pool/xen
SHA256
2
0
Fork 1
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

- Update to Xen Version 4.4.1 FCS

Browse Source 
xen-4.4.1-testing-src.tar.bz2
- Dropped patches now contained in tarball
  53d7b781-x86-cpu-undo-BIOS-CPUID-max_leaf-limit-earlier.patch
  53df71c7-lz4-check-for-underruns.patch
  53e47d6b-x86_emulate-properly-do-IP-updates-and-other-side-effects.patch

- bnc#882089 - Windows 2012 R2 fails to boot up with greater than
  60 vcpus 
  53df727b-x86-HVM-extend-LAPIC-shortcuts-around-P2M-lookups.patch
  53e8be5f-x86-vHPET-use-rwlock-instead-of-simple-one.patch
  53ff3659-x86-consolidate-boolean-inputs-in-hvm-and-p2m.patch
  53ff36ae-x86-hvm-treat-non-insn-fetch-NPF-also-as-read-violations.patch
  53ff36d5-x86-mem_event-deliver-gla-fault-EPT-violation-information.patch
  54005472-EPT-utilize-GLA-GPA-translation-known-for-certain-faults.patch
- Upstream patches from Jan
  53f737b1-VMX-fix-DebugCtl-MSR-clearing.patch
  53f7386d-x86-irq-process-softirqs-in-irq-keyhandlers.patch
  53ff3716-x86-ats-Disable-Address-Translation-Services-by-default.patch
  53ff3899-x86-NMI-allow-processing-unknown-NMIs-with-watchdog.patch

- bnc#864801 - VUL-0: CVE-2013-4540: qemu: zaurus: buffer overrun
  on invalid state load
  CVE-2013-4540-qemu.patch

OBS-URL: https://build.opensuse.org/package/show/Virtualization/xen?expand=0&rev=328
This commit is contained in:
Charles Arnold 2014-09-03 21:59:31 +00:00 committed by Git OBS Bridge
parent d757dbb7b0
commit 99885eadf2
19 changed files with 1437 additions and 498 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
53d7b781-x86-cpu-undo-BIOS-CPUID-max_leaf-limit-earlier.patch
View File

@ -1,46 +0,0 @@
# Commit a1ac4cf52e38386bac7ac3440c7da0099662ca5c
# Date 2014-07-29 17:02:25 +0200
# Author Andrew Cooper <andrew.cooper3@citrix.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/cpu: undo BIOS CPUID max_leaf limit before querying for features
If IA32_MISC_ENABLE[22] is set by the BIOS, CPUID.0.EAX will be limited to 3.
Lift this limit before considering whether to query CPUID.7[ECX=0].EBX for
features.
Without this change, dom0 is able to see this feature leaf (as the limit was
subsequently lifted), and will set features appropriately in HVM domain cpuid
policies.
The specific bug XenServer observed was the advertisement of the FSGSBASE
feature, but an inability to set CR4.FSGSBASE as Xen considered the bit to be
reserved as cpu_has_fsgsbase incorrectly evaluated as false.
This is a regression introduced by c/s 44e24f8567 "x86: don't call
generic_identify() redundantly" where the redundant call actually resampled
CPUID.7[ECX=0] properly to obtain the feature flags.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/arch/x86/cpu/common.c
+++ b/xen/arch/x86/cpu/common.c
@@ -234,6 +234,9 @@ static void __cpuinit generic_identify(s
paddr_bits = cpuid_eax(0x80000008) & 0xff;
}
+ /* Might lift BIOS max_leaf=3 limit. */
+ early_intel_workaround(c);
+
/* Intel-defined flags: level 0x00000007 */
if ( c->cpuid_level >= 0x00000007 ) {
u32 dummy;
@@ -241,8 +244,6 @@ static void __cpuinit generic_identify(s
c->x86_capability[X86_FEATURE_FSGSBASE / 32] = ebx;
}
- early_intel_workaround(c);
-
#ifdef CONFIG_X86_HT
c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff;
#endif

54
53df71c7-lz4-check-for-underruns.patch
View File

@ -1,54 +0,0 @@
# Commit 9143a6c55ef7e8f630857cb08c03844d372c2345
# Date 2014-08-04 13:43:03 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
lz4: check for underruns
While overruns are already being taken care of, underruns (resulting
from overflows in the respective "op + length" (or similar) operations
weren't.
This is CVE-2014-4611.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
--- a/xen/common/lz4/decompress.c
+++ b/xen/common/lz4/decompress.c
@@ -84,6 +84,8 @@ static int INIT lz4_uncompress(const uns
ip += length;
break; /* EOF */
}
+ if (unlikely((unsigned long)cpy < (unsigned long)op))
+ goto _output_error;
LZ4_WILDCOPY(ip, op, cpy);
ip -= (op - cpy);
op = cpy;
@@ -142,6 +144,8 @@ static int INIT lz4_uncompress(const uns
goto _output_error;
continue;
}
+ if (unlikely((unsigned long)cpy < (unsigned long)op))
+ goto _output_error;
LZ4_SECURECOPY(ref, op, cpy);
op = cpy; /* correction */
}
@@ -207,6 +211,8 @@ static int lz4_uncompress_unknownoutputs
op += length;
break;/* Necessarily EOF, due to parsing restrictions */
}
+ if (unlikely((unsigned long)cpy < (unsigned long)op))
+ goto _output_error;
LZ4_WILDCOPY(ip, op, cpy);
ip -= (op - cpy);
op = cpy;
@@ -270,6 +276,8 @@ static int lz4_uncompress_unknownoutputs
goto _output_error;
continue;
}
+ if (unlikely((unsigned long)cpy < (unsigned long)op))
+ goto _output_error;
LZ4_SECURECOPY(ref, op, cpy);
op = cpy; /* correction */
}

2
53df727b-x86-HVM-extend-LAPIC-shortcuts-around-P2M-lookups.patch
View File

@ -1,3 +1,5 @@
References: bnc#882089
# Commit fd1863847af15c3676348447755e1a1801f9d394 # Commit fd1863847af15c3676348447755e1a1801f9d394
# Date 2014-08-04 13:46:03 +0200 # Date 2014-08-04 13:46:03 +0200
# Author Jan Beulich <jbeulich@suse.com> # Author Jan Beulich <jbeulich@suse.com>

378
53e47d6b-x86_emulate-properly-do-IP-updates-and-other-side-effects.patch
View File

