On 7/28/19 2:12 PM, Thomas Gleixner wrote: > The 32bit variants of vdso_clock_gettime()/getres() have a NULL pointer > check for the timespec pointer. That's inconsistent vs. 64bit. > > But the vdso implementation will never be consistent versus the syscall > because the only case which it can handle is NULL. Any other invalid > pointer will cause a segfault. So special casing NULL is not really useful. > > Remove it along with the superflouos syscall fallback invocation as that > will return -EFAULT anyway. That also gets rid of the dubious typecast > which only works because the pointer is NULL.
Reviewed-by: Vincenzo Frascino <[email protected]> Tested-by: Vincenzo Frascino <[email protected]> > > Fixes: 00b26474c2f1 ("lib/vdso: Provide generic VDSO implementation") > Signed-off-by: Thomas Gleixner <[email protected]> > --- > lib/vdso/gettimeofday.c | 18 ++---------------- > 1 file changed, 2 insertions(+), 16 deletions(-) > > --- a/lib/vdso/gettimeofday.c > +++ b/lib/vdso/gettimeofday.c > @@ -115,9 +115,6 @@ static __maybe_unused int > struct __kernel_timespec ts; > int ret; > > - if (res == NULL) > - goto fallback; > - > ret = __cvdso_clock_gettime(clock, &ts); > > if (ret == 0) { > @@ -126,9 +123,6 @@ static __maybe_unused int > } > > return ret; > - > -fallback: > - return clock_gettime_fallback(clock, (struct __kernel_timespec *)res); > } > > static __maybe_unused int > @@ -204,10 +198,8 @@ int __cvdso_clock_getres(clockid_t clock > goto fallback; > } > > - if (res) { > - res->tv_sec = 0; > - res->tv_nsec = ns; > - } > + res->tv_sec = 0; > + res->tv_nsec = ns; > > return 0; > > @@ -221,9 +213,6 @@ static __maybe_unused int > struct __kernel_timespec ts; > int ret; > > - if (res == NULL) > - goto fallback; > - > ret = __cvdso_clock_getres(clock, &ts); > > if (ret == 0) { > @@ -232,8 +221,5 @@ static __maybe_unused int > } > > return ret; > - > -fallback: > - return clock_getres_fallback(clock, (struct __kernel_timespec *)res); > } > #endif /* VDSO_HAS_CLOCK_GETRES */ > > -- Regards, Vincenzo
