemo-ssu00:19
*** delphi has quit IRC00:28
*** trx has joined #maemo-ssu00:33
*** trx has joined #maemo-ssu00:33
*** xes_ is now known as xes01:08
*** LauRoman|Alt has quit IRC01:11
*** LauRoman|Alt has joined #maemo-ssu02:31
*** LauRoman has joined #maemo-ssu02:32
*** LauRoman|Alt has quit IRC02:35
*** LauRoman has quit IRC03:16
*** drrz has joined #maemo-ssu03:17
*** drrrz has quit IRC03:20
*** drrrz has joined #maemo-ssu03:20
*** drrz has quit IRC03:25
*** drrrz has quit IRC03:26
*** freemangordon has quit IRC04:16
*** freemangordon has joined #maemo-ssu04:18
*** APic has quit IRC05:04
*** DocScrutinizer05 has quit IRC07:00
*** DocScrutinizer05 has joined #maemo-ssu07:00
*** chainsawbike has quit IRC08:12
*** chainsawbike has joined #maemo-ssu08:13
*** peetah has quit IRC11:10
*** Pali has joined #maemo-ssu12:51
*** peetah has joined #maemo-ssu13:30
*** dafox has joined #maemo-ssu14:30
*** macmaN has joined #maemo-ssu15:19
*** APic has joined #maemo-ssu18:17
*** APic has quit IRC18:21
*** APic has joined #maemo-ssu18:21
*** APic has quit IRC18:22
*** APic has joined #maemo-ssu18:23
*** handaxe has joined #maemo-ssu22:19
*** handaxe has quit IRC22:24
*** DrCode has quit IRC22:26
DocScrutinizer05RFC: backport TRIM ioctl cmd support to MMC driver22:32
DocScrutinizer05rationale: particularly for swap volume a TRIM will significantly improve performance22:33
kerioi don't think TRIM is a thing, for microSDs22:33
DocScrutinizer05I'm talking about mmc22:34
kerioisn't it basically just a big SD card anyway22:34
DocScrutinizer05though TRIM even exists in ATA22:34
keriothe issue is that the controller needs to understand it22:34
DocScrutinizer05well, that's why TRIM got standardized22:34
kerioyeah but22:34
kerioi don't think the one in the N900 understands it22:35
DocScrutinizer05subject to evaluation: can *N900* eMMC understand TRIM22:35
keriocan the neo900 eMMC understand TRIM?22:35
DocScrutinizer05https://en.wikipedia.org/wiki/Trim_(computing)#SD.2FMMC22:37
keriook so it's still a thing22:37
DocScrutinizer05http://datasheet.octopart.com/THGBM1G8D8EBAI2-Toshiba-datasheet-11973698.pdf22:48
DocScrutinizer05Full compliance w ith JEDEC/MMCA Ver. 4.322:48
DocScrutinizer05https://en.wikipedia.org/wiki/MultiMediaCard#Open_standard22:54
kerioaww, 4.5 is the one with the proper TRIM22:55
*** handaxe has joined #maemo-ssu22:56
DocScrutinizer05yes, with even better TRIM, but 4.4 and prolly 4.3 already support something similar23:00
*** handaxe has quit IRC23:01
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.2FMMC23:02
kerioyeah but in addition to supporting the command, the controller should also know how to use it in a smart way, shouldn't it23:05
kerioand since it's likely not very used, we'd need some testing done on it23:05
DocScrutinizer05ohmy, you realy think a command is implemented in a chip in a way so it returns no error and does nothing else?23:13
DocScrutinizer05I doubt JEDEC would agree on such implementation23:13
*** DrCode has joined #maemo-ssu23:16
kerioDocScrutinizer05: turn ERASE into "write 0"23:17
keriothrough the normal channels23:17
DocScrutinizer05aha23:17
DocScrutinizer05and then?23:17
kerioand then you have an implementation of ERASE23:17
keriothat doesn't help at all23:17
DocScrutinizer05nonsense23:17
keriobut it's valid23:17
kerioyou're underestimating the hardware manufacturers' cheapness23:18
DocScrutinizer05there'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 zeroes23:18
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 why23:26
DocScrutinizer05prolly 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 write23:28
DocScrutinizer05my SSD does for sure, writing all zero files I get write speeds at 600MB/s23:29
DocScrutinizer05sustainable23:29
kerioDocScrutinizer05: writing zeroes on my ssd yields 650MB/s23:33
keriowhich is the normal write speed23:33
DocScrutinizer05sure, please tell me which SSD you got23:35
DocScrutinizer05a SSD that can do a sustained write speed of 650MB/s on random data - that thing must be a monster23:36
DocScrutinizer05It also clearly proves that we need better interfaces since 650MB/s is the theoretical max of a 6G SATA23:37
DocScrutinizer05well, ok, fastest SSDs seem to be near 512MB/s23:45
DocScrutinizer05http://ssd.userbenchmark.com/Patriot-Ignite-960GB/Rating/357523:46
DocScrutinizer05nah, we already got better interfaces: SATA Express 8 Gbit/s and 16 Gbit/s23:54
keriothis is a PCIe SSD, yeah23:57
keriowhat are you talking about23:58
DocScrutinizer05TRIM was the topic23:59
keriothe intel 750 gets up to 1200MB/s of sequential writes23:59

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!