@ -1,378 +0,0 @@
# Commit 3af450fd2d9403f208d3ac6459716f027b8597ad
# Date 2014-08-08 09:34:03 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86_emulate: properly do IP updates and other side effects on success
The two MMX/SSE/AVX code blocks failed to update IP properly, and these
as well as get_reg_refix(), which "manually" updated IP so far, failed
to do the TF and RF processing needed at the end of successfully
emulated instructions.
Fix the test utility at once to check IP is properly getting updated,
and while at it macroize the respective code quite a bit, hopefully
making it easier to add further tests when the need arises.
Reported-by: Andrei LUTAS <vlutas@bitdefender.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Tested-by: Razvan Cojocaru <rcojocaru@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.cooper@citrix.com>
--- a/tools/tests/x86_emulator/test_x86_emulator.c
+++ b/tools/tests/x86_emulator/test_x86_emulator.c
@@ -597,23 +597,32 @@ int main(int argc, char **argv)
printf("skipped\n");
#endif
+#define decl_insn(which) extern const unsigned char which[], which##_len[]
+#define put_insn(which, insn) ".pushsection .test, \"ax\", @progbits\n" \
+ #which ": " insn "\n" \
+ ".equ " #which "_len, .-" #which "\n" \
+ ".popsection"
+#define set_insn(which) (regs.eip = (unsigned long)memcpy(instr, which, \
+ (unsigned long)which##_len))
+#define check_eip(which) (regs.eip == (unsigned long)instr + \
+ (unsigned long)which##_len)
+
printf("%-40s", "Testing movq %mm3,(%ecx)...");
if ( stack_exec && cpu_has_mmx )
{
- extern const unsigned char movq_to_mem[];
+ decl_insn(movq_to_mem);
asm volatile ( "pcmpeqb %%mm3, %%mm3\n"
- ".pushsection .test, \"a\", @progbits\n"
- "movq_to_mem: movq %%mm3, (%0)\n"
- ".popsection" :: "c" (NULL) );
+ put_insn(movq_to_mem, "movq %%mm3, (%0)")
+ :: "c" (NULL) );
- memcpy(instr, movq_to_mem, 15);
+ set_insn(movq_to_mem);
memset(res, 0x33, 64);
memset(res + 8, 0xff, 8);
- regs.eip = (unsigned long)&instr[0];
regs.ecx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) )
+ if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) ||
+ !check_eip(movq_to_mem) )
goto fail;
printf("okay\n");
}
@@ -623,19 +632,17 @@ int main(int argc, char **argv)
printf("%-40s", "Testing movq (%edx),%mm5...");
if ( stack_exec && cpu_has_mmx )
{
- extern const unsigned char movq_from_mem[];
+ decl_insn(movq_from_mem);
asm volatile ( "pcmpgtb %%mm5, %%mm5\n"
- ".pushsection .test, \"a\", @progbits\n"
- "movq_from_mem: movq (%0), %%mm5\n"
- ".popsection" :: "d" (NULL) );
+ put_insn(movq_from_mem, "movq (%0), %%mm5")
+ :: "d" (NULL) );
- memcpy(instr, movq_from_mem, 15);
- regs.eip = (unsigned long)&instr[0];
+ set_insn(movq_from_mem);
regs.ecx = 0;
regs.edx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( rc != X86EMUL_OKAY )
+ if ( rc != X86EMUL_OKAY || !check_eip(movq_from_mem) )
goto fail;
asm ( "pcmpeqb %%mm3, %%mm3\n\t"
"pcmpeqb %%mm5, %%mm3\n\t"
@@ -650,20 +657,19 @@ int main(int argc, char **argv)
printf("%-40s", "Testing movdqu %xmm2,(%ecx)...");
if ( stack_exec && cpu_has_sse2 )
{
- extern const unsigned char movdqu_to_mem[];
+ decl_insn(movdqu_to_mem);
asm volatile ( "pcmpeqb %%xmm2, %%xmm2\n"
- ".pushsection .test, \"a\", @progbits\n"
- "movdqu_to_mem: movdqu %%xmm2, (%0)\n"
- ".popsection" :: "c" (NULL) );
+ put_insn(movdqu_to_mem, "movdqu %%xmm2, (%0)")
+ :: "c" (NULL) );
- memcpy(instr, movdqu_to_mem, 15);
+ set_insn(movdqu_to_mem);
memset(res, 0x55, 64);
memset(res + 8, 0xff, 16);
- regs.eip = (unsigned long)&instr[0];
regs.ecx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) )
+ if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) ||
+ !check_eip(movdqu_to_mem) )
goto fail;
printf("okay\n");
}
@@ -673,19 +679,17 @@ int main(int argc, char **argv)
printf("%-40s", "Testing movdqu (%edx),%xmm4...");
if ( stack_exec && cpu_has_sse2 )
{
- extern const unsigned char movdqu_from_mem[];
+ decl_insn(movdqu_from_mem);
asm volatile ( "pcmpgtb %%xmm4, %%xmm4\n"
- ".pushsection .test, \"a\", @progbits\n"
- "movdqu_from_mem: movdqu (%0), %%xmm4\n"
- ".popsection" :: "d" (NULL) );
+ put_insn(movdqu_from_mem, "movdqu (%0), %%xmm4")
+ :: "d" (NULL) );
- memcpy(instr, movdqu_from_mem, 15);
- regs.eip = (unsigned long)&instr[0];
+ set_insn(movdqu_from_mem);
regs.ecx = 0;
regs.edx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( rc != X86EMUL_OKAY )
+ if ( rc != X86EMUL_OKAY || !check_eip(movdqu_from_mem) )
goto fail;
asm ( "pcmpeqb %%xmm2, %%xmm2\n\t"
"pcmpeqb %%xmm4, %%xmm2\n\t"
@@ -700,21 +704,20 @@ int main(int argc, char **argv)
printf("%-40s", "Testing vmovdqu %ymm2,(%ecx)...");
if ( stack_exec && cpu_has_avx )
{
- extern const unsigned char vmovdqu_to_mem[];
+ decl_insn(vmovdqu_to_mem);
asm volatile ( "vpcmpeqb %%xmm2, %%xmm2, %%xmm2\n"
- ".pushsection .test, \"a\", @progbits\n"
- "vmovdqu_to_mem: vmovdqu %%ymm2, (%0)\n"
- ".popsection" :: "c" (NULL) );
+ put_insn(vmovdqu_to_mem, "vmovdqu %%ymm2, (%0)")
+ :: "c" (NULL) );
- memcpy(instr, vmovdqu_to_mem, 15);
+ set_insn(vmovdqu_to_mem);
memset(res, 0x55, 128);
memset(res + 16, 0xff, 16);
memset(res + 20, 0x00, 16);
- regs.eip = (unsigned long)&instr[0];
regs.ecx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 16, 64) )
+ if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 16, 64) ||
+ !check_eip(vmovdqu_to_mem) )
goto fail;
printf("okay\n");
}
@@ -724,7 +727,7 @@ int main(int argc, char **argv)
printf("%-40s", "Testing vmovdqu (%edx),%ymm4...");
if ( stack_exec && cpu_has_avx )
{
- extern const unsigned char vmovdqu_from_mem[];
+ decl_insn(vmovdqu_from_mem);
#if 0 /* Don't use AVX2 instructions for now */
asm volatile ( "vpcmpgtb %%ymm4, %%ymm4, %%ymm4\n"
@@ -732,17 +735,15 @@ int main(int argc, char **argv)
asm volatile ( "vpcmpgtb %%xmm4, %%xmm4, %%xmm4\n\t"
"vinsertf128 $1, %%xmm4, %%ymm4, %%ymm4\n"
#endif
- ".pushsection .test, \"a\", @progbits\n"
- "vmovdqu_from_mem: vmovdqu (%0), %%ymm4\n"
- ".popsection" :: "d" (NULL) );
+ put_insn(vmovdqu_from_mem, "vmovdqu (%0), %%ymm4")
+ :: "d" (NULL) );
- memcpy(instr, vmovdqu_from_mem, 15);
+ set_insn(vmovdqu_from_mem);
memset(res + 4, 0xff, 16);
- regs.eip = (unsigned long)&instr[0];
regs.ecx = 0;
regs.edx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( rc != X86EMUL_OKAY )
+ if ( rc != X86EMUL_OKAY || !check_eip(vmovdqu_from_mem) )
goto fail;
#if 0 /* Don't use AVX2 instructions for now */
asm ( "vpcmpeqb %%ymm2, %%ymm2, %%ymm2\n\t"
@@ -769,20 +770,19 @@ int main(int argc, char **argv)
memset(res + 10, 0x66, 8);
if ( stack_exec && cpu_has_sse2 )
{
- extern const unsigned char movsd_to_mem[];
+ decl_insn(movsd_to_mem);
asm volatile ( "movlpd %0, %%xmm5\n\t"
"movhpd %0, %%xmm5\n"
- ".pushsection .test, \"a\", @progbits\n"
- "movsd_to_mem: movsd %%xmm5, (%1)\n"
- ".popsection" :: "m" (res[10]), "c" (NULL) );
+ put_insn(movsd_to_mem, "movsd %%xmm5, (%1)")
+ :: "m" (res[10]), "c" (NULL) );
- memcpy(instr, movsd_to_mem, 15);
- regs.eip = (unsigned long)&instr[0];
+ set_insn(movsd_to_mem);
regs.ecx = (unsigned long)(res + 2);
regs.edx = 0;
rc = x86_emulate(&ctxt, &emulops);
- if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) )
+ if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) ||
+ !check_eip(movsd_to_mem) )
goto fail;
printf("okay\n");
}
@@ -795,19 +795,17 @@ int main(int argc, char **argv)
printf("%-40s", "Testing movaps (%edx),%xmm7...");
if ( stack_exec && cpu_has_sse )
{
- extern const unsigned char movaps_from_mem[];
+ decl_insn(movaps_from_mem);
asm volatile ( "xorps %%xmm7, %%xmm7\n"
- ".pushsection .test, \"a\", @progbits\n"
- "movaps_from_mem: movaps (%0), %%xmm7\n"
- ".popsection" :: "d" (NULL) );
+ put_insn(movaps_from_mem, "movaps (%0), %%xmm7")
+ :: "d" (NULL) );
- memcpy(instr, movaps_from_mem, 15);
- regs.eip = (unsigned long)&instr[0];
+ set_insn(movaps_from_mem);
regs.ecx = 0;
regs.edx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( rc != X86EMUL_OKAY )
+ if ( rc != X86EMUL_OKAY || !check_eip(movaps_from_mem) )
goto fail;
asm ( "cmpeqps %1, %%xmm7\n\t"
"movmskps %%xmm7, %0" : "=r" (rc) : "m" (res[8]) );
@@ -823,19 +821,18 @@ int main(int argc, char **argv)
memset(res + 10, 0x77, 8);
if ( stack_exec && cpu_has_avx )
{
- extern const unsigned char vmovsd_to_mem[];
+ decl_insn(vmovsd_to_mem);
asm volatile ( "vbroadcastsd %0, %%ymm5\n"
- ".pushsection .test, \"a\", @progbits\n"
- "vmovsd_to_mem: vmovsd %%xmm5, (%1)\n"
- ".popsection" :: "m" (res[10]), "c" (NULL) );
+ put_insn(vmovsd_to_mem, "vmovsd %%xmm5, (%1)")
+ :: "m" (res[10]), "c" (NULL) );
- memcpy(instr, vmovsd_to_mem, 15);
- regs.eip = (unsigned long)&instr[0];
+ set_insn(vmovsd_to_mem);
regs.ecx = (unsigned long)(res + 2);
regs.edx = 0;
rc = x86_emulate(&ctxt, &emulops);
- if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) )
+ if ( (rc != X86EMUL_OKAY) || memcmp(res, res + 8, 32) ||
+ !check_eip(vmovsd_to_mem) )
goto fail;
printf("okay\n");
}
@@ -848,19 +845,17 @@ int main(int argc, char **argv)
printf("%-40s", "Testing vmovaps (%edx),%ymm7...");
if ( stack_exec && cpu_has_avx )
{
- extern const unsigned char vmovaps_from_mem[];
+ decl_insn(vmovaps_from_mem);
asm volatile ( "vxorps %%ymm7, %%ymm7, %%ymm7\n"
- ".pushsection .test, \"a\", @progbits\n"
- "vmovaps_from_mem: vmovaps (%0), %%ymm7\n"
- ".popsection" :: "d" (NULL) );
+ put_insn(vmovaps_from_mem, "vmovaps (%0), %%ymm7")
+ :: "d" (NULL) );
- memcpy(instr, vmovaps_from_mem, 15);
- regs.eip = (unsigned long)&instr[0];
+ set_insn(vmovaps_from_mem);
regs.ecx = 0;
regs.edx = (unsigned long)res;
rc = x86_emulate(&ctxt, &emulops);
- if ( rc != X86EMUL_OKAY )
+ if ( rc != X86EMUL_OKAY || !check_eip(vmovaps_from_mem) )
goto fail;
asm ( "vcmpeqps %1, %%ymm7, %%ymm0\n\t"
"vmovmskps %%ymm0, %0" : "=r" (rc) : "m" (res[8]) );
@@ -871,6 +866,11 @@ int main(int argc, char **argv)
else
printf("skipped\n");
+#undef decl_insn
+#undef put_insn
+#undef set_insn
+#undef check_eip
+
for ( j = 1; j <= 2; j++ )
{
#if defined(__i386__)
--- a/xen/arch/x86/x86_emulate/x86_emulate.c
+++ b/xen/arch/x86/x86_emulate/x86_emulate.c
@@ -720,29 +720,26 @@ do{ uint8_t stub[] = { _bytes, 0xc3 };
put_fpu(&fic); \
} while (0)
-static unsigned long __get_rep_prefix(
- struct cpu_user_regs *int_regs,
- struct cpu_user_regs *ext_regs,
+static unsigned long _get_rep_prefix(
+ const struct cpu_user_regs *int_regs,
int ad_bytes)
{
- unsigned long ecx = ((ad_bytes == 2) ? (uint16_t)int_regs->ecx :
- (ad_bytes == 4) ? (uint32_t)int_regs->ecx :
- int_regs->ecx);
-
- /* Skip the instruction if no repetitions are required. */
- if ( ecx == 0 )
- ext_regs->eip = int_regs->eip;
-
- return ecx;
+ return (ad_bytes == 2) ? (uint16_t)int_regs->ecx :
+ (ad_bytes == 4) ? (uint32_t)int_regs->ecx :
+ int_regs->ecx;
}
#define get_rep_prefix() ({ \
unsigned long max_reps = 1; \
if ( rep_prefix() ) \
- max_reps = __get_rep_prefix(&_regs, ctxt->regs, ad_bytes); \
+ max_reps = _get_rep_prefix(&_regs, ad_bytes); \
if ( max_reps == 0 ) \
- goto done; \
- max_reps; \
+ { \
+ /* Skip the instruction if no repetitions are required. */ \
+ dst.type = OP_NONE; \
+ goto writeback; \
+ } \
+ max_reps; \
})
static void __put_rep_prefix(
@@ -3921,7 +3918,8 @@ x86_emulate(
if ( !rc && (b & 1) && (ea.type == OP_MEM) )
rc = ops->write(ea.mem.seg, ea.mem.off, mmvalp,
ea.bytes, ctxt);
- goto done;
+ dst.type = OP_NONE;
+ break;
}
case 0x20: /* mov cr,reg */
@@ -4188,7 +4186,8 @@ x86_emulate(
if ( !rc && (b != 0x6f) && (ea.type == OP_MEM) )
rc = ops->write(ea.mem.seg, ea.mem.off, mmvalp,
ea.bytes, ctxt);
- goto done;
+ dst.type = OP_NONE;
+ break;
}
case 0x80 ... 0x8f: /* jcc (near) */ {

