#
#asahi-dev · 2026-09-01 10:55
sofusi meant nv15
#asahi-dev · 2026-09-01 10:55
sofus[m]uploaded an image: (26KiB) < https://matrix.org/oftc/media/v1/media/download/AVvjsqlI9ZiILMizaC9un18RXK_qtpyiD3IS8spRCv2eEsexCDknNMh26ZT7aYmkS704OisUipjtw5tVfsEXEmBCegpLVzVAAG1hdHJpeC5vcmcvS2JMRW5YUUdqUVpOR2VTeW1SV3BRVEZX >
#asahi-dev · 2026-09-01 10:55
JamesCalligeros[m]nv24? that's the 4:4:4 one
#asahi-dev · 2026-09-01 10:51
sofusmaybe its nv20 or whatever its called
#asahi-dev · 2026-09-01 10:51
sofusoh!
#asahi-dev · 2026-09-01 10:51
chaos_princessthe stripes do suggest that it is reading the padding bits and interpreting them as something else
#asahi-dev · 2026-09-01 10:50
sofusim just trying to figure out what a flag does
#asahi-dev · 2026-09-01 10:48
sofusp010
#asahi-dev · 2026-09-01 10:48
chaos_princesswhat is the bit format?
#asahi-dev · 2026-09-01 10:48
chaos_princessreinterpreting 10-bit data as 8-bit?
#asahi-dev · 2026-09-01 10:47
sofusits 256x256 10 bit red, the 8 bit one looks fine
#asahi-dev · 2026-09-01 10:47
sofusdoes anyone know why it could look likes this?
#asahi-dev · 2026-09-01 10:46
sofus[m]uploaded an image: (8KiB) < https://matrix.org/oftc/media/v1/media/download/AfiG0Re3yc-sKZTX-BlhjA-bQGbFNdEC_I2py497LrKhfoxNE8t_WYzgCJIqZKrDsqcUwgCBaG6hI1FtPs-ylvNCegpK09QgAG1hdHJpeC5vcmcvQklNdFd3eGtVd2tlUVNnS1hhblpxTGFv >
#asahi-dev · 2026-09-01 07:25
djthat doc doesn't look like it would get you anywhere either way. speaking of which, I just got sbio talking to mesa for quite a bit before crashing :D
#asahi-dev · 2026-09-01 06:05
svenyou also absolutely do not want a "linux implementation roadmap" with pseudo code in there
#asahi-dev · 2026-09-01 05:57
svenwhich is one of the reasons for that strict LLM policy we have
#asahi-dev · 2026-09-01 05:57
svenbut they are very good at translating between languages, like assembly and english, which is why those attempts always end up like yours unfortunately
#asahi-dev · 2026-09-01 05:56
svenLLMs aren't very good at this
#asahi-dev · 2026-09-01 05:56
httperry[m]Ok
#asahi-dev · 2026-09-01 05:55
sven*not
#asahi-dev · 2026-09-01 05:55
svenno how XNU drives it, and certainly no references to "verified virtual function addresses" and similar
#asahi-dev · 2026-09-01 05:54
svenfor clean-room documentation you want to describe _behaviur_ of the hardware or firmware interface
#asahi-dev · 2026-09-01 05:54
httperry[m]Alr
#asahi-dev · 2026-09-01 05:54
svenlooking at those docs is no better than disassembling the binary yourself and quite worse, those docs claim to be clean-room when they clearly aren't
#asahi-dev · 2026-09-01 05:53
sventhis is not helpful
#asahi-dev · 2026-09-01 05:53
svenwhich brings us back to https://asahilinux.org/slop/ and the binary re policy
#asahi-dev · 2026-09-01 05:53
httperry[m]Yeah I made it add that
#asahi-dev · 2026-09-01 05:53
sventhis is as tainted as looking at the binary yourself
#asahi-dev · 2026-09-01 05:53
svenas i said, there are tons of references to disassmebly in there
#asahi-dev · 2026-09-01 05:52
httperry[m]:despair:
#asahi-dev · 2026-09-01 05:52
svenno, you obviously didn't.
#asahi-dev · 2026-09-01 05:52
httperry[m]I mean i verified them manually
#asahi-dev · 2026-09-01 05:51
sveni opened a random .md file and there are tons of references to disassembled code which is precisely what you don't want
#asahi-dev · 2026-09-01 05:50
svena) https://asahilinux.org/slop/ and b) it's _very_ obviously not clean room despite the LLM claiming that confidently
#asahi-dev · 2026-09-01 05:42
httperry[m]github.com/httperry/QuickTouch/
#asahi-dev · 2026-09-01 05:42
httperry[m]Hi, I've tried to reverse engineer the SEP communication between the OS and the SEPOS and extracted a table of opcodes and sequences to initialise it properly. I've decompiled the apple's kernel code and manually extracted the endpoints, can anyone verify them?
#asahi-dev · 2026-08-31 21:11
JamesCalligeros[m]B*
#asahi-dev · 2026-08-31 21:11
JamesCalligeros[m]sofus: it should be 128b aligned aiui
#asahi-dev · 2026-08-31 19:50
yukathere are still a lot of differences in the trace which I can try to apply but I'm done for today
#asahi-dev · 2026-08-31 19:49
yukalowering the max_queue_size to 63 results in the same errors at size: 64, tail/head: 63/62
#asahi-dev · 2026-08-31 19:49
yukasorry about the misinfo earlier, macos sets MAX_PENDS to 0x400040 on both t8132 an t8140
#asahi-dev · 2026-08-31 18:30
sofusi also wanna test the alignment, i think its 128 or 256 for avd
#asahi-dev · 2026-08-31 18:30
sofuschadmed: re meta_tile order i think its actually the same, and as expected it still works
#asahi-dev · 2026-08-31 18:27
yukawhy is size 65 though
#asahi-dev · 2026-08-31 18:27
svenwait.. with size 65 that should be okay
#asahi-dev · 2026-08-31 18:26
sventhat looks wrong, yes
#asahi-dev · 2026-08-31 18:26
yukais the tail 64 maybe the problem?
#asahi-dev · 2026-08-31 18:25
yukaI/O CQ : 0x1001880c000, size: 65, tail/head: 64/63
#asahi-dev · 2026-08-31 17:59
yukais the full error when writing
#asahi-dev · 2026-08-31 17:59
yuka"assert failed: [7454]:CQ (Host I/O) DB error, status_reg: 0x4, head: 63, valid_status: 0x4, err_info_0: 0x1, err_info_1: 0x20000"

plus ›