IRC logs for #openrisc Friday, 2017-01-13

--- Log opened Fri Jan 13 00:00:11 2017
stekernshorne: what's the problem now?04:18
shornestekern: havent had time to trace it... getting some time now05:17
shorneActually, I was trying to compile the kernel in debug mode05:17
shorneso I could better tie the code to the assembly05:18
shornebut got an error "unaligned opcodes"... I fixed that and am adding a patch.  The issue is the ".strings" in head.S are in executable section.  SO dwarf generation complains.05:19
shornenow running debug kernel05:19
shorneand it works fine05:19
shorneI think my pc adjustment fails06:06
shorneafter l.rfe, code dies, adjustment logic shouldnt be hit and it is06:07
shorneyup boots without my code :)06:18
shorneso my qemu patch works, but not my kernel patch06:20
shornestekern: this is failing on this line06:21
shornehttps://github.com/stffrdhrn/linux/blob/openrisc-4.11/arch/openrisc/kernel/traps.c#L34406:21
shorneWhen exception is not in delay slot this if condition evaluates true.06:22
shorneif (regs->sr & SPR_SR_DSX) {06:22
shornedid I write something wrong there?06:23
shorneIt should be false... maybe another bug in qemu06:23
stekernmaybe it doesn't implement the DSX flag?06:24
shorneit does...06:26
shorneI think its management of when its in the delay slot is wrong06:26
shornestekern: fyi https://github.com/qemu/qemu/blob/master/target/openrisc/interrupt.c#L3906:28
shornebut I think D_FLAG has a bug06:28
shorneah... doh, qemu will DSX in delay, but then never clear it ever16:39
shorneok, qemu fixed16:41
shornemy adjust pc code works16:41
shorneok, qemu patch sent, now I think linux patches are ready16:58
shorneI would have to say qemu is pretty nice (compared to or1ksim) it might be a bit faster17:26
shorneah... checkpatch warnings18:33
--- Log closed Sat Jan 14 00:00:13 2017

Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!