194
53e8be5f-x86-vHPET-use-rwlock-instead-of-simple-one.patch Normal file
View File

@ -0,0 +1,194 @@
References: bnc#882089
# Commit ded2100990d1688b96c2edc7221887c56c1a8e04
# Date 2014-08-11 15:00:15 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/vHPET: use rwlock instead of simple one
This namely benefits guests heavily reading the main counter, but not
touching the HPET much otherwise. Note that due to the way
hpet_get_comparator() works hpet_read() has to special cases reads from
the comparator registers and use a write lock there instead of the read
one used for all other registers.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/arch/x86/hvm/hpet.c
+++ b/xen/arch/x86/hvm/hpet.c
@@ -75,7 +75,7 @@
static inline uint64_t hpet_read_maincounter(HPETState *h)
{
- ASSERT(spin_is_locked(&h->lock));
+ ASSERT(rw_is_locked(&h->lock));
if ( hpet_enabled(h) )
return guest_time_hpet(h) + h->mc_offset;
@@ -88,6 +88,8 @@ static uint64_t hpet_get_comparator(HPET
uint64_t comparator;
uint64_t elapsed;
+ ASSERT(rw_is_write_locked(&h->lock));
+
comparator = h->hpet.comparator64[tn];
if ( timer_is_periodic(h, tn) )
{
@@ -172,16 +174,24 @@ static int hpet_read(
goto out;
}
- spin_lock(&h->lock);
+ result = addr < HPET_Tn_CMP(0) ||
+ ((addr - HPET_Tn_CMP(0)) % (HPET_Tn_CMP(1) - HPET_Tn_CMP(0))) > 7;
+ if ( result )
+ read_lock(&h->lock);
+ else
+ write_lock(&h->lock);
val = hpet_read64(h, addr);
+ if ( result )
+ read_unlock(&h->lock);
+ else
+ write_unlock(&h->lock);
+
result = val;
if ( length != 8 )
result = (val >> ((addr & 7) * 8)) & ((1ULL << (length * 8)) - 1);
- spin_unlock(&h->lock);
-
out:
*pval = result;
return X86EMUL_OKAY;
@@ -190,7 +200,7 @@ static int hpet_read(
static void hpet_stop_timer(HPETState *h, unsigned int tn)
{
ASSERT(tn < HPET_TIMER_NUM);
- ASSERT(spin_is_locked(&h->lock));
+ ASSERT(rw_is_write_locked(&h->lock));
destroy_periodic_time(&h->pt[tn]);
/* read the comparator to get it updated so a read while stopped will
* return the expected value. */
@@ -208,7 +218,7 @@ static void hpet_set_timer(HPETState *h,
unsigned int oneshot;
ASSERT(tn < HPET_TIMER_NUM);
- ASSERT(spin_is_locked(&h->lock));
+ ASSERT(rw_is_write_locked(&h->lock));
if ( (tn == 0) && (h->hpet.config & HPET_CFG_LEGACY) )
{
@@ -289,7 +299,7 @@ static int hpet_write(
if ( hpet_check_access_length(addr, length) != 0 )
goto out;
- spin_lock(&h->lock);
+ write_lock(&h->lock);
old_val = hpet_read64(h, addr);
new_val = val;
@@ -448,7 +458,7 @@ static int hpet_write(
#undef set_start_timer
#undef set_restart_timer
- spin_unlock(&h->lock);
+ write_unlock(&h->lock);
out:
return X86EMUL_OKAY;
@@ -473,7 +483,7 @@ static int hpet_save(struct domain *d, h
HPETState *hp = domain_vhpet(d);
int rc;
- spin_lock(&hp->lock);
+ write_lock(&hp->lock);
/* Write the proper value into the main counter */
hp->hpet.mc64 = hp->mc_offset + guest_time_hpet(hp);
@@ -507,7 +517,7 @@ static int hpet_save(struct domain *d, h
rec->timers[2].cmp = hp->hpet.comparator64[2];
}
- spin_unlock(&hp->lock);
+ write_unlock(&hp->lock);
return rc;
}
@@ -519,12 +529,12 @@ static int hpet_load(struct domain *d, h
uint64_t cmp;
int i;
- spin_lock(&hp->lock);
+ write_lock(&hp->lock);
/* Reload the HPET registers */
if ( _hvm_check_entry(h, HVM_SAVE_CODE(HPET), HVM_SAVE_LENGTH(HPET), 1) )
{
- spin_unlock(&hp->lock);
+ write_unlock(&hp->lock);
return -EINVAL;
}
@@ -564,7 +574,7 @@ static int hpet_load(struct domain *d, h
if ( timer_enabled(hp, i) )
hpet_set_timer(hp, i);
- spin_unlock(&hp->lock);
+ write_unlock(&hp->lock);
return 0;
}
@@ -578,7 +588,7 @@ void hpet_init(struct vcpu *v)
memset(h, 0, sizeof(HPETState));
- spin_lock_init(&h->lock);
+ rwlock_init(&h->lock);
h->stime_freq = S_TO_NS;
@@ -607,14 +617,14 @@ void hpet_deinit(struct domain *d)
int i;
HPETState *h = domain_vhpet(d);
- spin_lock(&h->lock);
+ write_lock(&h->lock);
if ( hpet_enabled(h) )
for ( i = 0; i < HPET_TIMER_NUM; i++ )
if ( timer_enabled(h, i) )
hpet_stop_timer(h, i);
- spin_unlock(&h->lock);
+ write_unlock(&h->lock);
}
void hpet_reset(struct domain *d)
--- a/xen/arch/x86/hvm/vpt.c
+++ b/xen/arch/x86/hvm/vpt.c
@@ -508,10 +508,10 @@ void pt_adjust_global_vcpu_target(struct
pt_adjust_vcpu(&pl_time->vrtc.pt, v);
spin_unlock(&pl_time->vrtc.lock);
- spin_lock(&pl_time->vhpet.lock);
+ write_lock(&pl_time->vhpet.lock);
for ( i = 0; i < HPET_TIMER_NUM; i++ )
pt_adjust_vcpu(&pl_time->vhpet.pt[i], v);
- spin_unlock(&pl_time->vhpet.lock);
+ write_unlock(&pl_time->vhpet.lock);
}
--- a/xen/include/asm-x86/hvm/vpt.h
+++ b/xen/include/asm-x86/hvm/vpt.h
@@ -96,7 +96,7 @@ typedef struct HPETState {
uint64_t hpet_to_ns_limit; /* max hpet ticks convertable to ns */
uint64_t mc_offset;
struct periodic_time pt[HPET_TIMER_NUM];
- spinlock_t lock;
+ rwlock_t lock;
} HPETState;
typedef struct RTCState {

39
53f737b1-VMX-fix-DebugCtl-MSR-clearing.patch Normal file
View File

@ -0,0 +1,39 @@
# Commit dfa625e15f3d6c374637f2bb789e1f444c2781c3
# Date 2014-08-22 14:29:37 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
VMX: fix DebugCtl MSR clearing
The previous shortcut was wrong, as it bypassed the necessary vmwrite:
All we really want to avoid if the guest writes zero is to add the MSR
to the host-load list.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -2170,8 +2170,6 @@ static int vmx_msr_write_intercept(unsig
int i, rc = 0;
uint64_t supported = IA32_DEBUGCTLMSR_LBR | IA32_DEBUGCTLMSR_BTF;
- if ( !msr_content )
- break;
if ( msr_content & ~supported )
{
/* Perhaps some other bits are supported in vpmu. */
@@ -2191,12 +2189,10 @@ static int vmx_msr_write_intercept(unsig
}
if ( (rc < 0) ||
- (vmx_add_host_load_msr(msr) < 0) )
+ (msr_content && (vmx_add_host_load_msr(msr) < 0)) )
hvm_inject_hw_exception(TRAP_machine_check, 0);
else
- {
__vmwrite(GUEST_IA32_DEBUGCTL, msr_content);
- }
break;
}

119
53f7386d-x86-irq-process-softirqs-in-irq-keyhandlers.patch Normal file
View File

@ -0,0 +1,119 @@
# Commit e13b3203990706db1313ec2aadd9a30b249ee793
# Date 2014-08-22 14:32:45 +0200
# Author Andrew Cooper <andrew.cooper3@citrix.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/irq: process softirqs in irq keyhandlers
Large machines with lots of interrupts can trip over the Xen watchdog.
Suggested-by: Santosh Jodh <Santosh.Jodh@citrix.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Tested-by: Santosh Jodh <Santosh.Jodh@citrix.com>
# Commit bd083922f9e78ed19ef98e7de372e5f568402ed3
# Date 2014-08-26 17:56:52 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/IO-APIC: don't process softirqs during early boot
Commit e13b320399 ("x86/irq: process softirqs in irq keyhandlers")
made this unconditional, but the boot time use of __print_IO_APIC()
(when "apic_verbosity=debug" was given) can't tolerate that.
Reported-by: Sander Eikelenboom <linux@eikelenboom.it>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Tested-by: Sander Eikelenboom <linux@eikelenboom.it>
--- a/xen/arch/x86/io_apic.c
+++ b/xen/arch/x86/io_apic.c
@@ -28,6 +28,7 @@
#include <xen/sched.h>
#include <xen/acpi.h>
#include <xen/keyhandler.h>
+#include <xen/softirq.h>
#include <asm/mc146818rtc.h>
#include <asm/smp.h>
#include <asm/desc.h>
@@ -1091,7 +1092,7 @@ static inline void UNEXPECTED_IO_APIC(vo
{
}
-static void /*__init*/ __print_IO_APIC(void)
+static void /*__init*/ __print_IO_APIC(bool_t boot)
{
int apic, i;
union IO_APIC_reg_00 reg_00;
@@ -1112,6 +1113,9 @@ static void /*__init*/ __print_IO_APIC(v
printk(KERN_INFO "testing the IO APIC.......................\n");
for (apic = 0; apic < nr_ioapics; apic++) {
+ if ( !boot )
+ process_pending_softirqs();
+
if (!nr_ioapic_entries[apic])
continue;
@@ -1215,6 +1219,10 @@ static void /*__init*/ __print_IO_APIC(v
printk(KERN_DEBUG "IRQ to pin mappings:\n");
for (i = 0; i < nr_irqs_gsi; i++) {
struct irq_pin_list *entry = irq_2_pin + i;
+
+ if ( !boot && !(i & 0x1f) )
+ process_pending_softirqs();
+
if (entry->pin < 0)
continue;
printk(KERN_DEBUG "IRQ%d ", irq_to_desc(i)->arch.vector);
@@ -1235,12 +1243,12 @@ static void /*__init*/ __print_IO_APIC(v
static void __init print_IO_APIC(void)
{
if (apic_verbosity != APIC_QUIET)
- __print_IO_APIC();
+ __print_IO_APIC(1);
}
static void _print_IO_APIC_keyhandler(unsigned char key)
{
- __print_IO_APIC();
+ __print_IO_APIC(0);
}
static struct keyhandler print_IO_APIC_keyhandler = {
.diagnostic = 1,
@@ -2454,6 +2462,9 @@ void dump_ioapic_irq_info(void)
for ( irq = 0; irq < nr_irqs_gsi; irq++ )
{
+ if ( !(irq & 0x1f) )
+ process_pending_softirqs();
+
entry = &irq_2_pin[irq];
if ( entry->pin == -1 )
continue;
--- a/xen/arch/x86/irq.c
+++ b/xen/arch/x86/irq.c
@@ -19,6 +19,7 @@
#include <xen/iommu.h>
#include <xen/symbols.h>
#include <xen/trace.h>
+#include <xen/softirq.h>
#include <xsm/xsm.h>
#include <asm/msi.h>
#include <asm/current.h>
@@ -2231,6 +2232,8 @@ static void dump_irqs(unsigned char key)
for ( irq = 0; irq < nr_irqs; irq++ )
{
+ if ( !(irq & 0x1f) )
+ process_pending_softirqs();
desc = irq_to_desc(irq);
@@ -2284,6 +2287,7 @@ static void dump_irqs(unsigned char key)
xfree(ssid);
}
+ process_pending_softirqs();
printk("Direct vector information:\n");
for ( i = FIRST_DYNAMIC_VECTOR; i < NR_VECTORS; ++i )
if ( direct_apic_vector[i] )

