selftests: kvm: Fix a compile error in selftests/kvm/rseq_test.c
PATCH 1/1 selftests: kvm: Fix a compile error in selftests/kvm/rseq_test.c
From: Jinrong Liang <[email protected]>
The following warning appears when executing:
make -C tools/testing/selftests/kvm
rseq_test.c: In function ‘main’:
rseq_test.c:237:33: warning: implicit declaration of function ‘gettid’; did you mean ‘getgid’? [-Wimplicit-function-declaration]
(void *)(unsigned long)gettid());
^~~~~~
getgid/usr/bin/ld: /tmp/ccr5mMko.o: in function `main':../kvm/tools/testing/selftests/kvm/rseq_test.c:237: undefined reference to `gettid'
collect2: error: ld returned 1 exit status
make: *** [../lib.mk:173: ../kvm/tools/testing/selftests/kvm/rseq_test] Error 1
Use the more compatible syscall(SYS_gettid) instead of gettid() to fix it. More subsequent reuse may cause it to be wrapped in a lib file.
--- tools/testing/selftests/kvm/rseq_test.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/testing/selftests/kvm/rseq_test.c +1 -1
@@ -229,7 +229,7 @@ int main(int argc, char *argv[])ucall_init (vm,NULL );pthread_create (&migration_thread,NULL , migration_worker, - (void *)(unsigned long )gettid ()); + (void *)(unsigned long )syscall (SYS_gettid));for (i =0 ; !done; i++) {vcpu_run (vcpu);
Signature · 1 line
2.37.1
Quoted 7 lines · Jinrong Liang
From: Jinrong Liang <[email protected]>
The following warning appears when executing:
make -Ctools/testing/selftests/kvm
rseq_test.c: In function ‘main’:
rseq_test.c:237:33: warning: implicit declaration of function ‘gettid’; did you mean ‘getgid’? [-Wimplicit-function-declaration]
Quoted 7 lines
(void *)(unsigned long)gettid()); ^~~~~~ getgid
/usr/bin/ld:/tmp/ccr5mMko.o: in function `main':../kvm/tools/testing/selftests/kvm/rseq_test.c:237: undefined reference to `gettid'
collect2: error: ld returned 1 exit status
make: *** [../lib.mk:173: ../kvm/tools/testing/selftests/kvm/rseq_test] Error 1
The man page says we need
#define _GNU_SOURCE
#include <unistd.h>
which rseq_test.c doesn't have. We have _GNU_SOURCE, but not unistd.h. IOW, I think this patch can be
tools/testing/selftests/kvm/rseq_test.c +1 -0
@@ -9,6 +9,7 @@#include <string.h> #include <signal.h> #include <syscall.h> +#include <unistd.h> #include <sys/ioctl.h> #include <sys/sysinfo.h> #include <asm/barrier.h>
Thanks,
drew
Quoted 25 lines
Use the more compatible
syscall(SYS_gettid)instead ofgettid()to fix it. More subsequent reuse may cause it to be wrapped in a lib file.Signed-off-by Jinrong Liang <[email protected]>--- tools/testing/selftests/kvm/rseq_test.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)tools/testing/selftests/kvm/rseq_test.c +1
-1@@ -229,7 +229,7 @@ int main(int argc, char *argv[])ucall_init (vm,NULL );pthread_create (&migration_thread,NULL , migration_worker, - (void *)(unsigned long )gettid ()); + (void *)(unsigned long )syscall (SYS_gettid));for (i =0 ; !done; i++) {vcpu_run (vcpu);Signature · 1 line
2.37.1
My ldd version is (GNU libc) 2.28, and I get a compilation error in this case. But I use another ldd (Ubuntu GLIBC 2.31-0ubuntu9.2) 2.31 is compiling fine. This shows that compilation errors may occur in different GNU libc environments. Would it be more appropriate to use syscall for better compatibility?
