| Message ID | 20231106132713.953501-8-adhemerval.zanella@linaro.org |
|---|---|
| State | Accepted |
| Commit | d0829302728954e0abacfc01551c17daf4d61c87 |
| Headers | show |
| Series | Multiple floating-point environment fixes | expand |
On 11/6/23 08:27, Adhemerval Zanella wrote: > From: Bruno Haible <bruno@clisp.org> > > The expression > > (excepts & FE_ALL_EXCEPT) << 27 > > produces a signed integer overflow when 'excepts' is specified as > FE_INVALID (= 0x10), because > - excepts is of type 'int', > - FE_ALL_EXCEPT is of type 'int', > - thus (excepts & FE_ALL_EXCEPT) is (int) 0x10, > - 'int' is 32 bits wide. Agreed. > > The patched code produces the same instruction sequence as > previosuly. LGTM. Reviewed-by: Carlos O'Donell <carlos@redhat.com> > --- > sysdeps/hppa/fpu/fclrexcpt.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/sysdeps/hppa/fpu/fclrexcpt.c b/sysdeps/hppa/fpu/fclrexcpt.c > index 055fb04ccc..46caf39ec1 100644 > --- a/sysdeps/hppa/fpu/fclrexcpt.c > +++ b/sysdeps/hppa/fpu/fclrexcpt.c > @@ -26,7 +26,7 @@ feclearexcept (int excepts) > /* Get the current status word. */ > __asm__ ("fstd %%fr0,0(%1)" : "=m" (s.l) : "r" (&s.l) : "%r0"); > /* Clear all the relevant bits. */ > - s.sw[0] &= ~((excepts & FE_ALL_EXCEPT) << 27); > + s.sw[0] &= ~(((unsigned int) excepts & FE_ALL_EXCEPT) << 27); > __asm__ ("fldd 0(%0),%%fr0" : : "r" (&s.l), "m" (s.l) : "%r0"); > > /* Success. */
diff --git a/sysdeps/hppa/fpu/fclrexcpt.c b/sysdeps/hppa/fpu/fclrexcpt.c index 055fb04ccc..46caf39ec1 100644 --- a/sysdeps/hppa/fpu/fclrexcpt.c +++ b/sysdeps/hppa/fpu/fclrexcpt.c @@ -26,7 +26,7 @@ feclearexcept (int excepts) /* Get the current status word. */ __asm__ ("fstd %%fr0,0(%1)" : "=m" (s.l) : "r" (&s.l) : "%r0"); /* Clear all the relevant bits. */ - s.sw[0] &= ~((excepts & FE_ALL_EXCEPT) << 27); + s.sw[0] &= ~(((unsigned int) excepts & FE_ALL_EXCEPT) << 27); __asm__ ("fldd 0(%0),%%fr0" : : "r" (&s.l), "m" (s.l) : "%r0"); /* Success. */
From: Bruno Haible <bruno@clisp.org> The expression (excepts & FE_ALL_EXCEPT) << 27 produces a signed integer overflow when 'excepts' is specified as FE_INVALID (= 0x10), because - excepts is of type 'int', - FE_ALL_EXCEPT is of type 'int', - thus (excepts & FE_ALL_EXCEPT) is (int) 0x10, - 'int' is 32 bits wide. The patched code produces the same instruction sequence as previosuly. --- sysdeps/hppa/fpu/fclrexcpt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)