Re: Memory Management on AMD64 in 32-bit mode

2004-01-19 Thread Joe Buehler
Benson Margulies wrote: Would anyone be willing to elaborate on how the existing code is going about accomplishing the task at hand? If not, at least knowing that this is the goal of the exercise should make it easier to get a clue. I believe that there is some documentation in a file in the CVS

Re: Memory Management on AMD64 in 32-bit mode

2004-01-18 Thread Benson Margulies
I've belatedly located Corinna Vinschen's email message about the requirement for the stack address in fork. Would anyone be willing to elaborate on how the existing code is going about accomplishing the task at hand? If not, at least knowing that this is the goal of the exercise should make it

Re: Memory Management on AMD64 in 32-bit mode

2003-12-10 Thread Corinna Vinschen
On Dec 9 12:55, Benson Margulies wrote: I assume that there's a very strong reason why this code can't just allocate a stack any-old-place (calling VirtualAlloc with first arg 0) and use it. What I don't understand is the nature of the constraints. If the parent needs to know, why not have

Re: Memory Management on AMD64 in 32-bit mode

2003-12-09 Thread jurgen . defurne
: (bcc: Jurgen Defurne/BRG/CE/PHILIPS) Subject:Re: Memory Management on AMD64 in 32-bit mode Classification: On Mon, Dec 08, 2003 at 07:30:17PM -0500, Benson Margulies wrote: I happen to have access to a an AMD64 system running Windows. All the cygwin shells fail

RE: Memory Management on AMD64 in 32-bit mode

2003-12-09 Thread Benson Margulies
The error message rather unambiguously indicates that VirtualAlloc is returning 0 with GetLastError() == 0. The call in question calls VirtualAlloc with parameters derived from a call to VirtualQuery against some stack storage. It seems that this version of Windows is not altogether pleased to see

RE: Memory Management on AMD64 in 32-bit mode

2003-12-09 Thread Benson Margulies
This looks like a logic error in fork.cc/dcrt0.cc to me, but I'm probably not understanding something. alloc_stack_hard_way assumes that the memory at ci-stacktop is available. ci-stacktop is set to be the region of memory that contains a stack variable in the parent process at the time that

Re: Memory Management on AMD64 in 32-bit mode

2003-12-08 Thread Christopher Faylor
On Mon, Dec 08, 2003 at 07:30:17PM -0500, Benson Margulies wrote: I happen to have access to a an AMD64 system running Windows. All the cygwin shells fail to fork with the following. I'd be willing to try coding and running a patch if the experts would care to offer an idea of what to try. Sorry