Re: [PATCH] Test case of multibyte NOP in emulation mode
On Wed, Jun 05, 2013 at 10:16:46AM +0800, 李春奇 Arthur Chunqi Li wrote: Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Adding the test to x86/realmode.c will be much easier. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- Gleb. -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Test case of multibyte NOP in emulation mode
Yes, that should be the point. x86/realmode.c is always running in emulation mode. I added the testing here there but no error occurred. I cannot find the reason. The code is as follows added to x86/realmode.c static void test_nopl(void) { MK_INSN(nopl, .byte 0x0f, 0x1f, 0x00\n\r); exec_in_big_real_mode(insn_nopl); report(nopl, 0, 1); } and I objdump from realmode.flat is as follows: 6458 insn_code_nopl: 6458: 0f 1f 00nopl (%eax) But there cause no error when executing this insn. Why? On Wed, Jun 5, 2013 at 2:42 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 10:16:46AM +0800, 李春奇 Arthur Chunqi Li wrote: Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Adding the test to x86/realmode.c will be much easier. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Test case of multibyte NOP in emulation mode
On Wed, Jun 05, 2013 at 03:00:33PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, that should be the point. x86/realmode.c is always running in emulation mode. I added the testing here there but no error occurred. I cannot find the reason. The code is as follows added to x86/realmode.c static void test_nopl(void) { MK_INSN(nopl, .byte 0x0f, 0x1f, 0x00\n\r); exec_in_big_real_mode(insn_nopl); report(nopl, 0, 1); } and I objdump from realmode.flat is as follows: 6458 insn_code_nopl: 6458: 0f 1f 00nopl (%eax) But there cause no error when executing this insn. Why? Because you probably use cpu that supports unrestricted mode or use AMD processor. Can you try loading kvm-intel with unrestricted_guest=0 option? On Wed, Jun 5, 2013 at 2:42 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 10:16:46AM +0800, 李春奇 Arthur Chunqi Li wrote: Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Adding the test to x86/realmode.c will be much easier. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Test case of multibyte NOP in emulation mode
Yes, I load kvm-intel with unrestricted_guest=0 and the emulator runs well. I will give another test case in x86/realmode.c later. BTW, what is the action when a 64-bit instruction executes in x86/realmode.c? Should I add 64-bit insn tests only in x86/emulator.c? On Wed, Jun 5, 2013 at 4:27 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 03:00:33PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, that should be the point. x86/realmode.c is always running in emulation mode. I added the testing here there but no error occurred. I cannot find the reason. The code is as follows added to x86/realmode.c static void test_nopl(void) { MK_INSN(nopl, .byte 0x0f, 0x1f, 0x00\n\r); exec_in_big_real_mode(insn_nopl); report(nopl, 0, 1); } and I objdump from realmode.flat is as follows: 6458 insn_code_nopl: 6458: 0f 1f 00nopl (%eax) But there cause no error when executing this insn. Why? Because you probably use cpu that supports unrestricted mode or use AMD processor. Can you try loading kvm-intel with unrestricted_guest=0 option? On Wed, Jun 5, 2013 at 2:42 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 10:16:46AM +0800, 李春奇 Arthur Chunqi Li wrote: Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Adding the test to x86/realmode.c will be much easier. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Test case of multibyte NOP in emulation mode
On Wed, Jun 05, 2013 at 05:23:18PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, I load kvm-intel with unrestricted_guest=0 and the emulator runs well. I will give another test case in x86/realmode.c later. The test fails for me on CPU without unrestricted guest support. This means you either test on fixed kernel or unrestricted_guest=0 is broken. BTW, what is the action when a 64-bit instruction executes in x86/realmode.c? Should I add 64-bit insn tests only in x86/emulator.c? Yes, 64-bit or 32-bit instructions should be added to x86/emulator.c. On Wed, Jun 5, 2013 at 4:27 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 03:00:33PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, that should be the point. x86/realmode.c is always running in emulation mode. I added the testing here there but no error occurred. I cannot find the reason. The code is as follows added to x86/realmode.c static void test_nopl(void) { MK_INSN(nopl, .byte 0x0f, 0x1f, 0x00\n\r); exec_in_big_real_mode(insn_nopl); report(nopl, 0, 1); } and I objdump from realmode.flat is as follows: 6458 insn_code_nopl: 6458: 0f 1f 00nopl (%eax) But there cause no error when executing this insn. Why? Because you probably use cpu that supports unrestricted mode or use AMD processor. Can you try loading kvm-intel with unrestricted_guest=0 option? On Wed, Jun 5, 2013 at 2:42 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 10:16:46AM +0800, 李春奇 Arthur Chunqi Li wrote: Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Adding the test to x86/realmode.c will be much easier. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Test case of multibyte NOP in emulation mode