331
53ff3659-x86-consolidate-boolean-inputs-in-hvm-and-p2m.patch Normal file
View File

@ -0,0 +1,331 @@
References: bnc#882089
# Commit 3d4d4f9336159f3f77a7b480ce9984fd3ff7949f
# Date 2014-08-28 16:02:01 +0200
# Author Tamas K Lengyel <tamas.lengyel@zentific.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86: consolidate boolean inputs in hvm and p2m into a shared bitmap
This patch consolidates the boolean input parameters of
hvm_hap_nested_page_fault and p2m_mem_access_check into a common bitmap
and defines the bitmap members accordingly.
Signed-off-by: Tamas K Lengyel <tamas.lengyel@zentific.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Tim Deegan <tim@xen.org>
# Commit 24857896a30105b7947e2cd36d63768054538bbc
# Date 2014-09-03 15:06:06 +0200
# Author Andrew Cooper <andrew.cooper3@citrix.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/hvm: fix operator precedence bug introduced by 3d4d4f9336
Bitwise or has greater precedence than the ternary operator, making the result
of the expression a constant P2M_UNSHARE.
Coverity-ID: 1234633
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Don Slutz <dslutz@verizon.com>
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -1464,12 +1464,8 @@ void hvm_inject_page_fault(int errcode,
hvm_inject_trap(&trap);
}
-int hvm_hap_nested_page_fault(paddr_t gpa,
- bool_t gla_valid,
- unsigned long gla,
- bool_t access_r,
- bool_t access_w,
- bool_t access_x)
+int hvm_hap_nested_page_fault(paddr_t gpa, unsigned long gla,
+ struct npfec npfec)
{
unsigned long gfn = gpa >> PAGE_SHIFT;
p2m_type_t p2mt;
@@ -1498,8 +1494,11 @@ int hvm_hap_nested_page_fault(paddr_t gp
* into l1 guest if not fixable. The algorithm is
* the same as for shadow paging.
*/
- rv = nestedhvm_hap_nested_page_fault(v, &gpa,
- access_r, access_w, access_x);
+
+ rv = nestedhvm_hap_nested_page_fault(v, &gpa,
+ npfec.read_access,
+ npfec.write_access,
+ npfec.insn_fetch);
switch (rv) {
case NESTEDHVM_PAGEFAULT_DONE:
case NESTEDHVM_PAGEFAULT_RETRY:
@@ -1538,47 +1537,49 @@ int hvm_hap_nested_page_fault(paddr_t gp
p2m = p2m_get_hostp2m(v->domain);
mfn = get_gfn_type_access(p2m, gfn, &p2mt, &p2ma,
- P2M_ALLOC | (access_w ? P2M_UNSHARE : 0), NULL);
+ P2M_ALLOC | (npfec.write_access ? P2M_UNSHARE : 0),
+ NULL);
/* Check access permissions first, then handle faults */
if ( mfn_x(mfn) != INVALID_MFN )
{
- int violation = 0;
+ bool_t violation;
+
/* If the access is against the permissions, then send to mem_event */
- switch (p2ma)
+ switch (p2ma)
{
case p2m_access_n:
case p2m_access_n2rwx:
default:
- violation = access_r || access_w || access_x;
+ violation = npfec.read_access || npfec.write_access || npfec.insn_fetch;
break;
case p2m_access_r:
- violation = access_w || access_x;
+ violation = npfec.write_access || npfec.insn_fetch;
break;
case p2m_access_w:
- violation = access_r || access_x;
+ violation = npfec.read_access || npfec.insn_fetch;
break;
case p2m_access_x:
- violation = access_r || access_w;
+ violation = npfec.read_access || npfec.write_access;
break;
case p2m_access_rx:
case p2m_access_rx2rw:
- violation = access_w;
+ violation = npfec.write_access;
break;
case p2m_access_wx:
- violation = access_r;
+ violation = npfec.read_access;
break;
case p2m_access_rw:
- violation = access_x;
+ violation = npfec.insn_fetch;
break;
case p2m_access_rwx:
+ violation = 0;
break;
}
if ( violation )
{
- if ( p2m_mem_access_check(gpa, gla_valid, gla, access_r,
- access_w, access_x, &req_ptr) )
+ if ( p2m_mem_access_check(gpa, gla, npfec, &req_ptr) )
{
fall_through = 1;
} else {
@@ -1594,7 +1595,7 @@ int hvm_hap_nested_page_fault(paddr_t gp
* to the mmio handler.
*/
if ( (p2mt == p2m_mmio_dm) ||
- (access_w && (p2mt == p2m_ram_ro)) )
+ (npfec.write_access && (p2mt == p2m_ram_ro)) )
{
put_gfn(p2m->domain, gfn);
@@ -1613,7 +1614,7 @@ int hvm_hap_nested_page_fault(paddr_t gp
paged = 1;
/* Mem sharing: unshare the page and try again */
- if ( access_w && (p2mt == p2m_ram_shared) )
+ if ( npfec.write_access && (p2mt == p2m_ram_shared) )
{
ASSERT(!p2m_is_nestedp2m(p2m));
sharing_enomem =
@@ -1630,7 +1631,7 @@ int hvm_hap_nested_page_fault(paddr_t gp
* a large page, we do not change other pages type within that large
* page.
*/
- if ( access_w )
+ if ( npfec.write_access )
{
paging_mark_dirty(v->domain, mfn_x(mfn));
p2m_change_type(v->domain, gfn, p2m_ram_logdirty, p2m_ram_rw);
@@ -1640,7 +1641,7 @@ int hvm_hap_nested_page_fault(paddr_t gp
}
/* Shouldn't happen: Maybe the guest was writing to a r/o grant mapping? */
- if ( access_w && (p2mt == p2m_grant_map_ro) )
+ if ( npfec.write_access && (p2mt == p2m_grant_map_ro) )
{
gdprintk(XENLOG_WARNING,
"trying to write to read-only grant mapping\n");
--- a/xen/arch/x86/hvm/svm/svm.c
+++ b/xen/arch/x86/hvm/svm/svm.c
@@ -1289,7 +1289,7 @@ const struct hvm_function_table * __init
}
static void svm_do_nested_pgfault(struct vcpu *v,
- struct cpu_user_regs *regs, uint32_t npfec, paddr_t gpa)
+ struct cpu_user_regs *regs, uint32_t pfec, paddr_t gpa)
{
int ret;
unsigned long gfn = gpa >> PAGE_SHIFT;
@@ -1298,10 +1298,13 @@ static void svm_do_nested_pgfault(struct
p2m_access_t p2ma;
struct p2m_domain *p2m = NULL;
- ret = hvm_hap_nested_page_fault(gpa, 0, ~0ul,
- 1, /* All NPFs count as reads */
- npfec & PFEC_write_access,
- npfec & PFEC_insn_fetch);
+ struct npfec npfec = {
+ .read_access = 1, /* All NPFs count as reads */
+ .write_access = !!(pfec & PFEC_write_access),
+ .insn_fetch = !!(pfec & PFEC_insn_fetch)
+ };
+
+ ret = hvm_hap_nested_page_fault(gpa, ~0ul, npfec);
if ( tb_init_done )
{
@@ -1329,7 +1332,7 @@ static void svm_do_nested_pgfault(struct
case -1:
ASSERT(nestedhvm_enabled(v->domain) && nestedhvm_vcpu_in_guestmode(v));
/* inject #VMEXIT(NPF) into guest. */
- nestedsvm_vmexit_defer(v, VMEXIT_NPF, npfec, gpa);
+ nestedsvm_vmexit_defer(v, VMEXIT_NPF, pfec, gpa);
return;
}
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -2278,6 +2278,11 @@ static void ept_handle_violation(unsigne
p2m_type_t p2mt;
int ret;
struct domain *d = current->domain;
+ struct npfec npfec = {
+ .read_access = !!(qualification & EPT_READ_VIOLATION),
+ .write_access = !!(qualification & EPT_WRITE_VIOLATION),
+ .insn_fetch = !!(qualification & EPT_EXEC_VIOLATION)
+ };
if ( tb_init_done )
{
@@ -2296,14 +2301,14 @@ static void ept_handle_violation(unsigne
}
if ( qualification & EPT_GLA_VALID )
+ {
__vmread(GUEST_LINEAR_ADDRESS, &gla);
+ npfec.gla_valid = 1;
+ }
else
gla = ~0ull;
- ret = hvm_hap_nested_page_fault(gpa,
- !!(qualification & EPT_GLA_VALID), gla,
- !!(qualification & EPT_READ_VIOLATION),
- !!(qualification & EPT_WRITE_VIOLATION),
- !!(qualification & EPT_EXEC_VIOLATION));
+
+ ret = hvm_hap_nested_page_fault(gpa, gla, npfec);
switch ( ret )
{
case 0: // Unhandled L1 EPT violation
--- a/xen/arch/x86/mm/p2m.c
+++ b/xen/arch/x86/mm/p2m.c
@@ -1261,9 +1261,9 @@ void p2m_mem_paging_resume(struct domain
}
}
-bool_t p2m_mem_access_check(paddr_t gpa, bool_t gla_valid, unsigned long gla,
- bool_t access_r, bool_t access_w, bool_t access_x,
- mem_event_request_t **req_ptr)
+bool_t p2m_mem_access_check(paddr_t gpa, unsigned long gla,
+ struct npfec npfec,
+ mem_event_request_t **req_ptr)
{
struct vcpu *v = current;
unsigned long gfn = gpa >> PAGE_SHIFT;
@@ -1281,7 +1281,7 @@ bool_t p2m_mem_access_check(paddr_t gpa,
gfn_lock(p2m, gfn, 0);
mfn = p2m->get_entry(p2m, gfn, &p2mt, &p2ma, 0, NULL);
- if ( access_w && p2ma == p2m_access_rx2rw )
+ if ( npfec.write_access && p2ma == p2m_access_rx2rw )
{
rc = p2m->set_entry(p2m, gfn, mfn, PAGE_ORDER_4K, p2mt, p2m_access_rw);
ASSERT(rc);
@@ -1290,7 +1290,7 @@ bool_t p2m_mem_access_check(paddr_t gpa,
}
else if ( p2ma == p2m_access_n2rwx )
{
- ASSERT(access_w || access_r || access_x);
+ ASSERT(npfec.write_access || npfec.read_access || npfec.insn_fetch);
rc = p2m->set_entry(p2m, gfn, mfn, PAGE_ORDER_4K,
p2mt, p2m_access_rwx);
ASSERT(rc);
@@ -1341,11 +1341,11 @@ bool_t p2m_mem_access_check(paddr_t gpa,
/* Send request to mem event */
req->gfn = gfn;
req->offset = gpa & ((1 << PAGE_SHIFT) - 1);
- req->gla_valid = gla_valid;
+ req->gla_valid = npfec.gla_valid;
req->gla = gla;
- req->access_r = access_r;
- req->access_w = access_w;
- req->access_x = access_x;
+ req->access_r = npfec.read_access;
+ req->access_w = npfec.write_access;
+ req->access_x = npfec.insn_fetch;
req->vcpu_id = v->vcpu_id;
}
--- a/xen/include/asm-x86/hvm/hvm.h
+++ b/xen/include/asm-x86/hvm/hvm.h
@@ -435,11 +435,8 @@ static inline void hvm_invalidate_regs_f
#endif
}
-int hvm_hap_nested_page_fault(paddr_t gpa,
- bool_t gla_valid, unsigned long gla,
- bool_t access_r,
- bool_t access_w,
- bool_t access_x);
+int hvm_hap_nested_page_fault(paddr_t gpa, unsigned long gla,
+ struct npfec npfec);
#define hvm_msr_tsc_aux(v) ({ \
struct domain *__d = (v)->domain; \
--- a/xen/include/asm-x86/mm.h
+++ b/xen/include/asm-x86/mm.h
@@ -551,6 +551,16 @@ void audit_domains(void);
#endif
+/*
+ * Nested page fault exception codes.
+ */
+struct npfec {
+ unsigned int read_access:1;
+ unsigned int write_access:1;
+ unsigned int insn_fetch:1;
+ unsigned int gla_valid:1;
+};
+
int new_guest_cr3(unsigned long pfn);
void make_cr3(struct vcpu *v, unsigned long mfn);
void update_cr3(struct vcpu *v);
--- a/xen/include/asm-x86/p2m.h
+++ b/xen/include/asm-x86/p2m.h
@@ -568,9 +568,9 @@ void p2m_mem_paging_resume(struct domain
* been promoted with no underlying vcpu pause. If the req_ptr has been populated,
* then the caller must put the event in the ring (once having released get_gfn*
* locks -- caller must also xfree the request. */
-bool_t p2m_mem_access_check(paddr_t gpa, bool_t gla_valid, unsigned long gla,
- bool_t access_r, bool_t access_w, bool_t access_x,
- mem_event_request_t **req_ptr);
+bool_t p2m_mem_access_check(paddr_t gpa, unsigned long gla,
+ struct npfec npfec,
+ mem_event_request_t **req_ptr);
/* Resumes the running of the VCPU, restarting the last instruction */
void p2m_mem_access_resume(struct domain *d);

