summary refs log tree commit diff stats
path: root/results/classifier/zero-shot/108/other/657329
blob: 835bc34e12e5302f679f0d0ac091762e903ee6d2 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
device: 0.693
graphic: 0.618
performance: 0.572
semantic: 0.569
other: 0.482
PID: 0.467
files: 0.382
socket: 0.261
vnc: 0.204
boot: 0.182
network: 0.160
permissions: 0.151
KVM: 0.105
debug: 0.103

APIC unusable on QEMU

The APIC is unusable with QEMU using x86-64 system emulation.  Problem exists in the latest stable QEMU 0.12.5 as well as the latest git head.  I am using Mac OS X 10.6, 64-bit version of QEMU.

The QEMU binary was configured with:

 ./configure --target-list=i386-softmmu,x86_64-softmmubck-i-search: conf_      

Problem is that the hw/apic.c file (as well as a few other naughty files) rely on the cpu_single_env global - which is set to NULL in cpu-exec.c.

Below is a test reading the local APIC version register:

Before taking it out:

(qemu) xp 0xfee00030
00000000fee00030: 0x00000000
(qemu)

After:

(qemu) xp 0xfee00030
00000000fee00030: 0x00050011
(qemu)

Quick fix below.  I don't know if there are any side effects with this, if this is OK maybe we can fix it like this for the stable versions and fix the HEAD to not rely on the cpu_single_env global.

diff --git a/cpu-exec.c b/cpu-exec.c
index dbdfdcc..3e966d7 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -674,7 +674,17 @@ int cpu_exec(CPUState *env1)
     env = (void *) saved_env_reg;
 
     /* fail safe : never use cpu_single_env outside cpu_exec() */
+#warning fixup devices which rely on this
+#if 0
+    /*
+     * Hello.  This is wrapped around an #if 0 ... #endif because that's
+     * what should happen.  However, certain naughty devices (like the APIC
+     * for instance, and a few others), access this global variable.
+     *
+     * So this is here for now ... until we fix up those devices.
+     */
     cpu_single_env = NULL;
+#endif
     return ret;
 }

Looking through old bug tickets... can you still reproduce this issue with the latest version of QEMU? Or could we close this ticket nowadays?


[Expired for QEMU because there has been no activity for 60 days.]