| 00:22:32 | | Quit Piece_Maker (Quit: ZNC 1.8.2 - https://znc.in) |
| 00:24:12 | | Join Piece_Maker [0] (~eddie@bolt-17-b2-v4wan-168143-cust4.vm18.cable.virginm.net) |
| 01:00 |
| 01:06:59 | | Quit othello7 (Quit: othello7) |
| 01:10:51 | | Join _bilgus_ [0] (~bilgus@162-154-213-134.res.spectrum.com) |
| 01:14:20 | | Quit _bilgus (Ping timeout: 268 seconds) |
| 01:46:56 | *** | Saving seen data "./dancer.seen" |
| 02:00 |
| 02:27:48 | | Quit CH23_M (Read error: Connection reset by peer) |
| 02:27:59 | | Join CH23_M [0] (~CH23@revspace/participant/ch23) |
| 03:00 |
| 03:46:58 | *** | Saving seen data "./dancer.seen" |
| 05:00 |
| 05:14:08 | | Quit Malinux (Ping timeout: 260 seconds) |
| 05:30:10 | | Join f_ [0] (~AUGESOUND@fases/developer/funderscore) |
| 05:41:52 | | Join Malinux [0] (~malin@2001:4641:4dfa::12c:c4a7) |
| 05:47:02 | *** | Saving seen data "./dancer.seen" |
| 05:57:24 | | Join jacobk [0] (~quassel@64.189.201.150) |
| 06:00 |
| 06:16:01 | | Join Malinux- [0] (~malin@2001:4641:4dfa::12c:c4a7) |
| 06:17:55 | | Quit Malinux (Ping timeout: 246 seconds) |
| 06:41:27 | | Join sebagala [0] (~Burak@185.25.123.34) |
| 07:00 |
| 07:47:04 | *** | Saving seen data "./dancer.seen" |
| 08:00 |
| 08:58:30 | | Quit f_ (Ping timeout: 260 seconds) |
| 09:00 |
| 09:08:54 | | Join othello7 [0] (~Thunderbi@pool-100-36-176-164.washdc.fios.verizon.net) |
| 09:15:52 | speachy | chris_s: let me know what target you are willing to sacrifice to my experimental builds. IMO it'll probably be immediately obvious. |
| 09:47:08 | *** | Saving seen data "./dancer.seen" |
| 10:00 |
| 10:32:36 | | Join chris_s [0] (~chris_s@ip-176-199-253-249.um44.pools.vodafone-ip.de) |
| 10:36:49 | chris_s | speachy: either one is fine (ipod4g,22 or ipodvideo,24); maybe preferably the 5g, since that one seemed slightly more picky in the past when it comes to SD/iFlash compatibility. If necessary or useful I can confirm with the other one at a later point. |
| 10:37:04 | speachy | ok. |
| 10:37:59 | chris_s | oops, mixed up the numbers: ipod4g,24. / ipodvideo,22 |
| 10:39:57 | speachy | https://www.shaftnet.org/~pizza/rockbox-ipod5g-exp.zip |
| 10:40:32 | chris_s | getting 403 |
| 10:40:54 | speachy | fixed |
| 10:41:11 | chris_s | yup, thx |
| 10:49:44 | speachy | this changes the powerdown behavior (turns it on, increases delay for SSDs), turns on write dma for SSDs, simplifies the sleep logic a bit... |
| 10:52:24 | speachy | seems to work on the mini2g (with the og microdrive) so far |
| 10:53:07 | chris_s | playing some music and letting the db build now |
| 10:54:20 | chris_s | wasn't write dma turned off because it didn't actually increase performance, or am I remembering that wrong from a comment somwhere in the code? |
| 10:54:53 | speachy | so claimed the comment |
| 10:55:23 | speachy | I have a forum-sourced patch that claims to make things work better for tehse things; one of the changes it made was turn DMA writes on. |
| 10:55:51 | speachy | there's a lot of question about PIO timings so I figured DMA might be safer, even if it's not necessarily faster. |
| 10:55:58 | chris_s | ah ok |
| 10:57:50 | chris_s | looks like it's hanging now while building the db and doesn't react to input anymore, although the music is still playing at this point |
| 10:57:57 | chris_s | ah no false alarm |
| 10:57:59 | speachy | plus let's face it, these things are all far faster than the OG drives could ever hope of being. |
| 10:58:01 | chris_s | continued now |
| 10:58:45 | speachy | logf is enabled but I didn't turn on any specifics in this build. |
| 11:00 |
| 11:05:30 | | Quit rogeliodh (Quit: The Lounge - https://thelounge.chat) |
| 11:05:51 | | Join rogeliodh [0] (~rogeliodh@rogeliodh.dev) |
| 11:40:33 | | Quit Malinux- (Ping timeout: 268 seconds) |
| 11:47:10 | *** | Saving seen data "./dancer.seen" |