69
53ff36ae-x86-hvm-treat-non-insn-fetch-NPF-also-as-read-violations.patch Normal file
View File

@ -0,0 +1,69 @@
References: bnc#882089
# Commit 401d5c5cc5a780cad160aa0e3c282c11ac11dd0c
# Date 2014-08-28 16:03:26 +0200
# Author Tamas K Lengyel <tamas.lengyel@zentific.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/hvm: treat non-instruction fetch nested page faults also as read violations
As pointed out by Jan Beulich in
http://lists.xen.org/archives/html/xen-devel/2014-08/msg01269.html:
"Read-modify-write instructions absolutely need to be treated as read
accesses, yet hardware doesn't guarantee to tell us so (they may
surface as just write accesses)." This patch addresses the issue in
both the VMX and the SVM side.
VMX: Treat all write data access violations also as read violations (in
addition to those that were already reported as read violations).
SVM: Refine the meaning of read data access violations to distinguish
between read/write and instruction fetch access violations.
With this patch both VMX and SVM specific nested page fault handling code reports violations the same way, thus abstracting the hardware specific behaviour from the layers above.
Suggested-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: Tamas K Lengyel <tamas.lengyel@zentific.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Tim Deegan <tim@xen.org>
--- a/xen/arch/x86/hvm/svm/svm.c
+++ b/xen/arch/x86/hvm/svm/svm.c
@@ -1298,8 +1298,13 @@ static void svm_do_nested_pgfault(struct
p2m_access_t p2ma;
struct p2m_domain *p2m = NULL;
+ /*
+ * Since HW doesn't explicitly provide a read access bit and we need to
+ * somehow describe read-modify-write instructions we will conservatively
+ * set read_access for all memory accesses that are not instruction fetches.
+ */
struct npfec npfec = {
- .read_access = 1, /* All NPFs count as reads */
+ .read_access = !(pfec & PFEC_insn_fetch),
.write_access = !!(pfec & PFEC_write_access),
.insn_fetch = !!(pfec & PFEC_insn_fetch)
};
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -2278,8 +2278,21 @@ static void ept_handle_violation(unsigne
p2m_type_t p2mt;
int ret;
struct domain *d = current->domain;
+
+ /*
+ * We treat all write violations also as read violations.
+ * The reason why this is required is the following warning:
+ * "An EPT violation that occurs during as a result of execution of a
+ * read-modify-write operation sets bit 1 (data write). Whether it also
+ * sets bit 0 (data read) is implementation-specific and, for a given
+ * implementation, may differ for different kinds of read-modify-write
+ * operations."
+ * - Intel(R) 64 and IA-32 Architectures Software Developer's Manual
+ * Volume 3C: System Programming Guide, Part 3
+ */
struct npfec npfec = {
- .read_access = !!(qualification & EPT_READ_VIOLATION),
+ .read_access = !!(qualification & EPT_READ_VIOLATION) ||
+ !!(qualification & EPT_WRITE_VIOLATION),
.write_access = !!(qualification & EPT_WRITE_VIOLATION),
.insn_fetch = !!(qualification & EPT_EXEC_VIOLATION)
};

133
53ff36d5-x86-mem_event-deliver-gla-fault-EPT-violation-information.patch Normal file
View File

@ -0,0 +1,133 @@
References: bnc#882089
# Commit 692f3cc7dd05b80dbd027e46372b1c25d7975332
# Date 2014-08-28 16:04:05 +0200
# Author Tamas K Lengyel <tamas.lengyel@zentific.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/mem_event: deliver gla fault EPT violation information
On Intel EPT the exit qualification generated by a violation also
includes a bit (EPT_GLA_FAULT) which describes the following
information: Set if the access causing the EPT violation is to a
guest-physical address that is the translation of a linear address.
Clear if the access causing the EPT violation is to a paging-structure
entry as part of a page walk or the update of an accessed or dirty bit.
For more information see Table 27-7 in the Intel SDM.
This patch extends the mem_event system to deliver this extra
information, which could be useful for determining the cause of a
violation.
Signed-off-by: Tamas K Lengyel <tamas.lengyel@zentific.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Tim Deegan <tim@xen.org>
--- a/xen/arch/x86/hvm/svm/svm.c
+++ b/xen/arch/x86/hvm/svm/svm.c
@@ -1289,7 +1289,7 @@ const struct hvm_function_table * __init
}
static void svm_do_nested_pgfault(struct vcpu *v,
- struct cpu_user_regs *regs, uint32_t pfec, paddr_t gpa)
+ struct cpu_user_regs *regs, uint64_t pfec, paddr_t gpa)
{
int ret;
unsigned long gfn = gpa >> PAGE_SHIFT;
@@ -1309,6 +1309,12 @@ static void svm_do_nested_pgfault(struct
.insn_fetch = !!(pfec & PFEC_insn_fetch)
};
+ /* These bits are mutually exclusive */
+ if ( pfec & NPT_PFEC_with_gla )
+ npfec.kind = npfec_kind_with_gla;
+ else if ( pfec & NPT_PFEC_in_gpt )
+ npfec.kind = npfec_kind_in_gpt;
+
ret = hvm_hap_nested_page_fault(gpa, ~0ul, npfec);
if ( tb_init_done )
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -2317,6 +2317,10 @@ static void ept_handle_violation(unsigne
{
__vmread(GUEST_LINEAR_ADDRESS, &gla);
npfec.gla_valid = 1;
+ if( qualification & EPT_GLA_FAULT )
+ npfec.kind = npfec_kind_with_gla;
+ else
+ npfec.kind = npfec_kind_in_gpt;
}
else
gla = ~0ull;
--- a/xen/arch/x86/mm/p2m.c
+++ b/xen/arch/x86/mm/p2m.c
@@ -1343,10 +1343,13 @@ bool_t p2m_mem_access_check(paddr_t gpa,
req->offset = gpa & ((1 << PAGE_SHIFT) - 1);
req->gla_valid = npfec.gla_valid;
req->gla = gla;
+ if ( npfec.kind == npfec_kind_with_gla )
+ req->fault_with_gla = 1;
+ else if ( npfec.kind == npfec_kind_in_gpt )
+ req->fault_in_gpt = 1;
req->access_r = npfec.read_access;
req->access_w = npfec.write_access;
req->access_x = npfec.insn_fetch;
-
req->vcpu_id = v->vcpu_id;
}
--- a/xen/include/asm-x86/hvm/svm/svm.h
+++ b/xen/include/asm-x86/hvm/svm/svm.h
@@ -105,4 +105,10 @@ extern u32 svm_feature_flags;
extern void svm_host_osvw_reset(void);
extern void svm_host_osvw_init(void);
+/* EXITINFO1 fields on NPT faults */
+#define _NPT_PFEC_with_gla 32
+#define NPT_PFEC_with_gla (1UL<<_NPT_PFEC_with_gla)
+#define _NPT_PFEC_in_gpt 33
+#define NPT_PFEC_in_gpt (1UL<<_NPT_PFEC_in_gpt)
+
#endif /* __ASM_X86_HVM_SVM_H__ */
--- a/xen/include/asm-x86/mm.h
+++ b/xen/include/asm-x86/mm.h
@@ -552,6 +552,16 @@ void audit_domains(void);
#endif
/*
+ * Extra fault info types which are used to further describe
+ * the source of an access violation.
+ */
+typedef enum {
+ npfec_kind_unknown, /* must be first */
+ npfec_kind_in_gpt, /* violation in guest page table */
+ npfec_kind_with_gla /* violation with guest linear address */
+} npfec_kind_t;
+
+/*
* Nested page fault exception codes.
*/
struct npfec {
@@ -559,6 +569,7 @@ struct npfec {
unsigned int write_access:1;
unsigned int insn_fetch:1;
unsigned int gla_valid:1;
+ unsigned int kind:2; /* npfec_kind_t */
};
int new_guest_cr3(unsigned long pfn);
--- a/xen/include/public/mem_event.h
+++ b/xen/include/public/mem_event.h
@@ -62,7 +62,9 @@ typedef struct mem_event_st {
uint16_t access_w:1;
uint16_t access_x:1;
uint16_t gla_valid:1;
- uint16_t available:12;
+ uint16_t fault_with_gla:1;
+ uint16_t fault_in_gpt:1;
+ uint16_t available:10;
uint16_t reason;
} mem_event_request_t, mem_event_response_t;

