mhps
|
 |
October 13, 2013, 01:25:32 AM Last edit: October 14, 2013, 02:16:32 AM by mhps |
|
On my Tesla c2075 running with -D the new released miner (x64) chose -F112x8 without doing the grid test. Is this the correct behavior? I get maybe 2-3 khash increase to 163k/s. Longer test is needed to see if there is a real increase. The last miner gave 160k/s
edit: a longer test showed that the increase was more like 5khash/s.
|
|
|
|
|
|
|
|
"Bitcoin: the cutting edge of begging technology." -- Giraffe.BTC
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
John122
|
 |
October 14, 2013, 12:53:54 AM Last edit: October 14, 2013, 01:20:56 AM by John122 |
|
I got a huge drop in hashrate with dual GTX580 with the October 10 version, from 240 Kh/s each to 180 Kh/s...  I autotuned again, which gave me the same config as I was using on the older version, and I can't get a better hashrate. Edit : I was able to retreive my old performance by using -H 1, seems my build doesn't like to send SHA-256 work to the CPU...
|
|
|
|
Eli0t
|
 |
October 14, 2013, 01:15:43 AM |
|
since it was requested, heres fruittool's 64bit build, sorry if you dont have a modern browser or are otherwise unable to use mega https://mega.co.nz/#!gZwRDSwQ!GyYVJTQSAWrS2WwKSwnb_U7a9jZ7cgfqltLeslkyTDk Note: cudart64_55.dll is a renamed version of cudart64_50_35.dll since 55 doesnt work with old drivers, the dll name was hardcoded and new drivers suck for 5xx series cards the real cudart64_55.dll is included as cudart64_55.rar. use at your own risk
|
LTC: LKpJf3uk7KsHU73kxq8iFJrP1AAKN7Yni7 DGC: DKXGvEbj3Rwgrm2QQbRyNPDDZDYoq4Y44d XPM: AWV5AKfLFyoBaMjg9C77rGUBhuFxz5DGGL
|
|
|
Jriker1
|
 |
October 14, 2013, 06:47:05 PM Last edit: October 14, 2013, 06:59:48 PM by Jriker1 |
|
since it was requested, heres fruittool's 64bit build, sorry if you dont have a modern browser or are otherwise unable to use mega https://mega.co.nz/#!gZwRDSwQ!GyYVJTQSAWrS2WwKSwnb_U7a9jZ7cgfqltLeslkyTDk Note: cudart64_55.dll is a renamed version of cudart64_50_35.dll since 55 doesnt work with old drivers, the dll name was hardcoded and new drivers suck for 5xx series cards the real cudart64_55.dll is included as cudart64_55.rar. use at your own risk I am using the 32 bit version of cudaminer from 7/13 and ran into this before but resolved a msvcr100.dll missing error. With this version error is back. Thoughts? Thanks. JR
|
|
|
|
cbuchner1 (OP)
|
 |
October 14, 2013, 07:12:46 PM |
|
get the Visual C++ 2010 SP1 runtime for 64 bit this time
|
|
|
|
Jriker1
|
 |
October 14, 2013, 09:21:13 PM |
|
get the Visual C++ 2010 SP1 runtime for 64 bit this time
Bingo. Thanks.
|
|
|
|
cbuchner1 (OP)
|
 |
October 16, 2013, 09:58:00 AM Last edit: October 17, 2013, 12:44:11 PM by cbuchner1 |
|
I got the dreaded hang again (in 2013-10-10 release), after the "failed to parse newline terminated string" message.
The data we get from the socket is just an empty null terminated string, then the subroutine to parse the string then returns a NULL pointer, after which the code tries to disconnect the tcp connection.
Then the program hangs in the subroutine that is supposed to disconnect the TCP socket. It could be that the code cannot acquire the mutex (held by some other thread), or one of the winsock calls might be blocking.
Christian
|
|
|
|
cbuchner1 (OP)
|
 |
October 18, 2013, 09:56:34 PM |
|
it would appear the stratum_send_line function is holding the socket mutex while the stratum_disconnect function wants to acquire the mutex. I was able to attach the debugger to a hanging instance of cudaminer.
Christian
|
|
|
|
cbuchner1 (OP)
|
 |
October 25, 2013, 09:58:21 PM |
|
I found the root cause of the socket issues in stratum
in miner.h my typedef size_t ssize_t; should have been a typedef int ssize_t; instead. Then negative return codes for send() and other socket calls can be interpreted and acted upon correctly. I will publish a fixed version this weekend.
