hch12907__ has quit [Ping timeout: 480 seconds]
hch12907__ has joined #zink
hch12907_ has joined #zink
hch12907__ has quit [Ping timeout: 480 seconds]
hch12907__ has joined #zink
hch12907_ has quit [Ping timeout: 480 seconds]
hch12907__ has quit [Ping timeout: 480 seconds]
hch12907_ has joined #zink
hch12907__ has joined #zink
hch12907 has joined #zink
hch12907_ has quit [Ping timeout: 480 seconds]
hch12907__ has quit [Ping timeout: 480 seconds]
hch12907_ has joined #zink
hch12907 has quit [Ping timeout: 480 seconds]
hch12907_ has quit [Ping timeout: 480 seconds]
hch12907_ has joined #zink
hch12907__ has joined #zink
hch12907_ has quit [Ping timeout: 480 seconds]
hch12907__ has quit [Ping timeout: 480 seconds]
hch12907__ has joined #zink
hch12907_ has joined #zink
hch12907 has joined #zink
hch12907__ has quit [Ping timeout: 480 seconds]
hch12907__ has joined #zink
hch12907_ has quit [Ping timeout: 480 seconds]
hch12907 has quit [Ping timeout: 480 seconds]
hch12907_ has joined #zink
hch12907__ has quit [Ping timeout: 480 seconds]
hch12907_ has quit [Ping timeout: 480 seconds]
hch12907 has joined #zink
hch12907_ has joined #zink
hch12907__ has joined #zink
hch12907 has quit [Ping timeout: 480 seconds]
hch12907_ has quit [Ping timeout: 480 seconds]
hch12907_ has joined #zink
hch12907__ has quit [Ping timeout: 480 seconds]
jekstrand has joined #zink
<
zmike>
I forgot to comment
damcclos has joined #zink
turol has joined #zink
Plagman has joined #zink
fahien has joined #zink
<
jekstrand>
zmike: Looking at these clear tests. I'm seeing a clear with a UNORM format and UINT data showing up in ANV.
<
jekstrand>
Yup. And, yes, I've got the !14932 branch
<
zmike>
I'm just passing the clear colors through as I get them
<
jekstrand>
zmike: Pushed a patch on top of that MR. They all pass now. Your bug, not mine. :-P
<
jekstrand>
No idea how they passed anywhere else
<
zmike>
do I need to ask Mr Nvidia so we get a fourth opinion?
<
jekstrand>
Oh, there's no opinion there. You need to do that. Why does it work elsewhere? I don't know.
<
zmike>
well if it fails for nvidia too then that means we need cts coverage
<
jekstrand>
It could be that Nvidia isn't getting the CmdClearColorImage path for some reason
<
zmike>
I haven't tried, but all drivers will always hit the same codepaths
<
jekstrand>
Oh, well then I don't know why it works anywhere else. It's clearly wrong. :P
<
zmike>
I had originally thought of doing something like this, but then it passed without, so 🤔
<
jekstrand>
I guess it could depend on how they interpret those NaNs you passed in.
* jekstrand
has done his Zink good deed for the day
<
zmike>
I'll take it
<
zmike>
like a genius, I've removed my own commit from that MR and added my own rb, which allows me to merge it without further delay
<
zmike>
jekstrand: thanks!
<
jekstrand>
Deleting all your own commits seems like a great way to get MRs landed...
<
zmike>
it's the next level of deleting code
<
jekstrand>
Too bad I can't retroactively delete those GL_ARB_copy_image commits from way back
<
zmike>
I could do a formatting patch to conceal them from git blame
<
zmike>
add in some nice trailing whitespace
ella-0[m] has joined #zink
damcclos has quit [Remote host closed the connection]