Re: [hatari-devel] build test fails for macos on cirrus ci

Hi,

Is that failure 100% reproducible?

Yes. it fails every time I run cmdfifo.sh on my Mac.

Try running the test in a loop to see whether it's 100% reproducible.
And provide the exact output from Hatari ("out.txt" file) when there's
error.

I attached out.txt to this email.. I turned on --log-level debug in hatari so there's debug output from Hatari in there as well.

If I ls the contents of $testdir then I see the following:

$ ls /var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/

grab0001.png hatari.nvram out.txt testmem.sav

So it has successfully saved the screenshot and the memory snapshot but it has silently failed to save the audio recording sndrec.wav and the video recording videorec.avi. These two files are specified in the temporary hatari.cfg that is created by the script so I'm wondering if the problem is caused by Hatari failing to correctly read and use the config. I'll try to find some time to investigate that next..

How does that sound and do you have any other suggestions for things I can look at?

Cheers,

Chris

Based on your first comment, it seems that problem might not be with the
cmdfifo functionality itself, but in the Hatari functionality that is
tested through it.

Nicolas, cmdfifo test is the only test checking that Hatari sound,
video, memory state and screenshot saving functionalities actually
produce a file.

I.e. IMHO we should not just disable cmdfifo test again, because that
would mean lacking even basic testing for several significant functions
in Hatari.

- Eero

> On Tue, 31 May 2022 at 20:58, Nicolas Pomarède <npomarede@xxxxxxxxxxxx>
> wrote:
>
>> Hi
>>
>> I see there's an error while running hatari's tests for the macOS build,
>> for example :
>>
>> https://cirrus-ci.com/task/5051027262013440
>>
>> This seems to be present since this commit :
>>
>> 73d9be6f32 "Fix test script name to enable cmdfifo testing"
>>
>> Eero, can you have a look ? It fails only on macOS, but maybe it's
>> something that you can test anyway ?
>>
>> Nicolas
>>
>>
>>
>

DEBUG: creating command FIFO: /var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/cmdfifo DEBUG: STMemory_RAM_SetBankSize total=1024 KB bank0=512 KB bank1=512 KB MMU=5 DEBUG: Video_SetSystemTimings 0 3 -> 2 (WS3) 503 503 64 INFO : Hatari v2.4.0-devel (May 31 2022), compiled on: May 31 2022, 17:55:56 DEBUG: Video_SetSystemTimings 0 3 -> 2 (WS3) 503 503 64 DEBUG: Configured max Hatari resolution = 832x588, optimal for ST = 832x552(+36) TODO : Support for Hatari window reparenting not built in DEBUG: STMemory_RAM_SetBankSize total=1024 KB bank0=512 KB bank1=512 KB MMU=5 DEBUG: update cyc speed 0 throttle 0.000000 clock_mult 512 DEBUG: CPU cycleunit: 256 (0.500) DEBUG: 1872 CPU functions DEBUG: memory_map_Standard_RAM total=1048576 ram0=524288 ram1=524288 mmu0=131072 mmu1=131072 DEBUG: Loaded TOS version 0.00, starting at $e00000, country code = 0, PAL DEBUG: Skipped TOS patches. DEBUG: Video_SetSystemTimings 0 3 -> 2 (WS3) 503 503 64 DEBUG: Video_SetSystemTimings 0 3 -> 2 (WS3) 503 503 64 INFO : NVRAM not found at '/var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/hatari.nvram' DEBUG: Host system has accurate delays. (1) DEBUG: update cyc speed 0 throttle 0.000000 clock_mult 512 DEBUG: CPU cycleunit: 256 (0.500) DEBUG: Building CPU, 44274 opcodes (0 -2 1) DEBUG: CPU=68000, FPU=0, MMU=0, JIT=0. DEBUG: prefetch and cycle-exact DEBUG: 24-bit DEBUG: DEBUG: hardreset, memory cleared DEBUG: m68k_run_1_ce DEBUG: CPU reset PC=e00034 (ROM memory).. DEBUG: memory_map_Standard_RAM total=1048576 ram0=524288 ram1=524288 mmu0=524288 mmu1=524288 > r D0 00000000 D1 00000000 D2 00000000 D3 00000000 D4 00000000 D5 00000000 D6 00000000 D7 00000000 A0 00000000 A1 00000000 A2 00000000 A3 00000000 A4 00000000 A5 00000000 A6 00000000 A7 00017FF8 USP 00017FF8 ISP 00020000 T=00 S=0 M=0 X=0 N=0 Z=0 V=0 C=0 IMASK=7 STP=0 Prefetch 4ef8 (JMP) 1100 (MOVE) Chip latch 00000000 00001100 4ef8 1100 jmp $1100 Next PC: 00001104 memsave: /var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/hatari.sav -> /var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/testmem.sav Screen dump saved to: /var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/grab0001.png INFO : WAV sound data recording has been started. INFO : AVI recording has been started Command line with '2' arguments: - '--spec512' - '256' DEBUG: Spec512 threshold = 256 palette writes per line. DEBUG: STMemory_RAM_SetBankSize total=1024 KB bank0=512 KB bank1=512 KB MMU=5 DEBUG: Video_SetSystemTimings 0 3 -> 2 (WS3) 503 503 64 printer: ON Embedded window ID change messages = ON DEBUG: SDL2 window event: 0x1 INFO : Memory state file saved: /var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/testmem.sav DEBUG: exit m68k_run DEBUG: removing command FIFO: /var/folders/kk/xv3dmf2n31vb5hvlh65f7cmc0000gp/T/tmp.VLDfkT9d/cmdfifo INFO : AVI recording has been stopped INFO : WAV Sound data recording has been stopped.