Skip to contentSkip to conversation

i386: Mask SVM features if nested SVM is disabled

PATCH In review Eduardo Habkost 3 participants 3 messages Jun 23, 2020 – Jun 24, 2020

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 i386: Mask SVM features if nested SVM is disabled

QEMU incorrectly validates FEAT_SVM feature flags against GET_SUPPORTED_CPUID even if SVM features are being masked out by cpu_x86_cpuid(). This can make QEMU print warnings on most AMD CPU models, even when SVM nesting is disabled (which is the default).

This bug was never detected before because of a Linux KVM bug:
until Linux v5.6, KVM was not filtering out SVM features in GET_SUPPORTED_CPUID when nested was disabled. This KVM bug was fixed in Linux v5.7-rc1, on Linux commit a50718cc3f43 ("KVM:
nSVM: Expose SVM features to L1 iff nested is enabled").

Fix the problem by adding a CPUID_EXT3_SVM dependency to all FEAT_SVM feature flags in the feature_dependencies table.

---
 target/i386/cpu.c | 4 ++++
 1 file changed, 4 insertions(+)
target/i386/cpu.c +4 -0
index b1b311baa2..a9edcaf531 100644
@@ -1404,6 +1404,10 @@ static FeatureDep feature_dependencies[] = {
         .from = { FEAT_VMX_SECONDARY_CTLS,  VMX_SECONDARY_EXEC_ENABLE_VMFUNC },
         .to = { FEAT_VMX_VMFUNC,            ~0ull },
     },
+    {
+        .from = { FEAT_8000_0001_ECX,       CPUID_EXT3_SVM },
+        .to = { FEAT_SVM,                   ~0ull },
+    },
 };
 
 typedef struct X86RegisterInfo32 {
Signature · 1 line
2.26.2
#1 lore raw
Patchew URL: https://patchew.org/QEMU/[email protected]/

Patchew URL: https://patchew.org/QEMU/[email protected]/

Hi,

This series failed the docker-quick@centos7 build test. Please find the testing commands and their output below. If you have Docker installed, you can probably reproduce it locally.

=== TEST SCRIPT BEGIN ===
#!/bin/bash
make docker-image-centos7 V=1 NETWORK=1
time make docker-test-quick@centos7 SHOW_ENV=1 J=14 NETWORK=1 === TEST SCRIPT END ===

  TEST    iotest-qcow2: 184
  TEST    check-qtest-x86_64: tests/qtest/test-x86-cpuid-compat
  TEST    iotest-qcow2: 186

qemu-system-x86_64: warning: This feature depends on other features that were not requested: CPUID.8000000AH:EDX.npt [bit 0]
**
ERROR:/tmp/qemu-test/src/tests/qtest/test-x86-cpuid-compat.c:64:test_cpuid_prop: assertion failed (val == args->expected_value): (0 == 2147483658)
ERROR - Bail out! ERROR:/tmp/qemu-test/src/tests/qtest/test-x86-cpuid-compat.c:64:test_cpuid_prop: assertion failed (val == args->expected_value): (0 == 2147483658)
make: *** [check-qtest-x86_64] Error 1
make: *** Waiting for unfinished jobs....
TEST iotest-qcow2: 187
TEST iotest-qcow2: 190
---
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['sudo', '-n', 'docker', 'run', '--label', 'com.qemu.instance.uuid=2a55c44dc54d40a4938f8487303f6dba', '-u', '1003', '--security-opt', 'seccomp=unconfined', '--rm', '-e', 'TARGET_LIST=', '-e', 'EXTRA_CONFIGURE_OPTS=', '-e', 'V=', '-e', 'J=14', '-e', 'DEBUG=', '-e', 'SHOW_ENV=1', '-e', 'CCACHE_DIR=/var/tmp/ccache', '-v', '/home/patchew2/.cache/qemu-docker-ccache:/var/tmp/ccache:z', '-v', '/var/tmp/patchew-tester-tmp-naaagll6/src/docker-src.2020-06-23-19.07.43.6102:/var/tmp/qemu:z,ro', 'qemu:centos7', '/var/tmp/qemu/run', 'test-quick']' returned non-zero exit status 2. filter=--filter=label=com.qemu.instance.uuid=2a55c44dc54d40a4938f8487303f6dba
make[1]: *** [docker-run] Error 1
make[1]: Leaving directory `/var/tmp/patchew-tester-tmp-naaagll6/src'
make: *** [docker-run-test-quick@centos7] Error 2

