selftests/mm: Skip test for non-LPA2 and non-LVA systems
Subsystems (3)
PATCH 1/1 selftests/mm: Skip test for non-LPA2 and non-LVA systems
Post my improvement of the test:
https://lore.kernel.org/all/[email protected]/
The test begins to fail on 4k and 16k pages, on non-LPA2 systems. To reduce noise in the CI systems, let us skip the test when higher address space is not implemented.
---
The patch applies on linux-next.
tools/testing/selftests/mm/va_high_addr_switch.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)
tools/testing/selftests/mm/va_high_addr_switch.c +13 -1
@@ -293,6 +293,18 @@ static int run_test(struct testcase *test, int count)return ret; } +/* Check if userspace VA > 48 bits */ +static int high_address_present (void ) +{ +void *ptr =mmap ((void *)(1UL <<50 ),1 ,PROT_READ |PROT_WRITE , +MAP_PRIVATE |MAP_ANONYMOUS |MAP_FIXED ,-1 ,0 ); +if (ptr ==MAP_FAILED ) +return 0 ; + +munmap (ptr,1 ); +return 1 ; +} +static int supported_arch (void ) {#if defined(__powerpc64__)
@@ -300,7 +312,7 @@ static int supported_arch(void)#elif defined(__x86_64__)return 1 ;#elif defined(__aarch64__) -return 1 ; +return high_address_present ();#else return 0 ;#endif
Signature · 1 line
2.34.1
Quoted 18 lines · Dev Jain
Post my improvement of the test:
https://lore.kernel.org/all/[email protected]/
The test begins to fail on 4k and 16k pages, on non-LPA2 systems. To reduce noise in the CI systems, let us skip the test when higher address space is not implemented.Signed-off-by Dev Jain <[email protected]>---
The patch applies on linux-next.tools/testing/selftests/mm/va_high_addr_switch.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)tools/testing/selftests/mm/va_high_addr_switch.c +0
-0
tools/testing/selftests/mm/va_high_addr_switch.c static int run_test(struct testcase *test, int count) · line 300
Quoted 8 lines
tools/testing/selftests/mm/va_high_addr_switch.c +5
-0@@ -293,6 +293,18 @@ static int run_test(struct testcase *test, int count)return ret; } +/* Check if userspace VA > 48 bits */ +static int high_address_present (void ) +{ +void *ptr =mmap ((void *)(1UL <<50 ),1 ,PROT_READ |PROT_WRITE , +MAP_PRIVATE |MAP_ANONYMOUS |MAP_FIXED ,-1 ,0 );
I think there is (very unlikely) possibility that something is already mapped at this address so it will be replaced due to MAP_FIXED. That could break the test. But the only way something could be already mapped is if ARM64_FORCE_52BIT is set and in that case, the test will fail anyway, right? So I think this is fine.
tools/testing/selftests/mm/va_high_addr_switch.c static int run_test(struct testcase *test, int count) · line 306
Quoted 6 lines
tools/testing/selftests/mm/va_high_addr_switch.c +6
-0+if (ptr ==MAP_FAILED ) +return 0 ; + +munmap (ptr,1 ); +return 1 ; +}
I'm guessing this will cause a function-not-used warning on arches other than arm64? Perhaps wrap it in #ifdef __aarch64__?
Thanks,
Ryan
Quoted 13 lines
tools/testing/selftests/mm/va_high_addr_switch.c +2
-1+static int supported_arch (void ) {#if defined(__powerpc64__)@@ -300,7 +312,7 @@ static int supported_arch(void)#elif defined(__x86_64__)return 1 ;#elif defined(__aarch64__) -return 1 ; +return high_address_present ();#else return 0 ;#endif
Quoted 22 lines · Ryan Roberts
Quoted 21 lines · Dev Jain
Post my improvement of the test:
https://lore.kernel.org/all/[email protected]/
The test begins to fail on 4k and 16k pages, on non-LPA2 systems. To reduce noise in the CI systems, let us skip the test when higher address space is not implemented.Signed-off-by Dev Jain <[email protected]>---
The patch applies on linux-next.tools/testing/selftests/mm/va_high_addr_switch.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)tools/testing/selftests/mm/va_high_addr_switch.c +0
-0@@ -293,6 +293,18 @@ static int run_test(struct testcase *test, int count)return ret; }
Quoted 9 lines
tools/testing/selftests/mm/va_high_addr_switch.c static int run_test(struct testcase *test, int count) · line 300
Quoted 5 lines
tools/testing/selftests/mm/va_high_addr_switch.c +5
-0+/* Check if userspace VA > 48 bits */ +static int high_address_present (void ) +{ +void *ptr =mmap ((void *)(1UL <<50 ),1 ,PROT_READ |PROT_WRITE , +MAP_PRIVATE |MAP_ANONYMOUS |MAP_FIXED ,-1 ,0 );I think there is (very unlikely) possibility that something is already mapped at this address so it will be replaced due to
MAP_FIXED. That could break the test. But the only way something could be already mapped is ifARM64_FORCE_52BITis set and in that case, the test will fail anyway, right? So I think this is fine.
