On Fri, Jul 1, 2011 at 4:15 PM, Laurent Desnogues <laurent.desnog...@gmail.com> wrote: > On Fri, Jul 1, 2011 at 2:57 PM, Jakub Jermar <ja...@jermar.eu> wrote: > [...] >> When _not_ singlestepping via GDB's `stepi`, the testcase will fail and >> crash Qemu like this: >> >> qemu: fatal: Trap 0x0101 while trap level (5) >= MAXTL (5), Error state >> <register dumped here> >> >> On the other hand, when I attach GDB to Qemu and step through all >> instructions using `stepi`, the testcase will succeed and crash Qemu >> like this: >> >> qemu: fatal: Trap 0x0100 while trap level (5) >= MAXTL (5), Error state >> <registers dumped here> >> >> Mind the difference in the trap type - 0x100 for success, 0x101 for failure. >> >> This is how I run the test: >> >> Without GDB: >> $ qemu-system-sparc64 -bios ./testcase >> >> With GDB: >> $ qemu-system-sparc64 -bios ./testcase -s -S >> >> From another terminal: >> $ /usr/local/cross/sparc64/bin/sparc64-linux-gnu-gdb >> (gdb) set architecture sparc:v9 >> (gdb) target remote localhost:1234 >> (gdb) stepi >> ... >> >> Hope this helps to fix the problem.
It definitely does, thanks a lot Jakub! > You don't have to use gdb to reproduce the issue, just add -singlestep > when running qemu. > > I find it odd that udivx is using cpu_cc_src and cpu_cc_src2. Using > dedicated local temps seems to fix the issue. Do we need to copy cpu_src* to further temps at all? IMHO - tcg_gen_mov_tl(cpu_cc_src, cpu_src1); - tcg_gen_mov_tl(cpu_cc_src2, cpu_src2); - gen_trap_ifdivzero_tl(cpu_cc_src2); - tcg_gen_divu_i64(cpu_dst, cpu_cc_src, cpu_cc_src2); + gen_trap_ifdivzero_tl(cpu_src2); + tcg_gen_divu_i64(cpu_dst, cpu_src1, cpu_src2); should do it. Or cpu_src is what you mean by dedicated? -- Regards, Artyom Tarasenko solaris/sparc under qemu blog: http://tyom.blogspot.com/
