alyssa changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard & Bifrost - Logs https://freenode.irclog.whitequark.org/panfrost - <daniels> avoiding X is a huge feature
raster has quit [Quit: Gettin' stinky!]
WoC has joined #panfrost
kinkinkijkin has quit [Ping timeout: 260 seconds]
youcai_ has joined #panfrost
kinkinkijkin has joined #panfrost
ezequielg has quit [Ping timeout: 260 seconds]
ezequielg_ has joined #panfrost
youcai has quit [Ping timeout: 264 seconds]
narmstrong has quit [Ping timeout: 264 seconds]
jstultz has quit [Ping timeout: 264 seconds]
youcai_ is now known as youcai
jstultz has joined #panfrost
narmstrong has joined #panfrost
WoC has quit [Remote host closed the connection]
WoC has joined #panfrost
WoC has quit [Remote host closed the connection]
WoC has joined #panfrost
WoC has quit [Ping timeout: 264 seconds]
WoC` has joined #panfrost
WoC` has quit [Remote host closed the connection]
popolon has quit [Quit: WeeChat 3.0]
stikonas has quit [Remote host closed the connection]
kaspter has joined #panfrost
atler is now known as Guest88774
Guest88774 has quit [Killed (tepper.freenode.net (Nickname regained by services))]
atler has joined #panfrost
vstehle has quit [Ping timeout: 256 seconds]
WoC has joined #panfrost
urjaman has quit [Read error: Connection reset by peer]
urjaman has joined #panfrost
WoC has quit [Remote host closed the connection]
WoC has joined #panfrost
WoC has quit [Remote host closed the connection]
kaspter has quit [Ping timeout: 265 seconds]
kaspter has joined #panfrost
WoC has joined #panfrost
WoC has quit [Read error: Connection reset by peer]
Green has quit [Quit: ...]
Green has joined #panfrost
davidlt has joined #panfrost
mixfix41 has joined #panfrost
mixfix41 has quit [Quit: Leaving.]
<icecream95> alyssa: Is there any documentation on what the source id means for GPU faults?
camus1 has joined #panfrost
kaspter has quit [Ping timeout: 256 seconds]
camus1 is now known as kaspter
vstehle has joined #panfrost
chewitt has quit [Quit: Zzz..]
mixfix41 has joined #panfrost
indy has quit [Ping timeout: 240 seconds]
_whitelogger has joined #panfrost
chewitt has joined #panfrost
felipealmeida has quit [Ping timeout: 246 seconds]
felipealmeida has joined #panfrost
kaspter has quit [Ping timeout: 246 seconds]
guillaume_g has joined #panfrost
kaspter has joined #panfrost
raster has joined #panfrost
raster has quit [Client Quit]
indy has joined #panfrost
raster has joined #panfrost
<daniels> icecream95: it's not standardised per-arch but varies per-SKU instead, and it's not exactly well documented
kaspter has quit [Ping timeout: 264 seconds]
kaspter has joined #panfrost
stikonas has joined #panfrost
davidlt_ has joined #panfrost
davidlt has quit [Ping timeout: 256 seconds]
robmur01 has joined #panfrost
alpernebbi has joined #panfrost
<robmur01> daniels: also they have totally different meanings for GPU faults vs. MMU faults...
nlhowell has joined #panfrost
karolherbst has joined #panfrost
davidlt__ has joined #panfrost
indy has quit [Quit: ZNC - http://znc.sourceforge.net]
dstzd has quit [Ping timeout: 246 seconds]
davidlt_ has quit [Ping timeout: 260 seconds]
jernej has quit [Ping timeout: 246 seconds]
megi has quit [Quit: WeeChat 3.0]
megi has joined #panfrost
indy has joined #panfrost
davidlt__ is now known as davidlt
<alyssa> icecream95: If there is I don't have it
dstzd has joined #panfrost
jernej has joined #panfrost
br has quit [Ping timeout: 265 seconds]
br has joined #panfrost
<alyssa> icecream95: bbrezillon: FWIW, I've been running the no-madvise branch on and off, and it's noticeably more stable.
karolherbst has quit [Ping timeout: 265 seconds]
karolherbst has joined #panfrost
kaspter has quit [Ping timeout: 265 seconds]
<bbrezillon> alyssa: acked
alpernebbi has quit [Quit: alpernebbi]
chewitt has quit [Quit: Zzz..]
davidlt has quit [Ping timeout: 240 seconds]
robmur01 has quit [Quit: Leaving]
Elpaulo has joined #panfrost
guillaume_g has quit [Quit: Konversation terminated!]
raster has quit [Quit: Gettin' stinky!]
cwabbott has quit [Ping timeout: 272 seconds]
narmstrong has quit [Read error: Connection reset by peer]
kinkinkijkin has quit [Read error: Connection reset by peer]
cwabbott has joined #panfrost
narmstrong has joined #panfrost
steev has quit [Ping timeout: 260 seconds]
kinkinkijkin has joined #panfrost
steev has joined #panfrost
alyssa has quit [Remote host closed the connection]
<icecream95> Okay, guess I'll have to randomly change GPU pointers to invalid addresses until I get a matching source id…
alyssa has joined #panfrost
<alyssa> icecream95: "Allocating large BOs is *probably* rare"
<alyssa> Thanks to Mali's weird varying allocation problems, some levels of SuperTuxKart allocate several MB of varyings alone every frame.
<alyssa> (200k triangles * 3 verts/tri * 16 bytes of varyings / vert = ~10MB)
<alyssa> In fact our BO cache originally exists to deal with that nonsense...
<icecream95> alyssa: Is that in a single BO?
<alyssa> uhm, let's see, I know I tried like 3 different allocation schemes at different points
<alyssa> Currently, we allocate varyings from the invisible pool
<icecream95> (which, by the way, is the source of a lot of overhead)
<alyssa> (it was so much worse before, if you can believe that)
<alyssa> So that means the varying BOs are 64k each, unless a single draw exceeds that.
<alyssa> (That should also be possible with sufficiently high-poly models)
* chrisf mumbles something about the vulkan blob exhausting 180MB of varying space and just giving up
<alyssa> chrisf: lol
<alyssa> I hope Valhall fixed this.
<chrisf> nope
<alyssa> Boo.
mixfix41 has quit [Ping timeout: 246 seconds]
karolherbst has quit [Ping timeout: 272 seconds]
<macc24> alyssa: whats !8698
<alyssa> macc24: fighting over why panfrost machines freeze so much
<macc24> is it ready to be tested?
<alyssa> sure, then you can pile on the argument
* macc24 says: testing is the future, and future starts here
<macc24> alyssa: wait how do i test this
<macc24> my only weakness - patches that i don't know how to test
<alyssa> If only I knew
<macc24> i wonder if this fixes chromium on my chromebook
<macc24> no change between master and !8698
<alyssa> If you've seen "Purging X bytes" in dmesg, that's related
<macc24> i have never seen this message