real 15m25.121s
user 0m8.515s

The full log is available at
http://patchew.org/logs/[email protected]/testing.docker-quick@centos7/?type=message.
---
Email generated automatically by Patchew [https://patchew.org/].
Please send your feedback to [email protected]

#2 lore raw
In reply to Eduardo Habkost
Quoted 27 lines · Eduardo Habkost

QEMU incorrectly validates FEAT_SVM feature flags against GET_SUPPORTED_CPUID even if SVM features are being masked out by cpu_x86_cpuid(). This can make QEMU print warnings on most AMD CPU models, even when SVM nesting is disabled (which is the default).

This bug was never detected before because of a Linux KVM bug:
until Linux v5.6, KVM was not filtering out SVM features in GET_SUPPORTED_CPUID when nested was disabled. This KVM bug was fixed in Linux v5.7-rc1, on Linux commit a50718cc3f43 ("KVM:
nSVM: Expose SVM features to L1 iff nested is enabled").

Fix the problem by adding a CPUID_EXT3_SVM dependency to all FEAT_SVM feature flags in the feature_dependencies table.

---
 target/i386/cpu.c | 4 ++++
 1 file changed, 4 insertions(+)
target/i386/cpu.c +0 -0
index b1b311baa2..a9edcaf531 100644

target/i386/cpu.c static FeatureDep feature_dependencies[] = { · line 1413

Quoted 10 lines
target/i386/cpu.c +4 -0
@@ -1404,6 +1404,10 @@ static FeatureDep feature_dependencies[] = {
         .from = { FEAT_VMX_SECONDARY_CTLS,  VMX_SECONDARY_EXEC_ENABLE_VMFUNC },
         .to = { FEAT_VMX_VMFUNC,            ~0ull },
     },
+    {
+        .from = { FEAT_8000_0001_ECX,       CPUID_EXT3_SVM },
+        .to = { FEAT_SVM,                   ~0ull },
+    },
 };
 
 typedef struct X86RegisterInfo32 {

Queued with this fixup:

tests/qtest/test-x86-cpuid-compat.c +2 -2
index 772287bdb4..8709e7d9ce 100644
@@ -256,7 +256,7 @@ int main(int argc, char **argv)
                    "-cpu 486,+invtsc", "xlevel", 0x80000007);
     /* CPUID[8000_000A].EDX: */
     add_cpuid_test("x86/cpuid/auto-xlevel/486/npt",
-                   "-cpu 486,+npt", "xlevel", 0x8000000A);
+                   "-cpu 486,+svm,+npt", "xlevel", 0x8000000A);
     /* CPUID[C000_0001].EDX: */
     add_cpuid_test("x86/cpuid/auto-xlevel2/phenom/xstore",
                    "-cpu phenom,+xstore", "xlevel2", 0xC0000001);
@@ -348,7 +348,7 @@ int main(int argc, char **argv)
                    "-machine pc-i440fx-2.4 -cpu SandyBridge,",
                    "xlevel", 0x80000008);
     add_cpuid_test("x86/cpuid/xlevel-compat/pc-i440fx-2.4/npt-on",
-                   "-machine pc-i440fx-2.4 -cpu SandyBridge,+npt",
+                   "-machine pc-i440fx-2.4 -cpu SandyBridge,+svm,+npt",
                    "xlevel", 0x80000008);
 
     /* Test feature parsing */

Paolo