Thanks,
Jinrong Liang
Quoted 66 lines · Andrew Jones <[email protected]> 于2022年8月2日周二 23:08写道:
Quoted 7 lines · Jinrong Liang
From: Jinrong Liang <[email protected]>
The following warning appears when executing:
make -Ctools/testing/selftests/kvm
rseq_test.c: In function ‘main’:
rseq_test.c:237:33: warning: implicit declaration of function ‘gettid’; did you mean ‘getgid’? [-Wimplicit-function-declaration]Quoted 7 lines
(void *)(unsigned long)gettid()); ^~~~~~ getgid
/usr/bin/ld:/tmp/ccr5mMko.o: in function `main':../kvm/tools/testing/selftests/kvm/rseq_test.c:237: undefined reference to `gettid'
collect2: error: ld returned 1 exit status
make: *** [../lib.mk:173: ../kvm/tools/testing/selftests/kvm/rseq_test] Error 1The man page says we need
#define
_GNU_SOURCE
#include <unistd.h>which
rseq_test.cdoesn't have. We have_GNU_SOURCE, but notunistd.h. IOW, I think this patch can betools/testing/selftests/kvm/rseq_test.c +1
-0@@ -9,6 +9,7 @@#include <string.h> #include <signal.h> #include <syscall.h> +#include <unistd.h> #include <sys/ioctl.h> #include <sys/sysinfo.h> #include <asm/barrier.h> Thanks,
drewQuoted 25 lines
Use the more compatible
syscall(SYS_gettid)instead ofgettid()to fix it. More subsequent reuse may cause it to be wrapped in a lib file.Signed-off-by Jinrong Liang <[email protected]>--- tools/testing/selftests/kvm/rseq_test.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)tools/testing/selftests/kvm/rseq_test.c +1
-2@@ -229,7 +229,7 @@ int main(int argc, char *argv[])ucall_init (vm,NULL );pthread_create (&migration_thread,NULL , migration_worker, - (void *)(unsigned long )gettid ()); + (void *)(unsigned long )syscall (SYS_gettid));for (i =0 ; !done; i++) {vcpu_run (vcpu); --2.37.1
Andrew Jones
Reviewed-byQuoted 4 lines · Jinrong Liang
My ldd version is (GNU libc) 2.28, and I get a compilation error in this case. But I use another ldd (Ubuntu GLIBC 2.31-0ubuntu9.2) 2.31 is compiling fine. This shows that compilation errors may occur in different GNU libc environments. Would it be more appropriate to use syscall for better compatibility?
OK, it's a pity, but no big deal to use syscall().
Thanks,
drew
Sean Christopherson
reviewerQuoted 7 lines · Andrew Jones
Quoted 4 lines · Jinrong Liang
My ldd version is (GNU libc) 2.28, and I get a compilation error in this case. But I use another ldd (Ubuntu GLIBC 2.31-0ubuntu9.2) 2.31 is compiling fine. This shows that compilation errors may occur in different GNU libc environments. Would it be more appropriate to use syscall for better compatibility?
OK, it's a pity, but no big deal to use
syscall().
Ya, https://man7.org/linux/man-pages/man2/gettid.2.html says:
The gettid() system call first appeared on Linux in kernel 2.4.11. Library support was added in glibc 2.30.
But there are already two other instances of syscall(SYS_gettid) in KVM selftests, tools/testing/selftests/kvm/lib/assert.c even adds a _gettid() wrapper.
So rather than having to remember (or discover) to use syscall(SYS_gettid), I wonder if it's possible to conditionally define gettid()? E.g. check for GLIBC version? Or do
#define gettid() syscall(SYS_gettid)
so that it's always available and simply overrides the library's gettid() if it's provided?
Quoted 9 lines · Sean Christopherson
Quoted 7 lines · Andrew Jones
Quoted 4 lines · Jinrong Liang
My ldd version is (GNU libc) 2.28, and I get a compilation error in this case. But I use another ldd (Ubuntu GLIBC 2.31-0ubuntu9.2) 2.31 is compiling fine. This shows that compilation errors may occur in different GNU libc environments. Would it be more appropriate to use syscall for better compatibility?
OK, it's a pity, but no big deal to use
syscall().
Quoted 7 lines
Ya, https://man7.org/linux/man-pages/man2/gettid.2.html says:
The
gettid()system call first appeared on Linux in kernel 2.4.11. Library support was added in glibc 2.30.But there are already two other instances of
syscall(SYS_gettid)in KVM selftests,tools/testing/selftests/kvm/lib/assert.ceven adds a_gettid()wrapper.