| 11:53:02 | chris_s | so far so good. I'll test some more, but nothing that immediately turns up |
| 12:00 |
| 12:01:39 | chris_s | seeing some image corruption in theme elements now |
| 12:02:55 | chris_s | and buflib *PANIC* "block len wacky" |
| 12:03:25 | chris_s | never seen that before |
| 12:06:25 | chris_s | database is corrupted |
| 12:07:16 | speachy | so much to that. |
| 12:13:59 | speachy | Hmm. it's probably the poweroff that's causing the problems. |
| 12:30:51 | speachy | chris_s: I have a new build up, |
| 12:38:30 | chris_s | checking it out now. I'm stepping away from the keyboard for an hour or two, but taking the iPod with me and will report back later |
| 12:44:57 | speachy | ok. I will have a third one shortly. |
| 13:00 |
| 13:07:04 | speachy | 3rd build up now. This is one I feel much more confident with. |
| 13:08:02 | speachy | short version: if device doesn't claim to support APM, _never_ kill power or even issue a standby_immediate command. |
| 13:08:21 | | Quit chris_s (Ping timeout: 256 seconds) |
| 13:08:28 | speachy | none of the SD/CF adapters seem to claim support for this |
| 13:08:53 | speachy | the one dump I have of a SATA SSD claims to support it, so we'll see how that shakes out |
| 13:41:15 | | Join lebellium [0] (~lebellium@2a01cb0405d07f00a3fad27af4baf24a.ipv6.abo.wanadoo.fr) |
| 13:47:12 | *** | Saving seen data "./dancer.seen" |
| 13:51:28 | speachy | I do have a 5.5g and the iflash-sata adapter that I need to assemble still |
| 13:54:29 | | Quit jacobk (Ping timeout: 240 seconds) |
| 13:58:37 | | Join chris_s [0] (~chris_s@ip-176-199-253-249.um44.pools.vodafone-ip.de) |
| 14:00 |
| 14:01:41 | chris_s | speachy: I wonder how the OF deals with this? Still testing the 2nd one. An initial database build didn't "take" which didn't inspire confidence but I can't rule out that this was caused by interacting with the player during the build in some way. Other than that, it's behaving like the first build at this point, so probably too early to tell |
| 14:02:15 | speachy | the OF barely writes anything at all |
| 14:02:42 | speachy | and the OF might also properly respect the APM flags and not try to put the drive to sleep or yank power |
| 14:03:19 | chris_s | right |
| 14:08:26 | | Join jacobk [0] (~quassel@utdpat242060.utdallas.edu) |
| 14:09:10 | speachy | the only possible regression I see here is for really old ipods (<4g) or the irivers with their original drives |
| 14:09:17 | chris_s | " if device doesn't claim to support APM, _never_ kill power or even issue a standby_immediate command." <- is that how it works in the current dev builds for iFlash ? |
| 14:09:37 | speachy | no, the current dev builds issue the standby_immediate command |
| 14:09:50 | speachy | and but never disable power due to an inverted test |
| 14:10:43 | speachy | the potential regression is if those ancient drives don't support APM either. |
| 14:11:27 | speachy | the 2nd iteration of that patch still potentially yanks power. |
| 14:11:40 | speachy | the 3rd iteration disables all of that altogether |
| 14:12:52 | speachy | I'd prefer to prevent our internal state from ever entering "sleep" (for hardwar that doesn't support it) but I've seen notes that the hardware goes to sleep when idle no matter what, so we'd need to "wake it up" safely after long idle periods anyway. |
| 14:14:25 | chris_s | i see |
| 14:15:06 | speachy | basically those CF->SD adapters leave a lot to be desired |
| 14:16:04 | speachy | ...the SATA-based ones are at least properly ATA-compliant. the standby_immediate command is supposed to put the device into a state where it's safe to remove power. |
| 14:16:49 | speachy | (the SD stuff _breaks_ when you send it the SLEEP command, but it accepts STANDBY_IMMEDIATE) |
| 14:17:03 | chris_s | :/ |
