[Bug math/30989] New: fesetexcept raises floating-point exception traps on i386

["bruno at clisp dot org" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=30989

            Bug ID: 30989
           Summary: fesetexcept raises floating-point exception traps on
                    i386
           Product: glibc
           Version: 2.35
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: math
          Assignee: unassigned at sourceware dot org
          Reporter: bruno at clisp dot org
  Target Milestone: ---

Created attachment 15188
  --> https://sourceware.org/bugzilla/attachment.cgi?id=15188&action=edit
test case foo.c

According to ISO C 23 § 7.6.4.4, fesetexcept is supposed to set floating-point
exception flags without raising a trap. (Unlike feraiseexcept, which is
supposed to raise a trap if feenableexcept() was called with the appropriate
argument.)
  "This function changes the state of the floating-point exception flags, but
does not cause any other side effects that might be associated with raising
floating-point exceptions. 267)
  Footnote 267) Implementation extensions like traps for floating-point
exceptions and IEC 60559 exception handling do not occur."

That's not how glibc's implementation does it on 32-bit x86.

How to reproduce:
1. Compile the attached program foo.c. On i386 hosts:
$ gcc -Wall foo.c -lm
or on x86_64 hosts:
$ i686-linux-gnu-gcc -Wall foo.c -lm
or
$ gcc -m32 -Wall foo.c -lm

2. Run it.
$ ./a.out
Floating point exception

3. Run it under gdb.
$ gdb a.out 
(gdb) run
Starting program: /home/bruno/a.out 
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".

Program received signal SIGFPE, Arithmetic exception.
0x565562a7 in main ()
(gdb) x/i $eip
=> 0x565562a7 <main+102>:       fld1   


I think it is a hardware limitation that on i386 CPUs, setting a
floating-point exception flag in the 387 unit triggers a trap, when
traps are enabled for the particular exception. But setting a
floating-point exception flag in the SSE unit does not have this
problem, cf. glibc/sysdeps/x86_64/fpu/fesetexcept.c.

Therefore I would expect the fesetexcept() function to operate correctly
when the hardware has an SSE unit (determined e.g. like in
sysdeps/i386/fpu/ftestexcept.c).
And only in the case where the CPU has only the 387 unit, I would expect
the fesetexcept() function to return a non-zero value (as error indicator)
and not modify the 387 floating-point control register, like I suggest
for powerpc hardware in https://sourceware.org/bugzilla/show_bug.cgi?id=30988 .

-- 
You are receiving this mail because:
You are on the CC list for the bug.