Armand has quit [Read error: Connection reset by peer]
armand_ is now known as Armand
KombuchaKip has quit [Quit: Leaving.]
KombuchaKip has joined #riscv
junaid_ has quit [Remote host closed the connection]
Armand has quit [Ping timeout: 240 seconds]
heat_ has quit [Remote host closed the connection]
heat_ has joined #riscv
pecastro has joined #riscv
MaxGanzII has joined #riscv
junaid_ has joined #riscv
solrize has quit [Ping timeout: 240 seconds]
ldevulder has joined #riscv
Stat_headcrabed has joined #riscv
foton has quit [Quit: %Bye, bye, ...%]
Stat_headcrabed has quit [Client Quit]
foton has joined #riscv
solrize has joined #riscv
solrize has quit [Changing host]
solrize has joined #riscv
MaxGanzII has quit [Ping timeout: 240 seconds]
danilogondolfo has joined #riscv
JanC has quit [Remote host closed the connection]
JanC has joined #riscv
<bjdooks>
ah -msave-restore and -fno-omit-frame-pointer incompatible on riscv
<Esmil>
arnd: i don't want to delay prabhakar's patch series any further, but just curious if there's a reason we don't want something like this: http://sprunge.us/yNosCT
<Esmil>
..so compliant cores should only suffer an extra nop
JanC_ has joined #riscv
JanC has quit [Killed (cadmium.libera.chat (Nickname regained by services))]
JanC_ is now known as JanC
<arnd>
Esmil: I have no objection to that, but I don't think anyone cares about the cost of a correctly predicted branch here, compared to the cost of the cache flush
<Esmil>
i see, thanks
<arnd>
static_call() is probably even better here, but that only seems to exist on arm, powerpc and x86 at the moment
<Esmil>
yes
junaid_ has quit [Ping timeout: 248 seconds]
<conchuod>
There were some patches posted for static call some weeks ago, but noone reviewed them nor were they in a great state.
<geertu>
arnd: hch does, he objects against any conditional code or function pointers
another is now known as another|
junaid_ has joined #riscv
prabhakarlad has joined #riscv
jacklsw has quit [Quit: Back to the real world]
MaxGanzII has joined #riscv
heat__ has joined #riscv
heat_ has quit [Read error: Connection reset by peer]
enoq has joined #riscv
prabhakarlad has quit [Quit: Client closed]
FL4SHK has quit [Ping timeout: 250 seconds]
prabhakar has quit [Remote host closed the connection]
wingsorc has quit [Quit: Leaving]
hays has quit [Remote host closed the connection]
prabhakar has joined #riscv
prabhakarlad has joined #riscv
prabhakar has quit [Remote host closed the connection]
prabhakar has joined #riscv
FL4SHK has joined #riscv
prabhakar has quit [Remote host closed the connection]
Tenkawa has joined #riscv
hays has joined #riscv
heat has joined #riscv
heat__ has quit [Read error: Connection reset by peer]
radu242753435752 has joined #riscv
radu242753435752 has quit [Ping timeout: 240 seconds]
prabhakar has joined #riscv
<bjdooks>
hmm, is it me or does uboot not compile with -msave-return ? wonder if it is an andes extension?
<jrtc27>
-msave-restore requires no ISA support
<bjdooks>
hmm, is it me or does uboot not have a config option to compile with -msave-return ? wonder if it is an andes extension?
aredridel has quit [Read error: Connection reset by peer]
aredridel has joined #riscv
<bjdooks>
jrtc27: i was looking to see if there's a config option to add -msave-restore to the CFLAGS, it does seem to build and work with it, but it can't work without framepointer w/o fixing libgcc
<jrtc27>
what's wrong with libgcc?
<bjdooks>
if you use -msave-restore it doesn't keep the frame-pointer
<jrtc27>
I see gcc -msave-restore -fno-omit-frame-pointer just not bothering to use the libcalls
<jrtc27>
which is odd
<jrtc27>
I guess because they're pushed onto the stack the "wrong" way round compared with the standard frame layout
heat has quit [Remote host closed the connection]
<jrtc27>
oh, no, they're not
heat has joined #riscv
<bjdooks>
i've not managed to read throgh this, it is possible the code just doesn't take into account the fp
<bjdooks>
the bug andes found in my uboot unwinder is due to this
<jrtc27>
eh, kinda, hm
<jrtc27>
seems they get put at the top of the "block"
<jrtc27>
so yeah -msave-restore breaks the stack frame convention :(
<jrtc27>
it doesn't *have* to, that's just what makes the libcall implementation smaller
<bjdooks>
not sure if it is worth trying to make a version with fp support and hve it use those for -fno-omit-frame-pointer or just find out how andes are building and then go "no unwind support for you"
<bjdooks>
git clone of gcc source code over conference wifi is not fast
BootLayer has quit [Read error: Connection reset by peer]
aredridel has quit [Read error: Connection reset by peer]
aredridel has joined #riscv
pedja has joined #riscv
prabhakarlad has quit [Quit: Client closed]
BootLayer has joined #riscv
junaid_ has quit [Remote host closed the connection]
MaxGanzII has quit [Ping timeout: 240 seconds]
<palmer>
atish was asking, the Linux PW meeting ended pretty early today
<bjdooks>
jrtc27: interesting, newer gcc looks like they turn -msave-restore off iff -fno-omit-frame-pointer
epony has quit [Ping timeout: 240 seconds]
MaxGanzII has joined #riscv
epony has joined #riscv
pedja has quit [Quit: Leaving]
<muurkha>
iff?
<muurkha>
hopefully you mean if
<muurkha>
'if and only if' would imply there was no other way to turn off -msave-restore
vagrantc has joined #riscv
Andre_Z has joined #riscv
<bjdooks>
i suppose there's a few other places it gets turned off too