diff options
Diffstat (limited to 'results/classifier/118/TCG-ppc')
| -rw-r--r-- | results/classifier/118/TCG-ppc/1065 | 65 | ||||
| -rw-r--r-- | results/classifier/118/TCG-ppc/1832353 | 117 | ||||
| -rw-r--r-- | results/classifier/118/TCG-ppc/212 | 61 | ||||
| -rw-r--r-- | results/classifier/118/TCG-ppc/2456 | 61 |
4 files changed, 304 insertions, 0 deletions
diff --git a/results/classifier/118/TCG-ppc/1065 b/results/classifier/118/TCG-ppc/1065 new file mode 100644 index 00000000..ac6c6e00 --- /dev/null +++ b/results/classifier/118/TCG-ppc/1065 @@ -0,0 +1,65 @@ +TCG: 0.982 +ppc: 0.952 +device: 0.886 +graphic: 0.829 +architecture: 0.785 +permissions: 0.723 +network: 0.720 +register: 0.626 +risc-v: 0.600 +debug: 0.512 +vnc: 0.511 +performance: 0.428 +socket: 0.422 +kernel: 0.381 +mistranslation: 0.369 +x86: 0.361 +files: 0.350 +i386: 0.342 +boot: 0.329 +arm: 0.325 +semantic: 0.313 +PID: 0.302 +VMM: 0.237 +assembly: 0.156 +virtual: 0.113 +user-level: 0.103 +peripherals: 0.065 +hypervisor: 0.064 +KVM: 0.059 +-------------------- +TCG: 0.975 +debug: 0.856 +kernel: 0.853 +register: 0.107 +hypervisor: 0.060 +assembly: 0.050 +virtual: 0.036 +performance: 0.036 +files: 0.028 +architecture: 0.027 +PID: 0.024 +permissions: 0.019 +semantic: 0.015 +device: 0.012 +x86: 0.011 +user-level: 0.010 +i386: 0.007 +risc-v: 0.006 +VMM: 0.004 +network: 0.004 +mistranslation: 0.003 +peripherals: 0.003 +boot: 0.003 +arm: 0.003 +socket: 0.002 +KVM: 0.002 +graphic: 0.002 +vnc: 0.002 +ppc: 0.001 + +cputlb: uninitialized local variable in tlb_set_page_with_attrs cause SIGSEGV when a CPU access an unmapped IOMMU page +Description of problem: +When a TCG cpu accesses an unmapped page within an IOMMU region that causes a translation fault, QEMU SIGSEGVs in `io_readx`. +The reason was that in `address_space_translate_for_iotlb`, `xlat` is not set on a permission fault. +As a result, `xlat` in `tlb_set_page_with_attr` is uninitialized. This in turn causes various mis-calculation and eventually crashes in `io_readx`. diff --git a/results/classifier/118/TCG-ppc/1832353 b/results/classifier/118/TCG-ppc/1832353 new file mode 100644 index 00000000..237d2473 --- /dev/null +++ b/results/classifier/118/TCG-ppc/1832353 @@ -0,0 +1,117 @@ +TCG: 0.879 +architecture: 0.868 +mistranslation: 0.824 +ppc: 0.802 +device: 0.784 +peripherals: 0.705 +files: 0.699 +arm: 0.636 +user-level: 0.632 +permissions: 0.625 +network: 0.621 +PID: 0.618 +debug: 0.601 +graphic: 0.592 +kernel: 0.592 +socket: 0.586 +vnc: 0.569 +performance: 0.567 +x86: 0.512 +semantic: 0.511 +i386: 0.476 +boot: 0.470 +register: 0.466 +risc-v: 0.446 +VMM: 0.443 +hypervisor: 0.364 +virtual: 0.341 +KVM: 0.245 +assembly: 0.213 +-------------------- +arm: 0.981 +TCG: 0.892 +debug: 0.146 +performance: 0.058 +virtual: 0.039 +hypervisor: 0.037 +files: 0.032 +kernel: 0.014 +semantic: 0.013 +user-level: 0.010 +architecture: 0.008 +register: 0.007 +PID: 0.007 +assembly: 0.006 +network: 0.006 +mistranslation: 0.003 +device: 0.003 +boot: 0.003 +socket: 0.002 +peripherals: 0.002 +graphic: 0.002 +vnc: 0.001 +VMM: 0.001 +risc-v: 0.001 +permissions: 0.001 +ppc: 0.000 +x86: 0.000 +KVM: 0.000 +i386: 0.000 + +cpu_exec: Assertion !have_mmap_lock() failed + +Hi, + +I have isolated a testcase from the GCC testsuite (actually gfortran, test proc_ptr_51.f90) which produces tons of: + +qemu-arm: /home/christophe.lyon/src/qemu/accel/tcg/cpu-exec.c:701: cpu_exec: Assertion `!have_mmap_lock()' failed. + +including with master qemu as of today. + +I'm attaching a tarball containing: +qemu-assert: +cmd lib proc_ptr_51.exe + +qemu-assert/lib: +ld-linux-armhf.so.3 libc.so.6 libgcc_s.so.1 libgfortran.so.5 libm.so.6 + +where cmd is the basic command used to launch the test & reproduce the failure. + +Note