Christian
|
|
|
|
mhps
|
 |
October 26, 2013, 01:22:16 AM |
|
I found the root cause of the socket issues in stratum
in miner.h my typedef size_t ssize_t; should have been a typedef int ssize_t; instead. Then negative return codes for send() and other socket calls can be interpreted and acted upon correctly. I will publish a fixed version this weekend.
Christian
Great!
|
|
|
|
Spoetnik
Legendary
Offline
Activity: 1540
Merit: 1011
FUD Philanthropist™
|
 |
October 26, 2013, 02:30:31 AM |
|
ya i looked at that before and kinda wondered what the point was lol
|
FUD first & ask questions later™
|
|
|
Icon
|
 |
October 26, 2013, 06:42:48 AM |
|
btw lub this miner finally can mine with my nvidia cards.. one thing next release please link to 64 bit windows version and tyvm.
Icon
|
|
|
|
|
Spoetnik
Legendary
Offline
Activity: 1540
Merit: 1011
FUD Philanthropist™
|
 |
October 27, 2013, 05:44:48 AM Last edit: October 27, 2013, 05:55:43 AM by Spoetnik |
|
@Batshit what version are you running because stratum has been supported by the last few updates. so if your using a version fairly new ? there is no reason to use a stratum proxy.. edit: i see you are running the newest one i think so no you should NOT need stratum proxy also Stratum will kick in by default i believe so no special cmd-lines need to be added for it try running program with --help command line option in a bat file also has anyone seen errors when mining on wallets ? (any of them) i have tinkered around and i don't why but i get a connection error message and a reconnect randomly.. it gives me an error message then reconnects and carries on.. weird cause i can go long periods of time with out one of those errors then wham tons of them out of the blue it seems.. 
|
FUD first & ask questions later™
|
|
|
Batshit
Newbie
Offline
Activity: 48
Merit: 0
|
 |
October 27, 2013, 07:31:33 AM |
|
I tried without the proxy with same results as in screen shot. Command line I used was: cudaminer.exe -a scrypt -u Batshit.GPU -p p4ssword I have since installed cpuminer and it is cranking along just fine at an awesome 54 khs/s. It's an always-on and underutilized home business server, so meh...whatever.  I really want to get the GPU hashing to maximize my losses. 
|
|
|
|
Batshit
Newbie
Offline
Activity: 48
Merit: 0
|
 |
October 27, 2013, 07:38:48 AM |
|
Also tried using cudaminer.exe -a scrypt -o stratum+tcp://us3.wemineltc.com:3333 -u Batshit.GPU -p p4ssword Still no joy. 
|
|
|
|
GoldBit89
|
 |
October 27, 2013, 08:12:10 AM |
|
Nooob Alert! The Stratum proxy seems to be running, but cudaminer is not doing anything after it launches. What am I doing wrong (see screen shots)? Windows 8 Enterprise x64 nVidia GT 610 I allowed cudaminer through the firewall when prompted by Windows 8.  mine does the same--at first i thought it wasnt working but on mining webpage it shows i was mining so i looked at logfiles and sure enough while it doesnt show on the command prompt it does log every line like usual but just in the log file and not on command screen prompt. so i would say check your miner webpage and see if yours is also running but not outputting to command prompt.
|
FTC 6nvzqqaCEizThvgMeC86MGzhAxGzKEtNH8 |WDC WckDxipCes2eBmxrUYEhrUfNNRZexKuYjR |BQC bSDm3XvauqWWnqrxfimw5wdHVDQDp2U8XU BOT EjcroqeMpZT4hphY4xYDzTQakwutpnufQR |BTG geLUGuJkhnvuft77ND6VrMvc8vxySKZBUz |LTC LhXbJMzCqLEzGBKgB2n73oce448BxX1dc4 BTC 1JPzHugtBtPwXgwMqt9rtdwRxxWyaZvk61 |ETH 0xA6cCD2Fb3AC2450646F8D8ebeb14f084F392ACFf
|
|
|
Batshit
Newbie
Offline
Activity: 48
Merit: 0
|
 |
October 27, 2013, 08:42:33 AM Last edit: October 28, 2013, 02:23:53 AM by Batshit |
|
Batshit.CPU is showing as hashing. Batshit.GPU is still 0 after @20+ minutes. How to access the log file you mentioned? BTW, I am running without the proxy now. Hurray. I also figured out how to launch the miner and automagically set the priority. So, I put it in a .bat file and it runs on startup. That way, it's always running in the background and doesn't get in the way of my machine's real workload which does tend to be intense sometimes (during backups and activities I do for work, for examples). I have the CPU (minerd.exe) running at low priority and the GPU (cudaminer.exe) [if it ever decides to start working] running at below normal priority so they don't bump. Here's the command line: C:\Windows\System32\cmd.exe /c start "WhateverWindowTitleNameYouWant" /LOW C:\CPUMiner\minerd.exe -a scrypt -o stratum+tcp://us3.wemineltc.com:3333 -u workernamehere -p p4sswordor, one could simply copy and paste the info below as-is (ahem). Works like a charm!  C:\Windows\System32\cmd.exe /c start "Batshit.CPU" /LOW C:\CPUMiner\minerd.exe -a scrypt -o stratum+tcp://us3.wemineltc.com:3333 -u Batshit.CPU -p p4sswordMOD: Please add to a sticky. It took me awhile to figure this out and others may find it useful.
