diff options
Diffstat (limited to 'gitlab/issues_text/target_missing/host_arm/accel_TCG')
3 files changed, 40 insertions, 0 deletions
diff --git a/gitlab/issues_text/target_missing/host_arm/accel_TCG/1147 b/gitlab/issues_text/target_missing/host_arm/accel_TCG/1147 new file mode 100644 index 000000000..e5a68adc4 --- /dev/null +++ b/gitlab/issues_text/target_missing/host_arm/accel_TCG/1147 @@ -0,0 +1,9 @@ +x86_64 emu on aarch64 host: cpu_exec: assertion failed: (cpu == current_cpu) +Description of problem: +Execution of some binaries crashes with `Bail out! ERROR:../qemu-7.0.0/accel/tcg/cpu-exec.c:933:cpu_exec: assertion failed: (cpu == current_cpu)`. Looking at the code, that code is wrapped in a gcc/clang ifdef. Recompiling with clang produces this crash instead: `... include/qemu/rcu.h:102: void rcu_read_unlock(void): Assertion 'p_rcu_reader->depth != 0' failed.` + +No easier steps to reproduce (yet) than `systemd-nspawn`ing into an x86_64 Ubuntu container invoking qemu-x86_64-static through binfmt. Commands such as `ls` work fine, while `apt-get` will immediately crash with the error listed above. + +Note that this happens running Asahi Linux on the bare metal of an M1-based Macbook Pro. This same issue does *not* occur running the *same* binaries with the *same* x86_64 Ubuntu image on an Arch or Ubuntu VM under macOS on the same machine - regardless of if the QEMU binaries were built in a VM or in Asahi. + +These are big.LITTLE chips. Using taskset/affinity to limit the target process to a single specific core does not help. The Asahi kernel has a 16K page-size, which is known to cause trouble for some programs. qemu-arm(-static) however works without any issues (the M1 cannot run 32-bit ARM code natively, only 64-bit). diff --git a/gitlab/issues_text/target_missing/host_arm/accel_TCG/1714 b/gitlab/issues_text/target_missing/host_arm/accel_TCG/1714 new file mode 100644 index 000000000..09eafb6fd --- /dev/null +++ b/gitlab/issues_text/target_missing/host_arm/accel_TCG/1714 @@ -0,0 +1,27 @@ +QEMU crashes on ARMv7 since at least commit 493c9b19 +Description of problem: +I'm trying to build QEMU for Android, Arm64 versions work well, but **Armv7** builds began to crash nearly since this series of commits (QEMU 7.2.50), related to 'TCG_TARGET_HAS_direct_jump' removal by @rth7680. +More precisely, this commit still works: + +https://gitlab.com/qemu-project/qemu/-/commit/82df11e78d0baef7ffb7e7933c6fb830ffed087c + +and this one crashes: + +https://gitlab.com/qemu-project/qemu/-/commit/493c9b19a7fb7f387c4fcf57d3836504d5242bf5 + +(I tracked commits of 'tcg' subfolder and didn't bisect finer, but it's possible if needed). + +Both qemu-system-x86_64 and qemu-system-i386 emulators crash. + +**The crash is related to translation buffer size** : if I don't specify "-accel tcg,thread=single **,tb-size=256** ", the machine works. + +The problem is that I can not run debugger on a phone, and crash dump does not show any useful information, just "segfault" reason ("Fatal signal 11 (SIGSEGV), code 2 (SEGV_ACCERR), fault addr 0xe19b8000"). + +Even more, the Linux starts and runs, but it crashes only when I'm trying to run the GIMP, between splash screen and main interface appearance. + +I know that 1) Android is not officially supported and 2) 32-bit hosts were considered deprecated recently, but maybe it's possible to do something with these crashes? + +Recent master (https://gitlab.com/qemu-project/qemu/-/commit/5692a39f329413a00020a61fff95aff6b9884a73) doesn't work as well. +All 8.0.x Arm64 builds are runnable. + +Thanks in advance. diff --git a/gitlab/issues_text/target_missing/host_arm/accel_TCG/2295 b/gitlab/issues_text/target_missing/host_arm/accel_TCG/2295 new file mode 100644 index 000000000..a525f8cd0 --- /dev/null +++ b/gitlab/issues_text/target_missing/host_arm/accel_TCG/2295 @@ -0,0 +1,4 @@ +Support Apple Silicon acceleration for x86 / x86_64 guests +Additional information: +* [Top-level discussion on UTM downstream](https://github.com/utmapp/UTM/issues/5460) +* [Discussion on memory access instructions on UTM downstream](https://github.com/utmapp/UTM/issues/2366) |