49
53ff3716-x86-ats-Disable-Address-Translation-Services-by-default.patch Normal file
View File

@ -0,0 +1,49 @@
# Commit ad6eddb742577d182e634785bcfaf92732a50024
# Date 2014-08-28 16:05:10 +0200
# Author Andrew Cooper <andrew.cooper3@citrix.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/ats: Disable Address Translation Services by default
Xen cannot safely use any ATS functionality until it gains asynchronous queued
invalidation support, because of the current synchronous wait for completion.
Do not turn ATS on by default.
While editing the default in the command line documentation, correct the
statement regarding PCI Passthrough. ATS is purely a performance
optimisation, and is certainly not required for PCI Passthrough to function.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
--- a/docs/misc/xen-command-line.markdown
+++ b/docs/misc/xen-command-line.markdown
@@ -167,10 +167,13 @@ developers wishing Xen to fall back to o
### ats
> `= <boolean>`
-> Default: `true`
+> Default: `false`
+
+Permits Xen to set up and use PCI Address Translation Services. This is a
+performance optimisation for PCI Passthrough.
-Permits Xen to set up and use PCI Address Translation Services, which
-is required for PCI Passthrough.
+**WARNING: Xen cannot currently safely use ATS because of its synchronous wait
+loops for Queued Invalidation completions.**
### availmem
> `= <size>`
--- a/xen/drivers/passthrough/x86/ats.c
+++ b/xen/drivers/passthrough/x86/ats.c
@@ -20,7 +20,7 @@
LIST_HEAD(ats_devices);
-bool_t __read_mostly ats_enabled = 1;
+bool_t __read_mostly ats_enabled = 0;
boolean_param("ats", ats_enabled);
int enable_ats_device(int seg, int bus, int devfn, const void *iommu)

220
53ff3899-x86-NMI-allow-processing-unknown-NMIs-with-watchdog.patch Normal file
View File

@ -0,0 +1,220 @@
# Commit 3ea2ba980afe7356c613c8e1ba00d223d1c25412
# Date 2014-08-28 16:11:37 +0200
# Author Ross Lagerwall <ross.lagerwall@citrix.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/NMI: allow processing unknown NMIs when watchdog is enabled
Change NMI processing so that if watchdog=force is passed on the
command-line and the NMI is not caused by a perf counter overflow (i.e.
likely not a watchdog "tick"), the NMI is handled by the unknown NMI
handler.
This allows injection of NMIs from IPMI controllers that don't set the
IOCK/SERR bits to trigger the unknown NMI handler rather than be
ignored.
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Fix command line parsing (don't enable the watchdog on e.g.
"watchdog=xyz").
Signed-off-by: Jan Beulich <jbeulich@suse.com>
# Commit fd553ae5f0f57baa63d033bedee84f607de57d33
# Date 2014-09-03 15:09:59 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/NMI: allow passing just "watchdog" again
This capability got inadvertently lost in commit 3ea2ba980a ("x86/NMI:
allow processing unknown NMIs when watchdog is enabled") due to an
oversight of mine.
Reported-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
--- a/docs/misc/xen-command-line.markdown
+++ b/docs/misc/xen-command-line.markdown
@@ -1039,12 +1039,14 @@ As the BTS virtualisation is not 100% sa
don't use the vpmu flag on production systems with Intel cpus!
### watchdog
-> `= <boolean>`
+> `= force | <boolean>`
> Default: `false`
Run an NMI watchdog on each processor. If a processor is stuck for
-longer than the **watchdog\_timeout**, a panic occurs.
+longer than the **watchdog\_timeout**, a panic occurs. When `force` is
+specified, in addition to running an NMI watchdog on each processor,
+unknown NMIs will still be processed.
### watchdog\_timeout
> `= <integer>`
--- a/xen/arch/x86/nmi.c
+++ b/xen/arch/x86/nmi.c
@@ -43,7 +43,32 @@ static DEFINE_PER_CPU(unsigned int, nmi_
/* opt_watchdog: If true, run a watchdog NMI on each processor. */
bool_t __initdata opt_watchdog = 0;
-boolean_param("watchdog", opt_watchdog);
+
+/* watchdog_force: If true, process unknown NMIs when running the watchdog. */
+bool_t watchdog_force = 0;
+
+static void __init parse_watchdog(char *s)
+{
+ if ( !*s )
+ {
+ opt_watchdog = 1;
+ return;
+ }
+
+ switch ( parse_bool(s) )
+ {
+ case 0:
+ opt_watchdog = 0;
+ return;
+ case 1:
+ opt_watchdog = 1;
+ return;
+ }
+
+ if ( !strcmp(s, "force") )
+ watchdog_force = opt_watchdog = 1;
+}
+custom_param("watchdog", parse_watchdog);
/* opt_watchdog_timeout: Number of seconds to wait before panic. */
static unsigned int opt_watchdog_timeout = 5;
@@ -82,6 +107,7 @@ int nmi_active;
#define K7_EVNTSEL_USR (1 << 16)
#define K7_EVENT_CYCLES_PROCESSOR_IS_RUNNING 0x76
#define K7_NMI_EVENT K7_EVENT_CYCLES_PROCESSOR_IS_RUNNING
+#define K7_EVENT_WIDTH 32
#define P6_EVNTSEL0_ENABLE (1 << 22)
#define P6_EVNTSEL_INT (1 << 20)
@@ -89,10 +115,12 @@ int nmi_active;
#define P6_EVNTSEL_USR (1 << 16)
#define P6_EVENT_CPU_CLOCKS_NOT_HALTED 0x79
#define CORE_EVENT_CPU_CLOCKS_NOT_HALTED 0x3c
+#define P6_EVENT_WIDTH 32
#define P4_ESCR_EVENT_SELECT(N) ((N)<<25)
#define P4_CCCR_OVF_PMI0 (1<<26)
#define P4_CCCR_OVF_PMI1 (1<<27)
+#define P4_CCCR_OVF (1<<31)
#define P4_CCCR_THRESHOLD(N) ((N)<<20)
#define P4_CCCR_COMPLEMENT (1<<19)
#define P4_CCCR_COMPARE (1<<18)
@@ -433,8 +461,10 @@ int __init watchdog_setup(void)
return 0;
}
-void nmi_watchdog_tick(struct cpu_user_regs * regs)
+/* Returns false if this was not a watchdog NMI, true otherwise */
+bool_t nmi_watchdog_tick(struct cpu_user_regs *regs)
{
+ bool_t watchdog_tick = 1;
unsigned int sum = this_cpu(nmi_timer_ticks);
if ( (this_cpu(last_irq_sums) == sum) && watchdog_enabled() )
@@ -460,8 +490,15 @@ void nmi_watchdog_tick(struct cpu_user_r
if ( nmi_perfctr_msr )
{
+ uint64_t msr_content;
+
+ /* Work out if this is a watchdog tick by checking for overflow. */
if ( nmi_perfctr_msr == MSR_P4_IQ_PERFCTR0 )
{
+ rdmsrl(MSR_P4_IQ_CCCR0, msr_content);
+ if ( !(msr_content & P4_CCCR_OVF) )
+ watchdog_tick = 0;
+
/*
* P4 quirks:
* - An overflown perfctr will assert its interrupt
@@ -474,14 +511,26 @@ void nmi_watchdog_tick(struct cpu_user_r
}
else if ( nmi_perfctr_msr == MSR_P6_PERFCTR0 )
{
+ rdmsrl(MSR_P6_PERFCTR0, msr_content);
+ if ( msr_content & (1ULL << P6_EVENT_WIDTH) )
+ watchdog_tick = 0;
+
/*
* Only P6 based Pentium M need to re-unmask the apic vector but
* it doesn't hurt other P6 variants.
*/
apic_write(APIC_LVTPC, APIC_DM_NMI);
}
+ else if ( nmi_perfctr_msr == MSR_K7_PERFCTR0 )
+ {
+ rdmsrl(MSR_K7_PERFCTR0, msr_content);
+ if ( msr_content & (1ULL << K7_EVENT_WIDTH) )
+ watchdog_tick = 0;
+ }
write_watchdog_counter(NULL);
}
+
+ return watchdog_tick;
}
/*
--- a/xen/arch/x86/traps.c
+++ b/xen/arch/x86/traps.c
@@ -3226,14 +3226,15 @@ void do_nmi(struct cpu_user_regs *regs)
{
unsigned int cpu = smp_processor_id();
unsigned char reason;
+ bool_t handle_unknown = 0;
++nmi_count(cpu);
if ( nmi_callback(regs, cpu) )
return;
- if ( nmi_watchdog )
- nmi_watchdog_tick(regs);
+ if ( !nmi_watchdog || (!nmi_watchdog_tick(regs) && watchdog_force) )
+ handle_unknown = 1;
/* Only the BSP gets external NMIs from the system. */
if ( cpu == 0 )
@@ -3243,7 +3244,7 @@ void do_nmi(struct cpu_user_regs *regs)
pci_serr_error(regs);
if ( reason & 0x40 )
io_check_error(regs);
- if ( !(reason & 0xc0) && !nmi_watchdog )
+ if ( !(reason & 0xc0) && handle_unknown )
unknown_nmi_error(regs, reason);
}
}
--- a/xen/include/asm-x86/apic.h
+++ b/xen/include/asm-x86/apic.h
@@ -206,7 +206,7 @@ extern void release_lapic_nmi(void);
extern void self_nmi(void);
extern void disable_timer_nmi_watchdog(void);
extern void enable_timer_nmi_watchdog(void);
-extern void nmi_watchdog_tick (struct cpu_user_regs *regs);
+extern bool_t nmi_watchdog_tick (struct cpu_user_regs *regs);
extern int APIC_init_uniprocessor (void);
extern void disable_APIC_timer(void);
extern void enable_APIC_timer(void);
--- a/xen/include/asm-x86/nmi.h
+++ b/xen/include/asm-x86/nmi.h
@@ -8,6 +8,9 @@ struct cpu_user_regs;
/* Watchdog boolean from the command line */
extern bool_t opt_watchdog;
+
+/* Watchdog force parameter from the command line */
+extern bool_t watchdog_force;
typedef int (*nmi_callback_t)(struct cpu_user_regs *regs, int cpu);