Ha! And I found four more in selftests...
testing/selftests/powerpc/include/utils.htesting/selftests/proc/proc.htesting/selftests/rseq/param_test.ctesting/selftests/sched/cs_prctl_test.c
and even more in tools...
Quoted 9 lines
So rather than having to remember (or discover) to use
syscall(SYS_gettid), I wonder if it's possible to conditionally definegettid()? E.g. check for GLIBC version? Or do#define gettid() syscall(SYS_gettid)so that it's always available and simply overrides the library's
gettid()if it's provided?
Sounds good to me. Now the question is where to put it? kvm_util.h, test_util.h, or maybe we should create a new header just for stuff like this?
It doesn't really "fit" in kvm_util.h, but if we put it there, then we greatly reduce the chance that we'll have to revisit this issue again. We could also create a new header just for stuff like this and then include that from kvm_util.h...
Thanks,
drew
Sean Christopherson
reviewerQuoted 18 lines · Andrew Jones
Quoted 9 lines · Sean Christopherson
Quoted 7 lines · Andrew Jones
Quoted 4 lines · Jinrong Liang
My ldd version is (GNU libc) 2.28, and I get a compilation error in this case. But I use another ldd (Ubuntu GLIBC 2.31-0ubuntu9.2) 2.31 is compiling fine. This shows that compilation errors may occur in different GNU libc environments. Would it be more appropriate to use syscall for better compatibility?
OK, it's a pity, but no big deal to use
syscall().Quoted 7 lines
Ya, https://man7.org/linux/man-pages/man2/gettid.2.html says:
The
gettid()system call first appeared on Linux in kernel 2.4.11. Library support was added in glibc 2.30.But there are already two other instances of
syscall(SYS_gettid)in KVM selftests,tools/testing/selftests/kvm/lib/assert.ceven adds a_gettid()wrapper.
Quoted 8 lines
Ha! And I found four more in selftests...
testing/selftests/powerpc/include/utils.htesting/selftests/proc/proc.htesting/selftests/rseq/param_test.ctesting/selftests/sched/cs_prctl_test.cand even more in tools...
Ha, and tools/testing/selftests/sched/cs_prctl_test.c even has the GLIBC crud.
#if __GLIBC_PREREQ(2, 30) == 0
#include <sys/syscall.h>
static pid_t gettid(void)
{
return syscall(SYS_gettid);
}#endif
Quoted 4 lines
Quoted 4 lines
So rather than having to remember (or discover) to use
syscall(SYS_gettid), I wonder if it's possible to conditionally definegettid()? E.g. check for GLIBC version? Or do
Quoted 8 lines
Quoted 4 lines
#define gettid() syscall(SYS_gettid)so that it's always available and simply overrides the library's
gettid()if it's provided?Sounds good to me. Now the question is where to put it?
kvm_util.h,test_util.h, or maybe we should create a new header just for stuff like this?
tools/include/uapi/linux/syscall.h?
Kind of dirty, but not thaaaat dirty.
Quoted 15 lines · Jinrong Liang
From: Jinrong Liang <[email protected]>
The following warning appears when executing:
make -Ctools/testing/selftests/kvm
rseq_test.c: In function ‘main’:
rseq_test.c:237:33: warning: implicit declaration of function ‘gettid’; did you mean ‘getgid’? [-Wimplicit-function-declaration](void *)(unsigned long)gettid()); ^~~~~~ getgid
/usr/bin/ld:/tmp/ccr5mMko.o: in function `main':../kvm/tools/testing/selftests/kvm/rseq_test.c:237: undefined reference to `gettid'
collect2: error: ld returned 1 exit status
make: *** [../lib.mk:173: ../kvm/tools/testing/selftests/kvm/rseq_test] Error 1
Quoted 5 lines
Use the more compatible
syscall(SYS_gettid)instead ofgettid()to fix it. More subsequent reuse may cause it to be wrapped in a lib file.Signed-off-by Jinrong Liang <[email protected]>---
Paolo, do you want to grab this for 6.0? It doesn't look like we're going to have a more elegant solution anytime soon...
