[PATCH libXi] 1UL may be 4 bytes, force to 1ULL
Jeremy Huddleston
jeremyhu at apple.com
Tue Sep 20 10:01:45 PDT 2011
I just realized I made (at least one) error in the inline functions I wrote below. s/< 0/< 1/ ... there're probably others, but that just emphasizes my point that there should be conversion routines rather than repeating this code with possible error in other locations.
On Sep 19, 2011, at 7:03 PM, Jeremy Huddleston wrote:
> I'm curious... why is the cls_wire->increment.frac getting cast to unsigned int? It's already typed as uint32_t.
>
> Looking at the context a bit more, I agree with Jamey that this should work (with or without explicit casts):
>
> cls_lib->increment = (double)cls_wire->increment.integral;
> cls_lib->increment += ldexp((double)cls_wire->increment.frac, -32);
>
> But I think it would be more useful to provide something like this in the proto to help with conversion:
>
> static inline double fp1616_to_double(FP1616 in) {
> double ret;
> ret = (double)(in >> 16);
> ret += ldexp((double)(in & 0xff), -16);
> return ret;
> }
>
> static inline double fp3232_to_double(FP3232 in) {
> double ret;
> ret = (double)in.integral;
> ret += ldexp((double)in.frac, -32);
> return ret;
> }
>
>
> static FP1616 double double_to_fp1616(double in) {
> FP1616 ret;
> int32_t integral
> double frac_f;
> uint32_t frac_d;
>
> integral = (int32_t)in << 16;
> frac_f = in - (double)((int32_t)(in));
>
> if (fabs(frac_f) < 0) {
> frac_d = (uint32_t)ldexp(frac_f,16);
> } else {
> /* crap */
> frac_d = 0;
> }
>
> ret = integral << 16;
> ret |= frac & 0xff;
> return ret;
> }
>
> static FP3232 double double_to_fp3232(double in) {
> FP3232 ret;
> int32_t integral
> double frac_f;
> uint32_t frac_d;
>
> integral = (int32_t)in << 16;
> frac_f = in - (double)((int32_t)(in));
>
> if (fabs(frac_f) < 0) {
> frac_d = (uint32_t)ldexp(frac_f,16);
> } else {
> /* crap */
> frac_d = 0;
> }
>
> ret.integral = integral;
> ret.frac = frac_d;
> return ret;
> }
>
> On Sep 19, 2011, at 5:37 PM, Jamey Sharp wrote:
>
>> I'm a bit skeptical... Would ldexp(frac, -32) work for you? Otherwise
>> I'd argue for (UINT64_C(1) << 32).
>>
>> Jamey
>>
>> On Tue, Sep 20, 2011 at 10:19:54AM +1000, Peter Hutterer wrote:
>>> Reported-by: Jeremy Huddleston <jeremyhu at apple.com>
>>> Signed-off-by: Peter Hutterer <peter.hutterer at who-t.net>
>>> ---
>>> something like this good enough?
>>>
>>> src/XExtInt.c | 2 +-
>>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/src/XExtInt.c b/src/XExtInt.c
>>> index d74a8d4..4009285 100644
>>> --- a/src/XExtInt.c
>>> +++ b/src/XExtInt.c
>>> @@ -1577,7 +1577,7 @@ copy_classes(XIDeviceInfo* to, xXIAnyInfo* from, int *nclasses)
>>> cls_lib->scroll_type= cls_wire->scroll_type;
>>> cls_lib->flags = cls_wire->flags;
>>> cls_lib->increment = cls_wire->increment.integral;
>>> - cls_lib->increment += (unsigned int)cls_wire->increment.frac/(double)(1UL << 32);
>>> + cls_lib->increment += (unsigned int)cls_wire->increment.frac/(double)(1ULL << 32);
>>>
>>> to->classes[cls_idx++] = any_lib;
>>> }
>>> --
>>> 1.7.6
>>>
>>> _______________________________________________
>>> xorg-devel at lists.x.org: X.Org development
>>> Archives: http://lists.x.org/archives/xorg-devel
>>> Info: http://lists.x.org/mailman/listinfo/xorg-devel
>
---
Jeremy Huddleston
Rebuild Sudan
- Board of Directors
- http://www.rebuildsudan.org
Berkeley Foundation for Opportunities in Information Technology
- Advisory Board
- http://www.bfoit.org
More information about the xorg-devel
mailing list