00:10
popolon has quit [Quit: WeeChat 3.0]
00:28
raster has quit [Quit: Gettin' stinky!]
00:34
icecream95 has joined #panfrost
00:44
italov has quit [Ping timeout: 240 seconds]
01:16
italov has joined #panfrost
01:18
icecream95 has quit [Ping timeout: 240 seconds]
01:21
robertfoss has quit [Quit: WeeChat 2.6]
01:46
icecream95 has joined #panfrost
01:49
kaspter has joined #panfrost
01:50
icecream95 has quit [Ping timeout: 240 seconds]
02:35
stikonas has quit [Remote host closed the connection]
02:36
stikonas has joined #panfrost
02:42
italov has quit [Ping timeout: 264 seconds]
02:43
italov has joined #panfrost
02:43
stikonas has quit [Ping timeout: 272 seconds]
03:03
camus has joined #panfrost
03:04
italov has quit [Ping timeout: 240 seconds]
03:05
kaspter has quit [Ping timeout: 264 seconds]
03:05
camus is now known as kaspter
03:17
icecream95 has joined #panfrost
03:42
italov has joined #panfrost
04:08
kaspter has quit [Ping timeout: 256 seconds]
04:09
kaspter has joined #panfrost
04:41
italov has quit [Ping timeout: 260 seconds]
04:59
lucmann has joined #panfrost
05:01
lucmann has quit [Remote host closed the connection]
05:32
davidlt has joined #panfrost
05:46
kaspter has quit [Ping timeout: 246 seconds]
05:47
kaspter has joined #panfrost
06:00
nlhowell has quit [Ping timeout: 256 seconds]
06:04
kinkinkijkin has quit [Remote host closed the connection]
06:09
chewitt has quit [Quit: Adios!]
06:11
lucmann has joined #panfrost
06:18
kinkinkijkin has joined #panfrost
07:28
kaspter has quit [Ping timeout: 246 seconds]
07:28
kaspter has joined #panfrost
07:38
_whitelogger has joined #panfrost
07:54
Elpaulo has joined #panfrost
08:38
icecream95 has quit [Quit: leaving]
08:50
raster has joined #panfrost
09:18
nlhowell has joined #panfrost
09:45
robertfoss has joined #panfrost
09:50
robertfoss has quit [Quit: WeeChat 2.6]
09:56
robertfoss has joined #panfrost
09:58
stikonas has joined #panfrost
10:01
yann has quit [Remote host closed the connection]
10:08
camus has joined #panfrost
10:08
nlhowell has quit [Ping timeout: 260 seconds]
10:09
kaspter has quit [Ping timeout: 260 seconds]
10:09
camus is now known as kaspter
10:10
nlhowell has joined #panfrost
10:15
sphalerite has quit [Quit: reboot time!]
10:20
sphalerite has joined #panfrost
10:34
wiizzard has quit [Quit: Bridge terminating on SIGTERM]
10:34
l-as has quit [Quit: Bridge terminating on SIGTERM]
10:34
Ke has quit [Quit: Bridge terminating on SIGTERM]
10:34
clementp[m] has quit [Quit: Bridge terminating on SIGTERM]
10:34
exit70[m] has quit [Quit: Bridge terminating on SIGTERM]
10:37
nlhowell has quit [Quit: WeeChat 3.0]
10:37
nlhowell has joined #panfrost
10:37
dstzd has joined #panfrost
10:41
wiizzard has joined #panfrost
10:44
dstzd has quit [Read error: Connection reset by peer]
10:45
dstzd has joined #panfrost
10:45
jernej has joined #panfrost
10:48
dstzd has quit [Client Quit]
10:50
jernej has quit [Client Quit]
10:50
jernej has joined #panfrost
10:51
dstzd has joined #panfrost
10:54
italov has joined #panfrost
10:55
Ke has joined #panfrost
10:55
l-as has joined #panfrost
10:55
clementp[m] has joined #panfrost
10:55
exit70[m] has joined #panfrost
11:02
nlhowell has quit [Ping timeout: 256 seconds]
11:02
camus has joined #panfrost
11:04
kaspter has quit [Ping timeout: 256 seconds]
11:04
camus is now known as kaspter
11:25
rando25892 has joined #panfrost
11:36
nlhowell has joined #panfrost
11:38
chewitt has joined #panfrost
11:50
camus has joined #panfrost
11:52
kaspter has quit [Ping timeout: 240 seconds]
11:52
camus is now known as kaspter
12:04
alpernebbi has joined #panfrost
12:28
nlhowell has quit [Ping timeout: 272 seconds]
12:43
nlhowell has joined #panfrost
13:25
nlhowell has quit [Remote host closed the connection]
13:27
nlhowell has joined #panfrost
13:47
nlhowell has quit [Ping timeout: 260 seconds]
13:51
alyssa has joined #panfrost
14:44
kaspter has quit [Quit: kaspter]
14:52
archetech__ has joined #panfrost
14:54
lucmann has quit [Remote host closed the connection]
15:18
archetech has quit [Quit: Leaving]
16:17
archetech__ has quit [Quit: Leaving]
17:04
<
bbrezillon >
alyssa: on bifrost, is LEA_ATTR supposed to load the desc from the varyings or attributes array?
