This 512 byte structure is saved by using the ia32 fxsave instruction. Thanks, Cristiano.
> I want to use the 512 byte structure which saves the > floating point register state and convert it into a > simulator that uses "long double" to represent each > st(i) register. Each "st" register is represented as > a > 16 bytes with 6 most significant reserved. So only > the > first 10 bytes are the actual register value. > > I want to take this 10 byte representation and > convert > it into a long double variable in my code. My > question > is how do I do it? Can I simply type cast it? > > Is there a correspondence between the 80bit register > value and the represenation that gcc uses for a long > double on ia32? > > Would the below work? > > unsigned char st[10]; // assume this contains the > content of one of the st registers > > long double n = *((long double *) st) > > Would this work? If not, can someone please tell me > how to do this exactly? > > Thanks, > Cristiano. > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam > protection around > http://mail.yahoo.com > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
