Open sxales opened 5 years ago
How much RAM did you give the VM?
Just tested the latest retroarch snap on my Ubuntu 18 desktop without issue.
I gave it 4Gb and increasing it doesn't have any effect. I also tried Ubuntu 18 on the same VM with the same result (flycast doesn't work while psx does). On my laptop with Ubuntu 18 it works perfectly. I'm starting to think that it has something to do with my hardware.
In case this helps, I also have a SIGSEGV in DEbian 10 with Retroarch v1.8.6 and Flycast commit 3b10113b40d22b495b1ea01e275b7be84e8f85cc
built from source. Here is what the emulator dumps in the console when run with Dolphin Blue (works in Windows 10 binaries):
[libretro INFO] 00:00:177 hw/mem/_vmem.cpp:556 N[VMEM]: Info: nvmem is enabled, with addr space of size 4GB
[libretro WARN] 00:00:215 hw/naomi/naomi_cart.cpp:237 W[NAOMI]: Warning: Region 3 bios not found in awbios
Protocol error: bad 3 (Window); Sequence Number 11
Opcode (20, 0) = GetProperty
Bad resource 0 (0x0)
at -e line 15.
[libretro INFO] 00:00:858 hw/naomi/naomi_cart.cpp:373 N[NAOMI]: NAOMI GAME ID [Dolphin Blue]
[libretro INFO] 00:00:860 nullDC.cpp:267 N[BOOT]: [LUT]: Naomi ROM name is: Dolphin Blue.
[INFO] [Environ]: SET_ROTATION: 0
[INFO] [Environ]: SET_INPUT_DESCRIPTORS:
[libretro ERROR] 00:00:860 libretro/common.cpp:376 E[COMMON]: SIGSEGV @ 7fca45e8ca72 ... 0x7fc927ef0000 -> was not in vram (dyna code 0)
[libretro INFO] Fatal error : segfault
in signal_handler -> core/libretro/common.cpp : 383
[libretro ERROR] 00:00:860 libretro/libretro.cpp:3230 E[COMMON]: DEBUGBREAK!
[libretro ERROR] 00:00:861 libretro/common.cpp:376 E[COMMON]: SIGSEGV @ 7fc917d2c828 ... 0x7fc917d2c828 -> was not in vram (dyna code 0)
[libretro INFO] Fatal error : segfault
in signal_handler -> core/libretro/common.cpp : 383
[libretro ERROR] 00:00:861 libretro/libretro.cpp:3230 E[COMMON]: DEBUGBREAK!
Illegal instruction
Other ROMs showcase the same behavior. They work properly in Windows 10 builds.
I was seeing this exact same crash with my Gentoo ebuild. After a lot of debugging and hair-pulling I discovered it only crashes with run-ahead enabled!
Runahead would not work with this core. It doesnt really work with hardware rendered cores.
@sjnewbury indeed I had run ahead enabled in the general latency config (not specifically for flycast), and by disabling it I can now use the latest version of flycast without any issue. However I wonder why this feature causes instability if it's not even supported by Flycast. Thanks a lot!
@twinaphex So I guess this bug is actually that the flycast core isn't failing to start with run-ahead, but crashes because the core tries to access memory from the second instance? Isn't there some way to flag run-ahead incompatibility?
This Issue is old but I'm facing the ame issue in Pi4 + Rarpbian Lite. What is interesting is that if I execute the retroarch command to launch a game by typing in the terminal it crashes with the segfault but if I connect to the Pi via SSH and execute the exact same command the game launches perfectly well.
I update yesterday for XU4.
Here is the log when trying to run dolphin.zip https://pastebin.com/VCKkrX2x Just in case it was an issue with the rom or bios I ran for a previously working rom of Crazy Taxi for dreamcast https://pastebin.com/F9kZg1Ys
I removed the naomi_boot.bin and dc_nvmem.bin based on #434 but no dice.