| 14:17:51 | speachy | so if old drives don't handle the latter, I'd have to put SLEEP back in, which means some sort of ssd-detection heuristics have to go back in...) |
| 14:27:55 | | Quit CH23_M (Read error: Connection reset by peer) |
| 14:28:34 | | Join CH23_M [0] (~CH23@revspace/participant/ch23) |
| 14:43:33 | | Join Malinux [0] (~malin@2001:4641:4dfa::12c:c4a7) |
| 15:00 |
| 15:29:48 | dook | chris_s apologies, only got back online now. Saw on reddit you managed to sort the issue out :) |
| 15:34:06 | chris_s | dook: np, thanks for directing me to it :) |
| 15:34:13 | chris_s | speachy: 2nd one (ee1f67) still chugging along nicely. Haven't been able to trip it up so far. |
| 15:35:07 | speachy | oh nice. |
| 15:47:14 | *** | Saving seen data "./dancer.seen" |
| 16:00 |
| 16:09:35 | | Quit jacobk (Ping timeout: 260 seconds) |
| 16:38:12 | speachy | it's now g#5648 btw |
| 16:38:15 | rb-bluebot | Gerrit review #5648 at https://gerrit.rockbox.org/r/c/rockbox/+/5648 : pp5020: Use DMA Writes by default for SSDs by Solomon Peachy |
| 16:38:24 | speachy | er, g#5647 |
| 16:38:27 | rb-bluebot | Gerrit review #5647 at https://gerrit.rockbox.org/r/c/rockbox/+/5647 : ata: Heavily rework sleep and poweroff logic by Solomon Peachy |
| 17:00 |
| 17:37:46 | chris_s | ok great, i'll be putting it on the 4g as well |
| 17:41:28 | speachy | that third build I uploaded is master + those two patches |
| 17:42:06 | speachy | just put rockbox on the ipod5.5g I got a while back, still need to swap out the drive though |
| 17:43:03 | speachy | frustratingly, no sound out of the headphone jack. I wonder if the ebay refurbisher didn't hook it up properly ... (IIRC it's on a separate board connected via a ribbon cable.. |
| 17:45:21 | chris_s | :/ ugh, does the hold switch work? https://www.ifixit.com/Guide/iPod+5th+Generation+(Video)+Headphone+Jack+&+Hold+Switch+Replacement/604 |
| 17:45:30 | speachy | yeah |
| 17:46:32 | speachy | I also have real CF cards and the 1st-gen iFlash CF->SD adapter I can swap into the mini2g |
| 17:47:18 | *** | Saving seen data "./dancer.seen" |
| 17:49:27 | speachy | but I won't have the bandwidth to tackle either until next week at the earliest. |
| 17:49:32 | speachy | hmm, no audio in this dock either |
| 17:51:39 | | Quit lebellium (Quit: Leaving) |
| 18:00 |
| 18:05:06 | speachy | really should have tested this thing when I bought it... 16 months ago |
| 18:05:39 | chris_s | :D no audio kind of takes the fun out of it |
| 18:07:27 | chris_s | weird that neither line out nor headphone jack work |
| 18:07:51 | speachy | yeah.. hopefully it's something obvious on the PCB. |
| 18:08:46 | speachy | but at least that won't matter for purposes of trying to test for (a lack of) disk corruption with sata adapters |
| 18:09:24 | | Quit othello7 (Ping timeout: 252 seconds) |
| 18:24:08 | | Join Moriar [0] (~moriar@107-200-193-159.lightspeed.stlsmo.sbcglobal.net) |
| 19:00 |
| 19:06:21 | | Join massiveH [0] (~massiveH@2600:4040:a982:c800:c8e8:e496:7b0c:c1fb) |
| 19:09:56 | | Quit chris_s (Quit: Connection closed) |
| 19:47:19 | *** | Saving seen data "./dancer.seen" |
| 20:00 |
| 20:14:29 | Bubblegumdrop | hello |
| 20:14:42 | Bubblegumdrop | speachy thanks! |
| 20:15:02 | Bubblegumdrop | https://phntsm.ddns.net/log/2024-02-27-ipod-video-iflash-msata-ssd-rockbox.html fwiw |
| 20:17:04 | Bubblegumdrop | I will try the latest builds |
| 21:00 |
| 21:32:35 | | Quit cstine (Quit: The Lounge - https://thelounge.chat) |
| 21:33:10 | | Join cstine [0] (~cstine@150.136.136.191) |
| 21:41:02 | | Quit Moriar (Quit: Leaving.) |
| 21:47:24 | *** | Saving seen data "./dancer.seen" |
| 22:00 |
| 22:26:57 | | Join jacobk [0] (~quassel@129.110.242.173) |
| 23:00 |
| 23:47:25 | *** | No seen item changed, no save performed. |
icon identifies that the person is a core developer (has commit access).