that the test or the generated may actually be buggy: I have reported failures on native aarch64 and arm machines. Yet, qemu should not fail with a loop of asserts. + + + +What version of qemu where you running? My HEAD is failing in a different way. + +It's fairly recent: +qemu-arm version 4.0.50 (v4.0.0-1215-ga578cdf-dirty) +Copyright (c) 2003-2019 Fabrice Bellard and the QEMU Project developers + +commit a578cdfbdd8f9beff5ced52b7826ddb1669abbbf +Merge: 19735c8 43b3952 +Author: Peter Maydell <email address hidden> +Date: Mon Jun 10 16:09:19 2019 +0100 + + Merge remote-tracking branch 'remotes/rth/tags/pull-tcg-20190610' into staging + + Move softmmu tlb into CPUNegativeOffsetState + +configured with: +--target-list=arm-softmmu,arm-linux-user,aarch64-softmmu,aarch64-linux-user --enable-debug + +Confirmed. The exact failure mode depends on debugging enabled or not. + +The test case is "buggy" in the sense that it makes a call to a NULL +function pointer, and the failure happens while trying to translate +the instructions at address 0. + +That said, the correct behaviour for QEMU is a SIGSEGV delivered to +the guest, not an assertion failure. + +The fix for this bug is now in master and will be in QEMU 4.1. + + +See series: https://lists.gnu.org/archive/html/qemu-devel/2019-07/msg02189.html + +https://git.qemu.org/?p=qemu.git;a=commitdiff;h=52ba13f042714c4086416 + diff --git a/results/classifier/118/TCG-ppc/212 b/results/classifier/118/TCG-ppc/212 new file mode 100644 index 00000000..8b8b83c1 --- /dev/null +++ b/results/classifier/118/TCG-ppc/212 @@ -0,0 +1,61 @@ +TCG: 0.966 +ppc: 0.808 +device: 0.721 +performance: 0.671 +graphic: 0.323 +files: 0.310 +arm: 0.231 +architecture: 0.218 +debug: 0.182 +network: 0.170 +hypervisor: 0.165 +permissions: 0.164 +virtual: 0.159 +user-level: 0.152 +mistranslation: 0.149 +register: 0.130 +risc-v: 0.121 +boot: 0.100 +peripherals: 0.099 +semantic: 0.094 +PID: 0.037 +vnc: 0.029 +socket: 0.015 +kernel: 0.013 +assembly: 0.010 +VMM: 0.008 +x86: 0.003 +i386: 0.002 +KVM: 0.002 +-------------------- +TCG: 0.979 +ppc: 0.969 +device: 0.509 +user-level: 0.082 +virtual: 0.071 +debug: 0.056 +performance: 0.014 +assembly: 0.007 +files: 0.006 +PID: 0.006 +semantic: 0.003 +peripherals: 0.002 +graphic: 0.002 +hypervisor: 0.001 +arm: 0.001 +socket: 0.000 +register: 0.000 +KVM: 0.000 +architecture: 0.000 +network: 0.000 +risc-v: 0.000 +mistranslation: 0.000 +VMM: 0.000 +kernel: 0.000 +boot: 0.000 +x86: 0.000 +vnc: 0.000 +permissions: 0.000 +i386: 0.000 + +ppc64 TCG application crashes diff --git a/results/classifier/118/TCG-ppc/2456 b/results/classifier/118/TCG-ppc/2456 new file mode 100644 index 00000000..b63b4fcb --- /dev/null +++ b/results/classifier/118/TCG-ppc/2456 @@ -0,0 +1,61 @@ +ppc: 0.950 +performance: 0.837 +TCG: 0.832 +device: 0.668 +peripherals: 0.470 +architecture: 0.423 +graphic: 0.339 +network: 0.331 +arm: 0.309 +semantic: 0.293 +debug: 0.255 +user-level: 0.243 +boot: 0.223 +hypervisor: 0.135 +register: 0.120 +mistranslation: 0.093 +VMM: 0.092 +permissions: 0.091 +virtual: 0.078 +vnc: 0.043 +PID: 0.040 +risc-v: 0.035 +socket: 0.031 +files: 0.017 +kernel: 0.015 +assembly: 0.014 +KVM: 0.004 +i386: 0.004 +x86: 0.002 +-------------------- +ppc: 0.991 +TCG: 0.885 +user-level: 0.100 +performance: 0.100 +debug: 0.068 +files: 0.031 +virtual: 0.026 +semantic: 0.007 +PID: 0.007 +assembly: 0.005 +architecture: 0.005 +kernel: 0.004 +VMM: 0.004 +device: 0.003 +register: 0.003 +boot: 0.003 +graphic: 0.002 +peripherals: 0.002 +risc-v: 0.001 +network: 0.001 +hypervisor: 0.001 +socket: 0.000 +permissions: 0.000 +KVM: 0.000 +mistranslation: 0.000 +vnc: 0.000 +x86: 0.000 +i386: 0.000 +arm: 0.000 + +check-tcg multi-threaded tests fail on ppc64 on clang-user CI job |