Windows x64/win32 Intel 11/10 pgo (non-smp) builds
Wael Deeb's official opening book will need to be updated to work with the new version.
I have included an opening book file created from Olivier Deville's 'Gm2001.pgn in the meantime.
Zach Wegner wrote:I was working a bit on a Windows build. I uploaded some fixes for Windows compatibility, now all that the source needs to compile is to comment out #define ZCT_POSIX and uncomment #define ZCT_WINDOWS. Also, since MSVC doesn't support inline assembly, it needs some intrinsics (BitScanForward64, etc) to run quickly.
Thanks for the update.
vc2005 express optimization stuff breaks it.
maybe it does not know what volatile means .
Windows x64/win32 Intel 11/10 pgo (non-smp) builds
Wael Deeb's official opening book will need to be updated to work with the new version.
I have included an opening book file created from Olivier Deville's 'Gm2001.pgn in the meantime.
Thanks Jim,I can send you a new updated pgn file if you wish
Dr.D
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
Zach Wegner wrote:I was working a bit on a Windows build. I uploaded some fixes for Windows compatibility, now all that the source needs to compile is to comment out #define ZCT_POSIX and uncomment #define ZCT_WINDOWS. Also, since MSVC doesn't support inline assembly, it needs some intrinsics (BitScanForward64, etc) to run quickly.
Thanks for the update.
vc2005 express optimization stuff breaks it.
maybe it does not know what volatile means .
Putting #pragma optimize( "", off ) right before every function that uses LOCK() fixes this issue.
Best
Fonzy
That's pretty strange. Now that I check it here at work, it's doing the same thing. Does not fail that assert on my home NetBSD machine. Not really sure why that is happening, but I will look at the code to see if I can figure it out. It's possible that it is an optimization bug, but it's also quite possible that removing optimization just hides the real problem. The bug is more likely caused by the fact that Unix compiles use processes but Windows use threads...
Windows x64/win32 Intel 11/10 pgo (non-smp) builds
Wael Deeb's official opening book will need to be updated to work with the new version.
I have included an opening book file created from Olivier Deville's 'Gm2001.pgn in the meantime.
F. Bluemers wrote:I managed to "track" it down to find_split_point* in smp2.c
If I turn optimization of for just that one,everything seems fine.
Best
Fonzy
(*and BOOL is_sb=FALSE; // in that one,its nice to be intialized )
Not sure what the issue is then.
That is_sb is also unused anyways, so I removed it. CVS update!
Dr.Wael Deeb wrote:Thanks Jim,I can send you a new updated pgn file if you wish
Dr.D
I could use that too.
I was going to release a new book, but wanted to wait for a binary. SMP should be enabled by default though.
I'll buzz you along with Jim when I am ready with the pgn
Dr.D
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….