I mean after adding unrestricted_guest=0, the error is reproduced. Sorry for confused expression. I have committed another patch in x86/realmode.c. On Wed, Jun 5, 2013 at 5:28 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 05:23:18PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, I load kvm-intel with unrestricted_guest=0 and the emulator runs well. I will give another test case in x86/realmode.c later. The test fails for me on CPU without unrestricted guest support. This means you either test on fixed kernel or unrestricted_guest=0 is broken. BTW, what is the action when a 64-bit instruction executes in x86/realmode.c? Should I add 64-bit insn tests only in x86/emulator.c? Yes, 64-bit or 32-bit instructions should be added to x86/emulator.c. On Wed, Jun 5, 2013 at 4:27 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 03:00:33PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, that should be the point. x86/realmode.c is always running in emulation mode. I added the testing here there but no error occurred. I cannot find the reason. The code is as follows added to x86/realmode.c static void test_nopl(void) { MK_INSN(nopl, .byte 0x0f, 0x1f, 0x00\n\r); exec_in_big_real_mode(insn_nopl); report(nopl, 0, 1); } and I objdump from realmode.flat is as follows: 6458 insn_code_nopl: 6458: 0f 1f 00nopl (%eax) But there cause no error when executing this insn. Why? Because you probably use cpu that supports unrestricted mode or use AMD processor. Can you try loading kvm-intel with unrestricted_guest=0 option? On Wed, Jun 5, 2013 at 2:42 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 10:16:46AM +0800, 李春奇 Arthur Chunqi Li wrote: Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Adding the test to x86/realmode.c will be much easier. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Test case of multibyte NOP in emulation mode
On Wed, Jun 05, 2013 at 05:46:31PM +0800, 李春奇 Arthur Chunqi Li wrote: I mean after adding unrestricted_guest=0, the error is reproduced. Ah, OK. unrestricted_guest=0 works then :) Sorry for confused expression. I have committed another patch in x86/realmode.c. On Wed, Jun 5, 2013 at 5:28 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 05:23:18PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, I load kvm-intel with unrestricted_guest=0 and the emulator runs well. I will give another test case in x86/realmode.c later. The test fails for me on CPU without unrestricted guest support. This means you either test on fixed kernel or unrestricted_guest=0 is broken. BTW, what is the action when a 64-bit instruction executes in x86/realmode.c? Should I add 64-bit insn tests only in x86/emulator.c? Yes, 64-bit or 32-bit instructions should be added to x86/emulator.c. On Wed, Jun 5, 2013 at 4:27 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 03:00:33PM +0800, 李春奇 Arthur Chunqi Li wrote: Yes, that should be the point. x86/realmode.c is always running in emulation mode. I added the testing here there but no error occurred. I cannot find the reason. The code is as follows added to x86/realmode.c static void test_nopl(void) { MK_INSN(nopl, .byte 0x0f, 0x1f, 0x00\n\r); exec_in_big_real_mode(insn_nopl); report(nopl, 0, 1); } and I objdump from realmode.flat is as follows: 6458 insn_code_nopl: 6458: 0f 1f 00nopl (%eax) But there cause no error when executing this insn. Why? Because you probably use cpu that supports unrestricted mode or use AMD processor. Can you try loading kvm-intel with unrestricted_guest=0 option? On Wed, Jun 5, 2013 at 2:42 PM, Gleb Natapov g...@redhat.com wrote: On Wed, Jun 05, 2013 at 10:16:46AM +0800, 李春奇 Arthur Chunqi Li wrote: Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Adding the test to x86/realmode.c will be much easier. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- Arthur Chunqi Li Department of Computer Science School of EECS Peking University Beijing, China -- Gleb. -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at
[PATCH] Test case of multibyte NOP in emulation mode
Add multibyte NOP test case to kvm-unit-tests. This case can test one of bugs when booting RHEL5.9 64-bit. Signed-off-by: Arthur Chunqi Li yzt...@gmail.com --- x86/emulator.c | 33 + 1 file changed, 33 insertions(+) diff --git a/x86/emulator.c b/x86/emulator.c index 96576e5..f26c70f 100644 --- a/x86/emulator.c +++ b/x86/emulator.c @@ -901,6 +901,37 @@ static void test_simplealu(u32 *mem) report(test, *mem == 0x8400); } +static void test_nopl(uint64_t *mem, uint8_t *insn_page, + uint8_t *alt_insn_page, void *insn_ram) +{ +ulong *cr3 = (ulong *)read_cr3(); + +// Pad with RET instructions +memset(insn_page, 0xc3, 4096); +memset(alt_insn_page, 0xc3, 4096); +// Place a trapping instruction in the page to trigger a VMEXIT +insn_page[0] = 0x89; // mov %eax, (%rax) +insn_page[1] = 0x00; +insn_page[2] = 0x90; // nop +// Place nopl 0x0(%eax) in alt_insn_page for emulator to execuate +alt_insn_page[0] = 0x0f; // nop DWORD ptr[EAX] +alt_insn_page[1] = 0x1f; +alt_insn_page[2] = 0x00; + +// Load the code TLB with insn_page, but point the page tables at +// alt_insn_page (and keep the data TLB clear, for AMD decode assist). +// This will make the CPU trap on the insn_page instruction but the +// hypervisor will see alt_insn_page. +install_page(cr3, virt_to_phys(insn_page), insn_ram); +// Load code TLB +invlpg(insn_ram); +asm volatile(call *%0 : : r(insn_ram + 3)); +// Trap, let hypervisor emulate at alt_insn_page +install_page(cr3, virt_to_phys(alt_insn_page), insn_ram); +asm volatile(call *%0 : : r(insn_ram), a(mem)); +report(nopl, 1); +} + int main() { void *mem; @@ -964,6 +995,8 @@ int main() test_string_io_mmio(mem); + test_nopl(mem, insn_page, alt_insn_page, insn_ram); + printf(\nSUMMARY: %d tests, %d failures\n, tests, fails); return fails ? 1 : 0; } -- 1.7.9.5 -- To unsubscribe from this list: send the line unsubscribe kvm in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html