[PATCH 5/6] d3dx9_36/tests: Added trigonometric functions test to HLSL test suite
Hello Travis Athougies, In this patch, you have this code: + void *ret = NULL; //... + ret = HeapAlloc(GetProcessHeap(), 0, 4 * samples); + for(i = 0;i < 32;i++) { //... + memcpy((void*) (((DWORD) ret) + (4 * i)), lr.pBits, 4); //... + } I'm not sure, but I think this will cause 64-bit issues (truncating to a 32-bit value?) Even if it doesn't, it seems simpler (more readable) to use a char pointer instead of a void pointer, like + char *ret; //... + ret = HeapAlloc(GetProcessHeap(), 0, 4 * samples); //... + memcpy(ret + (4 * i), lr.pBits, 4); HTH, regards, Joris
On Thu, Aug 5, 2010 at 3:58 PM, Joris Huizer <joris_huizer(a)yahoo.com> wrote:
In this patch, you have this code:
+ void *ret = NULL; //... + ret = HeapAlloc(GetProcessHeap(), 0, 4 * samples); + for(i = 0;i < 32;i++) { //... + memcpy((void*) (((DWORD) ret) + (4 * i)), lr.pBits, 4); //... + }
I'm not sure, but I think this will cause 64-bit issues (truncating to a 32-bit value?) Even if it doesn't, it seems simpler (more readable) to use a char pointer instead of a void pointer, like
+ char *ret; //... + ret = HeapAlloc(GetProcessHeap(), 0, 4 * samples); //... + memcpy(ret + (4 * i), lr.pBits, 4);
Or just use a DWORD pointer: + DWORD *ret; //... + ret = HeapAlloc(GetProcessHeap(), 0, 4 * samples); + for(i = 0;i < 32;i++) { //... + ret[i] = lr.pBits; //... + }
participants (2)
-
Joris Huizer -
Octavian Voicu