163
54005472-EPT-utilize-GLA-GPA-translation-known-for-certain-faults.patch Normal file
View File

@ -0,0 +1,163 @@
References: bnc#882089
# Commit ecb69533582e51999e5d76bce513be870222908f
# Date 2014-08-29 12:22:42 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
EPT: utilize GLA->GPA translation known for certain faults
Rather than doing the translation ourselves in __hvmemul_{read,write}()
leverage that we know the association for faults other than such having
occurred when translating addresses of page tables.
There is one intentional but not necessarily obvious (and possibly
subtle) adjustment to behavior: __hvmemul_read() no longer blindly
bails on instruction fetches matching the MMIO GVA (the callers of
handle_mmio_with_translation() now control the behavior via the struct
npfec they pass, and it didn't seem right to bail here rather than just
falling through to the unaccelerated path)
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Tim Deegan <tim@xen.org>
--- a/xen/arch/x86/hvm/emulate.c
+++ b/xen/arch/x86/hvm/emulate.c
@@ -496,10 +496,11 @@ static int __hvmemul_read(
while ( off & (chunk - 1) )
chunk >>= 1;
- if ( unlikely(vio->mmio_gva == (addr & PAGE_MASK)) && vio->mmio_gva )
+ if ( ((access_type != hvm_access_insn_fetch
+ ? vio->mmio_access.read_access
+ : vio->mmio_access.insn_fetch)) &&
+ (vio->mmio_gva == (addr & PAGE_MASK)) )
{
- if ( access_type == hvm_access_insn_fetch )
- return X86EMUL_UNHANDLEABLE;
gpa = (((paddr_t)vio->mmio_gpfn << PAGE_SHIFT) | off);
while ( (off + chunk) <= PAGE_SIZE )
{
@@ -639,7 +640,8 @@ static int hvmemul_write(
while ( off & (chunk - 1) )
chunk >>= 1;
- if ( unlikely(vio->mmio_gva == (addr & PAGE_MASK)) && vio->mmio_gva )
+ if ( vio->mmio_access.write_access &&
+ (vio->mmio_gva == (addr & PAGE_MASK)) )
{
gpa = (((paddr_t)vio->mmio_gpfn << PAGE_SHIFT) | off);
while ( (off + chunk) <= PAGE_SIZE )
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -1529,7 +1529,7 @@ int hvm_hap_nested_page_fault(paddr_t gp
&& is_hvm_vcpu(v)
&& hvm_mmio_internal(gpa) )
{
- if ( !handle_mmio() )
+ if ( !handle_mmio_with_translation(gla, gpa >> PAGE_SHIFT, npfec) )
hvm_inject_hw_exception(TRAP_gp_fault, 0);
rc = 1;
goto out;
@@ -1603,7 +1603,7 @@ int hvm_hap_nested_page_fault(paddr_t gp
if ( unlikely(is_pvh_vcpu(v)) )
goto out;
- if ( !handle_mmio() )
+ if ( !handle_mmio_with_translation(gla, gpa >> PAGE_SHIFT, npfec) )
hvm_inject_hw_exception(TRAP_gp_fault, 0);
rc = 1;
goto out;
--- a/xen/arch/x86/hvm/io.c
+++ b/xen/arch/x86/hvm/io.c
@@ -189,7 +189,7 @@ int handle_mmio(void)
if ( vio->io_state == HVMIO_awaiting_completion )
vio->io_state = HVMIO_handle_mmio_awaiting_completion;
else
- vio->mmio_gva = 0;
+ vio->mmio_access = (struct npfec){};
switch ( rc )
{
@@ -218,9 +218,14 @@ int handle_mmio(void)
return 1;
}
-int handle_mmio_with_translation(unsigned long gva, unsigned long gpfn)
+int handle_mmio_with_translation(unsigned long gva, unsigned long gpfn,
+ struct npfec access)
{
struct hvm_vcpu_io *vio = &current->arch.hvm_vcpu.hvm_io;
+
+ vio->mmio_access = access.gla_valid &&
+ access.kind == npfec_kind_with_gla
+ ? access : (struct npfec){};
vio->mmio_gva = gva & PAGE_MASK;
vio->mmio_gpfn = gpfn;
return handle_mmio();
--- a/xen/arch/x86/mm/shadow/multi.c
+++ b/xen/arch/x86/mm/shadow/multi.c
@@ -2839,6 +2839,11 @@ static int sh_page_fault(struct vcpu *v,
p2m_type_t p2mt;
uint32_t rc;
int version;
+ struct npfec access = {
+ .read_access = 1,
+ .gla_valid = 1,
+ .kind = npfec_kind_with_gla
+ };
#if SHADOW_OPTIMIZATIONS & SHOPT_FAST_EMULATION
int fast_emul = 0;
#endif
@@ -2849,6 +2854,9 @@ static int sh_page_fault(struct vcpu *v,
perfc_incr(shadow_fault);
+ if ( regs->error_code & PFEC_write_access )
+ access.write_access = 1;
+
#if SHADOW_OPTIMIZATIONS & SHOPT_FAST_EMULATION
/* If faulting frame is successfully emulated in last shadow fault
* it's highly likely to reach same emulation action for this frame.
@@ -2950,7 +2958,7 @@ static int sh_page_fault(struct vcpu *v,
SHADOW_PRINTK("fast path mmio %#"PRIpaddr"\n", gpa);
reset_early_unshadow(v);
trace_shadow_gen(TRC_SHADOW_FAST_MMIO, va);
- return (handle_mmio_with_translation(va, gpa >> PAGE_SHIFT)
+ return (handle_mmio_with_translation(va, gpa >> PAGE_SHIFT, access)
? EXCRET_fault_fixed : 0);
}
else
@@ -3447,7 +3455,7 @@ static int sh_page_fault(struct vcpu *v,
paging_unlock(d);
put_gfn(d, gfn_x(gfn));
trace_shadow_gen(TRC_SHADOW_MMIO, va);
- return (handle_mmio_with_translation(va, gpa >> PAGE_SHIFT)
+ return (handle_mmio_with_translation(va, gpa >> PAGE_SHIFT, access)
? EXCRET_fault_fixed : 0);
not_a_shadow_fault:
--- a/xen/include/asm-x86/hvm/io.h
+++ b/xen/include/asm-x86/hvm/io.h
@@ -119,7 +119,8 @@ static inline void register_buffered_io_
void send_timeoffset_req(unsigned long timeoff);
void send_invalidate_req(void);
int handle_mmio(void);
-int handle_mmio_with_translation(unsigned long gva, unsigned long gpfn);
+int handle_mmio_with_translation(unsigned long gva, unsigned long gpfn,
+ struct npfec);
int handle_pio(uint16_t port, unsigned int size, int dir);
void hvm_interrupt_post(struct vcpu *v, int vector, int type);
void hvm_io_assist(ioreq_t *p);
--- a/xen/include/asm-x86/hvm/vcpu.h
+++ b/xen/include/asm-x86/hvm/vcpu.h
@@ -54,8 +54,9 @@ struct hvm_vcpu_io {
* HVM emulation:
* Virtual address @mmio_gva maps to MMIO physical frame @mmio_gpfn.
* The latter is known to be an MMIO frame (not RAM).
- * This translation is only valid if @mmio_gva is non-zero.
+ * This translation is only valid for accesses as per @mmio_access.
*/
+ struct npfec mmio_access;
unsigned long mmio_gva;
unsigned long mmio_gpfn;

52
CVE-2013-4540-qemu.patch Normal file
View File

@ -0,0 +1,52 @@
References: bnc#864801
Subject: zaurus: fix buffer overrun on invalid state load
From: Michael S. Tsirkin mst@redhat.com Thu Apr 3 19:52:13 2014 +0300
Date: Mon May 5 22:15:02 2014 +0200:
Git: 52f91c3723932f8340fe36c8ec8b18a757c37b2b
CVE-2013-4540
Within scoop_gpio_handler_update, if prev_level has a high bit set, then
we get bit > 16 and that causes a buffer overrun.
Since prev_level comes from wire indirectly, this can
happen on invalid state load.
Similarly for gpio_level and gpio_dir.
To fix, limit to 16 bit.
Reported-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Index: xen-4.4.1-testing/tools/qemu-xen-dir-remote/hw/gpio/zaurus.c
===================================================================
--- xen-4.4.1-testing.orig/tools/qemu-xen-dir-remote/hw/gpio/zaurus.c
+++ xen-4.4.1-testing/tools/qemu-xen-dir-remote/hw/gpio/zaurus.c
@@ -203,6 +203,15 @@ static bool is_version_0 (void *opaque,
return version_id == 0;
}
+static bool vmstate_scoop_validate(void *opaque, int version_id)
+{
+ ScoopInfo *s = opaque;
+
+ return !(s->prev_level & 0xffff0000) &&
+ !(s->gpio_level & 0xffff0000) &&
+ !(s->gpio_dir & 0xffff0000);
+}
+
static const VMStateDescription vmstate_scoop_regs = {
.name = "scoop",
.version_id = 1,
@@ -215,6 +224,7 @@ static const VMStateDescription vmstate_
VMSTATE_UINT32(gpio_level, ScoopInfo),
VMSTATE_UINT32(gpio_dir, ScoopInfo),
VMSTATE_UINT32(prev_level, ScoopInfo),
+ VMSTATE_VALIDATE("irq levels are 16 bit", vmstate_scoop_validate),
VMSTATE_UINT16(mcr, ScoopInfo),
VMSTATE_UINT16(cdr, ScoopInfo),
VMSTATE_UINT16(ccr, ScoopInfo),

