Quoting Thomas Huth (2023-03-31 15:17:33) > On 31/03/2023 13.30, Nico Boehr wrote: > > Hi Paolo and/or Thomas, > > > > here comes the first pull request from me. :) > > Thanks! > > ... I gave it a try, but I'm hitting a failure in the spec_ex test: > > $ cat logs/spec_ex.log > timeout -k 1s --foreground 90s qemu-system-s390x -nodefaults -nographic -machine s390-ccw-virtio,accel=kvm -chardev stdio,id=con0 -device sclpconsole,chardev=con0 -kernel s390x/spec_ex.elf -smp 1 # -initrd /tmp/tmp.cxHP06rT1F > PASS: specification exception: psw_bit_12_is_1: Program interrupt: expected(6) == received(6) > PASS: specification exception: short_psw_bit_12_is_0: Program interrupt: expected(6) == received(6) > FAIL: specification exception: psw_odd_address: Expected exception due to invalid PSW > PASS: specification exception: odd_ex_target: did not perform ex with odd target > PASS: specification exception: odd_ex_target: Program interrupt: expected(6) == received(6) > PASS: specification exception: bad_alignment_lqp: Program interrupt: expected(6) == received(6) > PASS: specification exception: bad_alignment_lrl: Program interrupt: expected(6) == received(6) > PASS: specification exception: not_even: Program interrupt: expected(6) == received(6) > PASS: specification exception during transaction: odd_ex_target: Program interrupt: expected(518) == received(518) > PASS: specification exception during transaction: bad_alignment_lqp: Program interrupt: expected(518) == received(518) > PASS: specification exception during transaction: bad_alignment_lrl: Program interrupt: expected(518) == received(518) > PASS: specification exception during transaction: not_even: Program interrupt: expected(518) == received(518) > SUMMARY: 12 tests, 1 unexpected failures > > EXIT: STATUS=3 > > I'm sure I'm missing something, I just cannot figure it out right > now (it's Friday afternoon...) - QEMU is the current version from > the master branch, so I thought that it should contain all the > recent fixes ... does this psw_odd_address test require a fix in > the kernel, too? (I'm currently running a RHEL9 kernel) Thomas, I cannot reproduce this with a QEMU master and 6.3-rc4, so you might be right that some kernel fix is required. Since the weekend is calling now, I will take a look at this on Monday.