Liam Merwick
Reviewed-byQuoted 16 lines · Sean Christopherson
Quoted 15 lines · Jinrong Liang
From: Jinrong Liang <[email protected]>
The following warning appears when executing:
make -Ctools/testing/selftests/kvm
rseq_test.c: In function ‘main’:
rseq_test.c:237:33: warning: implicit declaration of function ‘gettid’; did you mean ‘getgid’? [-Wimplicit-function-declaration](void *)(unsigned long)gettid()); ^~~~~~ getgid
/usr/bin/ld:/tmp/ccr5mMko.o: in function `main':../kvm/tools/testing/selftests/kvm/rseq_test.c:237: undefined reference to `gettid'
collect2: error: ld returned 1 exit status
make: *** [../lib.mk:173: ../kvm/tools/testing/selftests/kvm/rseq_test] Error 1
Quoted 8 lines
Quoted 5 lines
Use the more compatible
syscall(SYS_gettid)instead ofgettid()to fix it. More subsequent reuse may cause it to be wrapped in a lib file.Signed-off-by Jinrong Liang <[email protected]>---
Reviewed-by Sean Christopherson <[email protected]>
Can a 'Cc: [email protected]' be added also as e923b0537d28 got backported to v5.15.58
Quoted 3 lines
Paolo, do you want to grab this for 6.0? It doesn't look like we're going to have a more elegant solution anytime soon...
Quoted 23 lines · Liam Merwick
Quoted 22 lines · Sean Christopherson
Quoted 20 lines · Jinrong Liang
From: Jinrong Liang <[email protected]>
The following warning appears when executing:
make -Ctools/testing/selftests/kvm
rseq_test.c: In function ‘main’:
rseq_test.c:237:33: warning: implicit declaration of function ‘gettid’; did you mean ‘getgid’? [-Wimplicit-function-declaration] (void *)(unsigned long)gettid());
^~~~~~
getgid/usr/bin/ld:/tmp/ccr5mMko.o: in function `main':../kvm/tools/testing/selftests/kvm/rseq_test.c:237: undefined reference to `gettid'
collect2: error: ld returned 1 exit status
make: *** [../lib.mk:173: ../kvm/tools/testing/selftests/kvm/rseq_test] Error 1Use the more compatible
syscall(SYS_gettid)instead ofgettid()to fix it. More subsequent reuse may cause it to be wrapped in a lib file.Signed-off-by Jinrong Liang <[email protected]>---
Quoted 12 lines
Quoted 2 lines
Reviewed-by Sean Christopherson <[email protected]>Can a 'Cc: [email protected]' be added also as
e923b0537d28got backported to v5.15.58Reviewed-by Liam Merwick <[email protected]>Quoted 3 lines
Paolo, do you want to grab this for 6.0? It doesn't look like we're going to have a more elegant solution anytime soon...
Ping?
Quoted 41 lines · JinrongLiang
Quoted 37 lines · Liam Merwick
Quoted 19 lines · Sean Christopherson
Quoted 18 lines · Jinrong Liang
From: Jinrong Liang <[email protected]>
The following warning appears when executing:
make -Ctools/testing/selftests/kvm
rseq_test.c: In function ‘main’:
rseq_test.c:237:33: warning: implicit declaration of function ‘gettid’; did you mean ‘getgid’? [-Wimplicit-function-declaration](void *)(unsigned long)gettid()); ^~~~~~ getgid
/usr/bin/ld:/tmp/ccr5mMko.o: in function `main':../kvm/tools/testing/selftests/kvm/rseq_test.c:237: undefined reference to `gettid'
collect2: error: ld returned 1 exit status
make: *** [../lib.mk:173:
../kvm/tools/testing/selftests/kvm/rseq_test] Error 1Quoted 9 lines
Quoted 6 lines
Use the more compatible
syscall(SYS_gettid)instead ofgettid()to fix it.
More subsequent reuse may cause it to be wrapped in a lib file.Signed-off-by Jinrong Liang <[email protected]>---
Reviewed-by Sean Christopherson <[email protected]>Can a 'Cc: [email protected]' be added also as
e923b0537d28got backported to v5.15.58Reviewed-by Liam Merwick <[email protected]>Quoted 1 line
Quoted 5 lines
Quoted 4 lines
Quoted 3 lines
Paolo, do you want to grab this for 6.0? It doesn't look like we're going to have
a more elegant solution anytime soon...Ping?
+1, please grab this. Every new branch I create (for new development and to review upstream code) I have to remember to cherry-pick this patch in because I use Clang.
Queued, thanks.
Paolo