The testcases already assume that high addresses must be empty. Yes, FORCE_52BIT is the only way something could already be mapped at high addresses, but in that case the test fails trivially.
Quoted 9 lines
tools/testing/selftests/mm/va_high_addr_switch.c static int run_test(struct testcase *test, int count) · line 306
Quoted 7 lines
tools/testing/selftests/mm/va_high_addr_switch.c +6
-0+if (ptr ==MAP_FAILED ) +return 0 ; + +munmap (ptr,1 ); +return 1 ; +}I'm guessing this will cause a function-not-used warning on arches other than arm64? Perhaps wrap it in
#ifdef __aarch64__?
Ah yes, I just checked and that is true. I shall post v2 in some time, shall wait if any more comments are there.
Quoted 17 lines
Thanks,
RyanQuoted 13 lines
tools/testing/selftests/mm/va_high_addr_switch.c +2
-1+static int supported_arch (void ) {#if defined(__powerpc64__)@@ -300,7 +312,7 @@ static int supported_arch(void)#elif defined(__x86_64__)return 1 ;#elif defined(__aarch64__) -return 1 ; +return high_address_present ();#else return 0 ;#endif
Ryan Roberts
Reviewed-byQuoted 34 lines · Dev Jain
Quoted 23 lines · Ryan Roberts
Quoted 22 lines · Dev Jain
Post my improvement of the test:
https://lore.kernel.org/all/[email protected]/
The test begins to fail on 4k and 16k pages, on non-LPA2 systems. To reduce noise in the CI systems, let us skip the test when higher address space is not implemented.Signed-off-by Dev Jain <[email protected]>---
The patch applies on linux-next.
tools/testing/selftests/mm/va_high_addr_switch.c| 14 +++++++++++++-1 file changed, 13 insertions(+), 1 deletion(-)tools/testing/selftests/mm/va_high_addr_switch.c +0
-0@@ -293,6 +293,18 @@ static int run_test(struct testcase *test, int count) return ret; }+/* Check if userspace VA > 48 bits */
Quoted 8 lines
tools/testing/selftests/mm/va_high_addr_switch.c static int run_test(struct testcase *test, int count) · line 300
Quoted 4 lines
tools/testing/selftests/mm/va_high_addr_switch.c +4
-0+static int high_address_present (void ) +{ + void *ptr = mmap((void *)(1UL << 50), 1, PROT_READ | PROT_WRITE, + MAP_PRIVATE | MAP_ANONYMOUS | MAP_FIXED, -1, 0);I think there is (very unlikely) possibility that something is already mapped at this address so it will be replaced due to
MAP_FIXED. That could break the test. But the only way something could be already mapped is ifARM64_FORCE_52BITis set and in that case, the test will fail anyway, right? So I think this is fine.
Quoted 3 lines
The testcases already assume that high addresses must be empty. Yes,
FORCE_52BITis the only way something could already be mapped at high addresses, but in that case the test fails trivially.
agreed.
Quoted 11 lines
Quoted 10 lines
tools/testing/selftests/mm/va_high_addr_switch.c static int run_test(struct testcase *test, int count) · line 306
Quoted 8 lines
+ if (ptr ==
MAP_FAILED)
+ return 0;tools/testing/selftests/mm/va_high_addr_switch.c +4
-0+ + munmap(ptr, 1); + return 1; +}I'm guessing this will cause a function-not-used warning on arches other than arm64? Perhaps wrap it in
#ifdef __aarch64__?
Quoted 2 lines
Ah yes, I just checked and that is true. I shall post v2 in some time, shall wait if any more comments are there.
With this fixup:
Quoted 18 lines
Quoted 18 lines
Thanks,
RyanQuoted 13 lines
+
static intsupported_arch(void)
{
#ifdefined(__powerpc64__)tools/testing/selftests/mm/va_high_addr_switch.c +1
-1@@ -300,7 +312,7 @@ static int supported_arch(void) #elif defined(__x86_64__) return 1; #elif defined(__aarch64__) - return 1; + return high_address_present(); #else return 0; #endif