Changeset 13748


Ignore:
Timestamp:
May 30, 2010, 1:57:18 PM (10 years ago)
Author:
gb
Message:

Complain about the lack of "trapped FP exceptions" on VFPv3.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • branches/arm/lisp-kernel/ARM-notes.txt

    r13734 r13748  
    1919don't see any reason to try to optimize for older variants.)
    2020
     21More: recent ARM variants (ARMv7, at least) support vfp v3 with NEON
     22SIMD extenstions.  If vfp v3, the exception-enable bits in the FPSCR
     23always read as 0; that means that FP exceptions can't cause SIGFPE.
     24
     25On a jailbroken iPod Touch, an enabled FP exception seems to force
     26a reboot instead of (or as well as) SIGFPE.
     27
     28Checking to see if an FP operation generated an interesting FP
     29exception involves something like:
     30
     31 (fp-operation fp1 fp2 fp3)
     32 (mov temp1 fpscr) ; the "exception occurred" bits get set and can be read
     33 (load temp2 (tcr.lisp-exception-bits))
     34 (ands temp3 temp1 temp2) ; set flags
     35 (uuo-if (:? ne) ....)
     36
     37That just sucks; it makes FP exception checking a time/space/safety tradeoff.
     38
     39 
     40
     41
    2142
    2243- instruction set
Note: See TracChangeset for help on using the changeset viewer.