|
|
|
|
GoldBit89
|
 |
October 27, 2013, 09:21:11 AM |
|
Batshit.CPU is showing as hashing. Batshit.GPU is still 0 after @20+ minutes. How to access the log file you mentioned? BTW, I am running without the proxy now. Hurray. I also figured out how to launch the miner and automagically set the priority. So, I put it in a .bat file and it runs on startup. That way, it's always running in the background and doesn't get in the way of my machine's real workload which does tend to be intense sometimes (during backups and activities I do for work, for examples). I have the CPU (minerd.exe) running at low priority and the GPU (cudaminer.exe) [if it ever decides to start working] running at above normal priority so they don't bump. Here's the command line: C:\Windows\System32\cmd.exe /c start "WhateverWindowTitleNameYouWant" /LOW C:\CPUMiner\minerd.exe -a scrypt -o stratum+tcp://us3.wemineltc.com:3333 -u workernamehere -p p4sswordor, one could simply copy and paste the info below as-is (ahem). Works like a charm!  C:\Windows\System32\cmd.exe /c start "Batshit.CPU" /LOW C:\CPUMiner\minerd.exe -a scrypt -o stratum+tcp://us3.wemineltc.com:3333 -u Batshit.CPU -p p4sswordMOD: Please add to a sticky. It took me awhile to figure this out and others may find it useful. its here : C:\Windows\System32\logfile.txt here is a small sample of what it looks like in logfile: [2013-10-27 04:08:27] GPU #0: GeForce 9600 GT, 769024 hashes, 13.88 khash/s [2013-10-27 04:09:22] LONGPOLL detected new block [2013-10-27 04:09:22] DEBUG: got new work [2013-10-27 04:09:22] GPU #0: GeForce 9600 GT, 773120 hashes, 13.99 khash/s [2013-10-27 04:09:46] DEBUG: hash <= target Hash: 0000003fa4e40222ec0de9e3d142af935494ede3436f4e521c475f5a5a672c30 Target: 000007ffffffffffffffffffffffffffffffffffffffffffffffffffffffff
|
FTC 6nvzqqaCEizThvgMeC86MGzhAxGzKEtNH8 |WDC WckDxipCes2eBmxrUYEhrUfNNRZexKuYjR |BQC bSDm3XvauqWWnqrxfimw5wdHVDQDp2U8XU BOT EjcroqeMpZT4hphY4xYDzTQakwutpnufQR |BTG geLUGuJkhnvuft77ND6VrMvc8vxySKZBUz |LTC LhXbJMzCqLEzGBKgB2n73oce448BxX1dc4 BTC 1JPzHugtBtPwXgwMqt9rtdwRxxWyaZvk61 |ETH 0xA6cCD2Fb3AC2450646F8D8ebeb14f084F392ACFf
|
|
|
Spoetnik
Legendary
Offline
Activity: 1540
Merit: 1011
FUD Philanthropist™
|
 |
October 27, 2013, 09:42:26 AM |
|
Batshit nice name by the way lol first off i have not used the version of cudaminer you are using, i use my own 64bit version i modded for myself so i have never seen that message in your picture before ever once.. but please go and double check and verify you are not running the cpu miner and gpu miner on the same worker that may explain why your getting an error like that.. also not sure if its available but i run my cpu/gpu miners with -q that is quiet mode  each miner should be running on their own worker to fetch their own work..
|
FUD first & ask questions later™
|
|
|
|