seninha has quit [Remote host closed the connection]
riff-IRC has quit [Remote host closed the connection]
riff-IRC has joined #riscv
handsome_feng has joined #riscv
davidlt has quit [Ping timeout: 248 seconds]
conordooley95 has joined #riscv
conordooley95 is now known as conchuod
___nick___ has joined #riscv
gordonDrogon has quit [Ping timeout: 256 seconds]
gordonDrogon has joined #riscv
aerkiaga has joined #riscv
corank_ has quit [Remote host closed the connection]
corank_ has joined #riscv
handsome_feng has quit [Quit: Connection closed for inactivity]
jmdaemon has quit [Ping timeout: 240 seconds]
jacklsw has quit [Ping timeout: 240 seconds]
pecastro has joined #riscv
drmpeg has quit [Ping timeout: 240 seconds]
drmpeg has joined #riscv
corank_ has quit [Remote host closed the connection]
corank_ has joined #riscv
BOKALDO has quit [Quit: Leaving]
aerkiaga has quit [Remote host closed the connection]
drmpeg has left #riscv [#riscv]
drmpeg has joined #riscv
adjtm has joined #riscv
littlebobeep has quit [Remote host closed the connection]
littlebobeep has joined #riscv
cousteau has joined #riscv
BOKALDO has joined #riscv
aerkiaga has joined #riscv
adjtm has quit [Ping timeout: 246 seconds]
adjtm has joined #riscv
kaph has joined #riscv
kaph_ has joined #riscv
kaph has quit [Read error: Connection reset by peer]
aerkiaga has quit [Remote host closed the connection]
Sofia has quit [Ping timeout: 240 seconds]
EchelonX has joined #riscv
Sofia has joined #riscv
Andre_H has joined #riscv
kaph_ has quit [Ping timeout: 246 seconds]
corank_ has quit [Remote host closed the connection]
corank_ has joined #riscv
motherfsck has joined #riscv
mahmutov has joined #riscv
Sofia has quit [Ping timeout: 240 seconds]
seninha has joined #riscv
Andre_H has quit [Quit: Leaving.]
Sofia has joined #riscv
<conordooley>
palmer: i gave the generic spinlocks a go today, from what limited stuff i did i saw no meaninful difference. you didnt come up with anything to specifically test since i last asked, did you?
kaph_ has joined #riscv
crabbedhaloablut has quit [Remote host closed the connection]
crabbedhaloablut has joined #riscv
jacklsw has joined #riscv
jacklsw has joined #riscv
<palmer>
nope, and I don't really have any specific ideas
<palmer>
I might push them to our internal Linux branch so some of the folks here can give them a shot, but anything we find there would likely be an internal issue anyway so not that exciting
Andre_H has joined #riscv
vagrantc has joined #riscv
Andre_H has quit [Quit: Leaving.]
toulene has joined #riscv
<conordooley>
ye, figured you hadn't. i'll do a bit of the same & dump it in our ci i guess :)
<palmer>
cool, thanks
<palmer>
Boqun had some feedback on an ordering issue, but I think it doesn't matter for RISC-V (not sure I fully understand what's going on, though)
jacklsw has quit [Read error: Connection reset by peer]
davidlt has joined #riscv
davidlt has quit [Client Quit]
<jimwilson>
nmeum, newlib doesn't know anything about zicsr, but I think it doesn't use any csr other than fscsr/frcsr/fflags which is implied by F
<nmeum>
jimwilson: yes, I think the problem in my case is that my gcc cross toolchain uses the default t-elf-multilib configuration file which doesn't have MULTILIB_REUSE rules for zicsr/zifence/… because I would expect gcc to pick up the newlib rv32imac library files when linking/compiling with -march=rv32imac_zicsr (which it doesn't out of the box)
<jimwilson>
nmeum, current gcc release has no zicsr/zifencei support, it uses the old ISA version without them, next gcc release, gcc-12, will support zicsr and zifencei, and we expect stuff will break
<jimwilson>
so currently csrs are still part of the base I extension
<nmeum>
but gcc 11.X does support -misa-spec=20191213, no? it's just not the default
<jimwilson>
it accepts it, but it doesn't do anything
<jimwilson>
at least not anything in gcc, it might do something inside binutils, but if you have a binutils that properly supports it, and gcc doesn't, things will break
<nmeum>
ahhhhaaa, you are right. the "unrecognized opcode" errors I am seeing for csr instruction when compiling with `-misa-spec=20191213 -march=rv32imac` are from the binutils 2.38 assembler, not from gcc itself
<jimwilson>
yes, the binutils release from January has the right support, and the gcc release from next month or so will have matching support, but using binutils-2.38 with gcc-11 can lead to some confusion
<nmeum>
will the upcoming gcc release also have multilib re-use rules for rv32imac → rv32imac_zicsr etc.? because looking at gcc git HEAD I don't see zicsr / zifence / … stuff in the default t-elf-multilib file. Conceptually speaking, I just don't understand how gcc would otherwise infer that it can use rv32imac newlib libraries for -march=rv32imac_zicsr
<jimwilson>
I don't think anyone has thought about multilibbing yet. Linux doesn't multilib. And riscv-gnu-toolchain can have its own multilib list pretty easily.
BOKALDO has quit [Quit: Leaving]
<palmer>
I think I saw another comment about the multilib rules, we should think about that a bit
EchelonX has quit [Quit: Leaving]
___nick___ has quit [Ping timeout: 248 seconds]
littlebobeep has quit [Remote host closed the connection]
littlebobeep has joined #riscv
littlebo1eep has joined #riscv
littlebobeep has quit [Remote host closed the connection]
kaph_ has quit [Ping timeout: 256 seconds]
seninha has quit [Remote host closed the connection]