René Berber, le Sat 26 Nov 2005 15:19:06 -0600, a écrit : > Samuel Thibault wrote: > [snip] > > This seems odd. In the cygwin case, pagesize() would not be sufficient ?? > > No, in windows there is a requirement that mmap uses memory aligned to 64k > (the > infamous granularity).
Then the result of pagesize() is not sufficiently big: POSIX says that mmap() can return EINVAL if « the address [...] is not a multiple of the page size, or is considered invalid by the implementation ». Well, we could consider non-16-pages-alignment as `considered invalid', but I guess this expression was meant as `not in a mmap()-able area'. > > And anyway, it should rather be > > data2 = (char *) malloc (2 * 16 * pagesize); > > Not only 16 (since it is further 16*pagesize -aligned). > > It's enough to allocate 16 x pagesize to use one page for mmap-ing (like > Corinna > said in one message 15/16 pages will not be aligned -- but one will). If you > see the intermediate results I printed, data2 is re-aligned to 64k correctly. Ah oops, didn't pay attention that the mmap was page-sized. 16 pages are still not enough: if malloc() returns 0x0001 for instance, you'll round up that to 16*pagesize, which is out the allocated area. 17 pages are necessary. BTW, I don't understand why using so complicated code: p += (align - ((long) p & (align-1))) & (align-1) Isn't p = ((long) p + (align-1)) & (align-1) both faster (tested) and more readable ? Regards, Samuel -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