12
README.SUSE
View File

@ -550,16 +550,16 @@ the 'xl dmesg' command will produce more output to better analyze problems.
Grub2 Example: Grub2 Example:
Edit /etc/default/grub and add, Edit /etc/default/grub and add,
GRUB_CMDLINE_XEN_DEFAULT="loglvl=all loglvl_guest=all" GRUB_CMDLINE_XEN_DEFAULT="loglvl=all guest_loglvl=all"
and then run, and then run,
grub2-mkconfig -o /boot/grub2/grub.cfg grub2-mkconfig -o /boot/grub2/grub.cfg
Grub1 Example: Grub1 Example:
Edit /boot/grub/menu.lst and edit the line containing xen.gz Edit /boot/grub/menu.lst and edit the line containing xen.gz
kernel /boot/xen.gz loglvl=all loglvl_guest=all kernel /boot/xen.gz loglvl=all guest_loglvl=all
2) With the log levels specified above and the host rebooted, more useful 2) With the log levels specified above and the host rebooted, more useful
information about domain 0 and running VMs can be obtained using using the information about domain 0 and running VMs can be obtained using the
'xl dmesg' and 'xl debug-keys' commands. For example, from the command line 'xl dmesg' and 'xl debug-keys' commands. For example, from the command line
run: run:
xl debug-keys h xl debug-keys h
@ -581,7 +581,7 @@ following to prepare Xen to send its output over this serial line.
Grub2 Example: Grub2 Example:
Edit /etc/default/grub and add, Edit /etc/default/grub and add,
GRUB_CMDLINE_XEN_DEFAULT="loglvl=all loglvl_guest=all console=com1 com1=115200,8n1" GRUB_CMDLINE_XEN_DEFAULT="loglvl=all guest_loglvl=all console=com1 com1=115200,8n1"
Also append additional serial flags to the option below such that it appears as, Also append additional serial flags to the option below such that it appears as,
GRUB_CMDLINE_LINUX_DEFAULT="<pre-existing flags> console=ttyS0, 115200" GRUB_CMDLINE_LINUX_DEFAULT="<pre-existing flags> console=ttyS0, 115200"
where pre-existing flags are those options already present and then run, where pre-existing flags are those options already present and then run,
@ -600,7 +600,7 @@ use the debug-enabled hypervisor, and/or to prevent automatic rebooting.
Grub2 Example: Grub2 Example:
Edit /etc/default/grub and add, Edit /etc/default/grub and add,
GRUB_CMDLINE_XEN_DEFAULT="noreboot loglvl=all loglvl_guest=all" GRUB_CMDLINE_XEN_DEFAULT="noreboot loglvl=all guest_loglvl=all"
Edit /etc/grub.d/20_linux_xen file. Look for this line: Edit /etc/grub.d/20_linux_xen file. Look for this line:
while [ "x${xen_list}" != "x" ] ; do while [ "x${xen_list}" != "x" ] ; do
and add *before* the above line something like this: and add *before* the above line something like this:
@ -616,7 +616,7 @@ Grub1 Example:
Edit your menu.lst configuration from something like this: Edit your menu.lst configuration from something like this:
kernel (hd0,5)/xen.gz kernel (hd0,5)/xen.gz
To something like this: To something like this:
kernel (hd0,5)/xen-dbg.gz noreboot loglvl=all loglvl_guest=all kernel (hd0,5)/xen-dbg.gz noreboot loglvl=all guest_loglvl=all
All hypervisor options require a reboot to take effect. After rebooting, the All hypervisor options require a reboot to take effect. After rebooting, the
Xen hypervisor will write any error messages to the log file (viewable with Xen hypervisor will write any error messages to the log file (viewable with

8
x86-ioapic-ack-default.patch
View File

@ -1,10 +1,8 @@
Change default IO-APIC ack mode for single IO-APIC systems to old-style. Change default IO-APIC ack mode for single IO-APIC systems to old-style.
Index: xen-4.3.0-testing/xen/arch/x86/io_apic.c --- a/xen/arch/x86/io_apic.c
=================================================================== +++ b/xen/arch/x86/io_apic.c
--- xen-4.3.0-testing.orig/xen/arch/x86/io_apic.c @@ -2034,7 +2034,10 @@ void __init setup_IO_APIC(void)
+++ xen-4.3.0-testing/xen/arch/x86/io_apic.c
@@ -2026,7 +2026,10 @@ void __init setup_IO_APIC(void)
io_apic_irqs = ~PIC_IRQS; io_apic_irqs = ~PIC_IRQS;
printk("ENABLING IO-APIC IRQs\n"); printk("ENABLING IO-APIC IRQs\n");

4
xen-4.4.1-testing-src.tar.bz2
View File

@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1 version https://git-lfs.github.com/spec/v1
oid sha256:f277f2bd5714773978d9a3688fd1f65577178aaac1add3b7d66c01e3648267b1 oid sha256:57062330e02f3d4d2911bb184ed89717a007e8e276e5922a1e50d6e90e72662d
size 4432926 size 4368624

34
xen.changes
View File

@ -1,3 +1,37 @@
-------------------------------------------------------------------
Tue Sep 2 09:01:24 MDT 2014 - carnold@suse.com
- Update to Xen Version 4.4.1 FCS
xen-4.4.1-testing-src.tar.bz2
- Dropped patches now contained in tarball
53d7b781-x86-cpu-undo-BIOS-CPUID-max_leaf-limit-earlier.patch
53df71c7-lz4-check-for-underruns.patch
53e47d6b-x86_emulate-properly-do-IP-updates-and-other-side-effects.patch
-------------------------------------------------------------------
Mon Sep 1 15:20:20 MDT 2014 - carnold@suse.com
- bnc#882089 - Windows 2012 R2 fails to boot up with greater than
60 vcpus
53df727b-x86-HVM-extend-LAPIC-shortcuts-around-P2M-lookups.patch
53e8be5f-x86-vHPET-use-rwlock-instead-of-simple-one.patch
53ff3659-x86-consolidate-boolean-inputs-in-hvm-and-p2m.patch
53ff36ae-x86-hvm-treat-non-insn-fetch-NPF-also-as-read-violations.patch
53ff36d5-x86-mem_event-deliver-gla-fault-EPT-violation-information.patch
54005472-EPT-utilize-GLA-GPA-translation-known-for-certain-faults.patch
- Upstream patches from Jan
53f737b1-VMX-fix-DebugCtl-MSR-clearing.patch
53f7386d-x86-irq-process-softirqs-in-irq-keyhandlers.patch
53ff3716-x86-ats-Disable-Address-Translation-Services-by-default.patch
53ff3899-x86-NMI-allow-processing-unknown-NMIs-with-watchdog.patch
-------------------------------------------------------------------
Fri Aug 29 09:25:47 MDT 2014 - carnold@suse.com
- bnc#864801 - VUL-0: CVE-2013-4540: qemu: zaurus: buffer overrun
on invalid state load
CVE-2013-4540-qemu.patch
------------------------------------------------------------------- -------------------------------------------------------------------
Fri Aug 15 15:09:27 MDT 2014 - carnold@suse.com Fri Aug 15 15:09:27 MDT 2014 - carnold@suse.com

28
xen.spec
View File

@ -21,7 +21,7 @@ Name: xen
ExclusiveArch: %ix86 x86_64 %arm aarch64 ExclusiveArch: %ix86 x86_64 %arm aarch64
%define xvers 4.4 %define xvers 4.4
%define xvermaj 4 %define xvermaj 4
%define changeset 28531 %define changeset 28541
%define xen_build_dir xen-4.4.1-testing %define xen_build_dir xen-4.4.1-testing
# #
%define with_kmp 0 %define with_kmp 0
@ -153,7 +153,7 @@ BuildRequires: xorg-x11-util-devel
%endif %endif
%endif %endif
Version: 4.4.1_02 Version: 4.4.1_04
Release: 0 Release: 0
PreReq: %insserv_prereq %fillup_prereq PreReq: %insserv_prereq %fillup_prereq
Summary: Xen Virtualization: Hypervisor (aka VMM aka Microkernel) Summary: Xen Virtualization: Hypervisor (aka VMM aka Microkernel)
@ -224,11 +224,17 @@ Patch9: 539ebe62-x86-EFI-improve-boot-time-diagnostics.patch
Patch10: 53aac342-x86-HVM-consolidate-and-sanitize-CR4-guest-reserved-bit-determination.patch Patch10: 53aac342-x86-HVM-consolidate-and-sanitize-CR4-guest-reserved-bit-determination.patch
Patch11: 53c9151b-Fix-xl-vncviewer-accesses-port-0-by-any-invalid-domid.patch Patch11: 53c9151b-Fix-xl-vncviewer-accesses-port-0-by-any-invalid-domid.patch
Patch12: 53d124e7-fix-list_domain_details-check-config-data-length-0.patch Patch12: 53d124e7-fix-list_domain_details-check-config-data-length-0.patch
Patch13: 53d7b781-x86-cpu-undo-BIOS-CPUID-max_leaf-limit-earlier.patch Patch13: 53dba447-x86-ACPI-allow-CMOS-RTC-use-even-when-ACPI-says-there-is-none.patch
Patch14: 53dba447-x86-ACPI-allow-CMOS-RTC-use-even-when-ACPI-says-there-is-none.patch Patch14: 53df727b-x86-HVM-extend-LAPIC-shortcuts-around-P2M-lookups.patch
Patch15: 53df71c7-lz4-check-for-underruns.patch Patch15: 53e8be5f-x86-vHPET-use-rwlock-instead-of-simple-one.patch
Patch16: 53df727b-x86-HVM-extend-LAPIC-shortcuts-around-P2M-lookups.patch Patch16: 53f737b1-VMX-fix-DebugCtl-MSR-clearing.patch
Patch17: 53e47d6b-x86_emulate-properly-do-IP-updates-and-other-side-effects.patch Patch17: 53f7386d-x86-irq-process-softirqs-in-irq-keyhandlers.patch
Patch18: 53ff3659-x86-consolidate-boolean-inputs-in-hvm-and-p2m.patch
Patch19: 53ff36ae-x86-hvm-treat-non-insn-fetch-NPF-also-as-read-violations.patch
Patch20: 53ff36d5-x86-mem_event-deliver-gla-fault-EPT-violation-information.patch
Patch21: 53ff3716-x86-ats-Disable-Address-Translation-Services-by-default.patch
Patch22: 53ff3899-x86-NMI-allow-processing-unknown-NMIs-with-watchdog.patch
Patch23: 54005472-EPT-utilize-GLA-GPA-translation-known-for-certain-faults.patch
# Upstream qemu # Upstream qemu
Patch250: VNC-Support-for-ExtendedKeyEvent-client-message.patch Patch250: VNC-Support-for-ExtendedKeyEvent-client-message.patch
Patch251: 0001-net-move-the-tap-buffer-into-TAPState.patch Patch251: 0001-net-move-the-tap-buffer-into-TAPState.patch
@ -239,6 +245,7 @@ Patch255: 0005-e1000-multi-buffer-packet-support.patch
Patch256: 0006-e1000-clear-EOP-for-multi-buffer-descriptors.patch Patch256: 0006-e1000-clear-EOP-for-multi-buffer-descriptors.patch
Patch257: 0007-e1000-verify-we-have-buffers-upfront.patch Patch257: 0007-e1000-verify-we-have-buffers-upfront.patch
Patch258: 0008-e1000-check-buffer-availability.patch Patch258: 0008-e1000-check-buffer-availability.patch
Patch259: CVE-2013-4540-qemu.patch
# Our platform specific patches # Our platform specific patches
Patch301: xen-destdir.patch Patch301: xen-destdir.patch
Patch302: xen-xmexample.patch Patch302: xen-xmexample.patch
@ -611,6 +618,12 @@ Authors:
%patch15 -p1 %patch15 -p1
%patch16 -p1 %patch16 -p1
%patch17 -p1 %patch17 -p1
%patch18 -p1
%patch19 -p1
%patch20 -p1
%patch21 -p1
%patch22 -p1
%patch23 -p1
# Upstream qemu patches # Upstream qemu patches
%patch250 -p1 %patch250 -p1
%patch251 -p1 %patch251 -p1
@ -621,6 +634,7 @@ Authors:
%patch256 -p1 %patch256 -p1
%patch257 -p1 %patch257 -p1
%patch258 -p1 %patch258 -p1
%patch259 -p1
# Our platform specific patches # Our platform specific patches
%patch301 -p1 %patch301 -p1
%patch302 -p1 %patch302 -p1