17:05
<
alyssa >
bbrezillon: varyings, unless overrided by a paired DTSEL_IMM instruction
17:05
<
alyssa >
(this is an optimization -- usually LEA_ATTR is used to write out a varying, paired with ST_CVT)
17:08
<
alyssa >
We don't currently support emitting DTSEL_IMM and I don't see why you would need it except for bindless
17:08
<
alyssa >
(If you have a gles3 disasm from the blob using it, let me know, I'm intrigued.)
17:11
<
bbrezillon >
nope, I was just trying to figure how the attribute array was selected
17:11
<
bbrezillon >
if it's varyings by default, I'm all good
17:23
<
alyssa >
bbrezillon: this does have implications for images
17:23
<
alyssa >
namely, writeable images should be in the varying table, not the attribute table, since LEA_ATTR_TEX will default to the varying table
17:25
<
bbrezillon >
I'm not there yet
17:26
<
bbrezillon >
trying to debug another issue, which I think is caused by bad gl_Position updates
17:34
<
italov >
alyssa: bbrezillon: is there a specific attribute type for AFBC? or is it just the same as 3d interleaved?
17:40
<
bbrezillon >
italov: same type
17:40
<
bbrezillon >
but the stride calculation is different
17:42
<
italov >
bbrezillon: which stride? only the slice strid or also the other 2?
17:57
<
bbrezillon >
italov: aren't the other strides directly coming from the slice stride?
18:14
<
alyssa >
italov: bbrezillon: Images can't be AFBC on Mali.
18:16
<
alyssa >
If an AFBC surface is bound as an image, the driver needs to transition the layout (via an explicit decompression blit).
18:17
<
alyssa >
This shouldn't be too surprising. Something like a LEA_TEX operation is not well-defined for block-based formats.
18:18
<
alyssa >
How would an atomic image operation work with AFBC? Atomics are defined per pixel but AFBC does not support random access finer than 4x4.
18:22
<
italov >
alyssa: ok, that makes sense
18:24
<
italov >
I'm not sure how to debug the case where store works for rgba32f but does for rgba32i even though the only differences should be the format and the opcode, and those are already matching the blob's output, any ideas?
18:25
<
italov >
I meant "does not work for rgba32i"
18:28
<
alyssa >
If those are right then something else is wrong
19:20
stikonas has quit [Ping timeout: 272 seconds]
19:23
davidlt has quit [Ping timeout: 240 seconds]
19:25
<
macc24 >
is it faster to draw 1px triangles or points on panfrost devices?
19:31
<
alyssa >
If you have to ask your app is doing something wrong
19:31
<
macc24 >
ok... how about 2 triangles or a quad?
19:33
<
HdkR >
If you're using two triangles for a screen copy then you're already doing it wrong :P
19:33
<
macc24 >
i am not copying screen or anything
19:33
<
macc24 >
i am attempting to rewrite my ray marching renderer into C
19:36
<
macc24 >
in lua,rendering results to screen was the bottleneck
19:45
alyssa has left #panfrost [#panfrost]
19:51
stikonas has joined #panfrost
20:20
raster has quit [Quit: Gettin' stinky!]
20:26
italov has quit [Ping timeout: 240 seconds]
20:27
raster has joined #panfrost
20:45
italov has joined #panfrost
20:50
italov has quit [Ping timeout: 260 seconds]
21:05
indy has quit [Max SendQ exceeded]
21:13
indy has joined #panfrost
21:21
italov has joined #panfrost
21:26
italov has quit [Ping timeout: 260 seconds]
21:29
tomboy64 has quit [Remote host closed the connection]
21:29
tomboy64 has joined #panfrost
21:36
italov has joined #panfrost
21:43
alpernebbi has quit [Quit: alpernebbi]
21:46
italov has quit [Ping timeout: 272 seconds]
21:56
<
macc24 >
is it normal for graphics in cube2 to be jittery?
22:06
tomboy64 has quit [Remote host closed the connection]
22:07
tomboy64 has joined #panfrost
22:55
raster has quit [Quit: Gettin' stinky!]
23:04
kherbst has joined #panfrost
23:04
alyssa has joined #panfrost
23:04
karolherbst has quit [Ping timeout: 256 seconds]
23:12
raster has joined #panfrost
23:13
italov has joined #panfrost