Re: [PATCH] d3dx9: D3DXVec3Transform should support input and output parameter overlap.
On 12 April 2016 at 12:09, Paul Gofman <gofmanp(a)gmail.com> wrote:
Fixes #40456.
Signed-off-by: Paul Gofman <gofmanp(a)gmail.com> --- dlls/d3dx9_36/math.c | 11 +++++++---- dlls/d3dx9_36/tests/math.c | 8 ++++++-- 2 files changed, 13 insertions(+), 6 deletions(-)
For what it's worth, the technical term is "aliasing". And yeah, I'd expect this to be an issue in more places, so if you have the time it would be good the check the entire DLL, particularly for things working on vectors, quaternions, matrices, etc.
On 04/12/2016 04:17 PM, Henri Verbeet wrote:
On 12 April 2016 at 12:09, Paul Gofman <gofmanp(a)gmail.com> wrote:
Fixes #40456.
Signed-off-by: Paul Gofman <gofmanp(a)gmail.com> --- dlls/d3dx9_36/math.c | 11 +++++++---- dlls/d3dx9_36/tests/math.c | 8 ++++++-- 2 files changed, 13 insertions(+), 6 deletions(-)
For what it's worth, the technical term is "aliasing". And yeah, I'd expect this to be an issue in more places, so if you have the time it would be good the check the entire DLL, particularly for things working on vectors, quaternions, matrices, etc.
I had a quick a look and actually lot of functions are OK in this sense, though still many Vec3 functions has the same problem. I can deal with them after this one. Should I resend the patch with the different title?
On 12 April 2016 at 16:16, Paul Gofman <gofmanp(a)gmail.com> wrote:
I had a quick a look and actually lot of functions are OK in this sense, though still many Vec3 functions has the same problem. I can deal with them after this one. Should I resend the patch with the different title?
I think it's fine, but it's up to Matteo.
participants (2)
-
Henri Verbeet -
Paul Gofman