@stonefoz
right on man thanks for your advice
the git build yesterday had compile errors for me i mentioned already
if needed i could do it again log them if anyone wants to see them, myself i just ignored them it looked like a mess..
especially compared to Visual Studio's build output lol
i first tried to port the program to VC++ on VS 2010 and almost got it working before i gave up.
i have already made a GUI version of cudaminer (another Pooler CPU miner fork) on VS using MFC so i kinda think its doable to support GCC and MSVC.
your comment on SSE2 was intriguing and so is that other guys test build "minerd64_CHEAT.exe"
i have hex compared the first v2 build uploaded that came with different Instruction sets and then compiled my own and compared.
i some experience using IDA, OllyDBG etc patching programs over the years so i know some asm.. never used it much in c/c++ code though.
point being i took one of the builds last not (not the newest) and tried -msse2 and msee4.1 and did see they produced different code..
using -O3 also which turns out via google search enables the highest level of optimizations.
Anyway the SSE4.1 build has been yielding me a noticeable increase in speed running it over night.. max high of 202 for me now to 215 regularly
and all i did was turn off AVX, AVX2 and XOP in the cpuminer.h config and rebuild with sse4.1 (the max my Intel C2D supports)
and i didn't do anything yet lol
for now all i did was add all the files to a project in Visual Studio and use GCC to compile.. that way i get best of both for now
What i'm really wondering right now is what is with that guy that said we all need to patch our miners a few pages back ?
He was saying something about it fixing reporting to P2Pools ? something to do with pseudo-share settings ?
215.09 on my C2D E7500 using both cores -s 1 mofo's
I can get more I KNOW it lol
right on man thanks for your advice
the git build yesterday had compile errors for me i mentioned already
if needed i could do it again log them if anyone wants to see them, myself i just ignored them it looked like a mess..
especially compared to Visual Studio's build output lol
i first tried to port the program to VC++ on VS 2010 and almost got it working before i gave up.
i have already made a GUI version of cudaminer (another Pooler CPU miner fork) on VS using MFC so i kinda think its doable to support GCC and MSVC.
your comment on SSE2 was intriguing and so is that other guys test build "minerd64_CHEAT.exe"
i have hex compared the first v2 build uploaded that came with different Instruction sets and then compiled my own and compared.
i some experience using IDA, OllyDBG etc patching programs over the years so i know some asm.. never used it much in c/c++ code though.
point being i took one of the builds last not (not the newest) and tried -msse2 and msee4.1 and did see they produced different code..
using -O3 also which turns out via google search enables the highest level of optimizations.
Anyway the SSE4.1 build has been yielding me a noticeable increase in speed running it over night.. max high of 202 for me now to 215 regularly
and all i did was turn off AVX, AVX2 and XOP in the cpuminer.h config and rebuild with sse4.1 (the max my Intel C2D supports)
and i didn't do anything yet lol
for now all i did was add all the files to a project in Visual Studio and use GCC to compile.. that way i get best of both for now
What i'm really wondering right now is what is with that guy that said we all need to patch our miners a few pages back ?
He was saying something about it fixing reporting to P2Pools ? something to do with pseudo-share settings ?
215.09 on my C2D E7500 using both cores -s 1 mofo's
I can get more I KNOW it lol