#asahi-dev · 2026-07-16 16:25
sofuswhy was "Heatpipe Power" chosen for PHPC? For my t6020 it seems to be ~6W less than PSTR#asahi-dev · 2026-07-16 14:23
sofusfingers crossed#asahi-dev · 2026-07-16 14:17
sofuscould it be connected to something else that you dont start?#asahi-dev · 2026-07-16 14:09
sofuswhich what?#asahi-dev · 2026-07-16 14:02
sofusbut they looked ok when macos used it? #asahi-dev · 2026-07-16 13:57
sofuswhat it the sac controller? Does it go by another name in the kernel?#asahi-dev · 2026-07-16 09:59
sofusi think it got merged yesterday#asahi-dev · 2026-07-16 09:34
sofus*the patched ffmpeg#asahi-dev · 2026-07-16 09:34
sofusthis should make ffmpeg work on videos without tiles https://pastebin.com/zWYnUnyy#asahi-dev · 2026-07-16 09:27
sofusffmpeg wont work for hevc at all#asahi-dev · 2026-07-16 09:26
sofusvery much :p Is that possible??#asahi-dev · 2026-07-16 09:24
sofusalthough i dont think anyone tested hevc#asahi-dev · 2026-07-16 09:23
sofushevc, h264 and vp9#asahi-dev · 2026-07-15 21:04
sofusi dont think they are, hmm#asahi-dev · 2026-07-15 20:57
sofusmaybe the offsets are wrong#asahi-dev · 2026-07-15 20:56
sofushere: https://paste.debian.net/hidden/8a5cab9b it passes#asahi-dev · 2026-07-15 20:55
sofuswait#asahi-dev · 2026-07-15 20:55
sofusoh yeah sorry, thought of RPS_E_qualcomm_5#asahi-dev · 2026-07-15 20:53
sofusi would need the output of eiln's tracer to fix this, not sure how many of the failing ones i need it for#asahi-dev · 2026-07-15 20:51
sofusPPS_A_qualcomm_7 fails on all#asahi-dev · 2026-07-15 20:42
sofuslooks like im passing both chroma/luma bitdepth#asahi-dev · 2026-07-15 20:33
sofusi never actually tested it myself#asahi-dev · 2026-07-15 20:32
sofuseiln blocked luma/chroma mismatch for h264, not sure if that translates to hevc#asahi-dev · 2026-07-15 20:30
sofussamples missing on macos side of things?#asahi-dev · 2026-07-15 17:00
sofusjannau: add this to fluster/decoders/ffmpeg.py https://pastebin.com/1iVE4Dua#asahi-dev · 2026-07-15 16:53
sofusi used ffmpeg#asahi-dev · 2026-07-15 13:43
sofusi dont think its a buffer issue. Would be interesting to see if it passes when using macos#asahi-dev · 2026-07-15 13:23
sofusheres my results: https://pastebin.com/i5uwvSbJ and macos for reference: https://pastebin.com/duWpLQPw#asahi-dev · 2026-07-15 13:21
sofusthis could just be an buffer issue#asahi-dev · 2026-07-15 13:21
sofuswould you mind posting the errors also? Thanks#asahi-dev · 2026-07-15 13:16
sofusthats weird, would you mind posting the results?#asahi-dev · 2026-07-15 13:05
sofusits *actually* added as fixup now#asahi-dev · 2026-07-15 13:02
sofusoh sorry i forgot to add it :p#asahi-dev · 2026-07-15 12:52
sofussorry again, and thanks for testing!#asahi-dev · 2026-07-15 12:51
sofussorry, im to fast#asahi-dev · 2026-07-15 12:50
sofusoh#asahi-dev · 2026-07-15 12:48
sofusfound it#asahi-dev · 2026-07-15 12:44
sofusthat should not matter#asahi-dev · 2026-07-15 12:43
sofuswell i must have forgotten something that t8103 needs... #asahi-dev · 2026-07-15 12:42
sofusanything in dmesg?#asahi-dev · 2026-07-15 12:41
sofuswhat soc?#asahi-dev · 2026-07-15 12:40
sofusyes and -j1#asahi-dev · 2026-07-15 12:17
sofusHRD_A_Fujitsu_3 works now#asahi-dev · 2026-07-15 10:04
sofuscurrently i think only gstreamer will work#asahi-dev · 2026-07-15 10:02
sofusneeds testing on t8103, t600x and one of the m3's#asahi-dev · 2026-07-15 10:01
sofuswith RPS_E_qualcomm_5 it could be something like that#asahi-dev · 2026-07-15 10:00
sofusHRD_A_Fujitsu_3 works if we only do the idr_pic check#asahi-dev · 2026-07-15 10:00
sofusim still failing to bitstreams... one of them is because of the is_intra check, im not sure why the other one fails#asahi-dev · 2026-07-15 09:58
sofushevc support https://github.com/AsahiLinux/linux/pull/544#asahi-dev · 2026-07-14 20:08
sofushttps://github.com/AsahiLinux/AsahiLinux.github.io/pull/125