ewibit
Legendary
 Offline
Activity: 2955
Merit: 1049
|
 |
October 22, 2013, 09:46:07 AM
Last edit: October 22, 2013, 01:12:34 PM by ewibit |
|
Still getting the error below with 3.6.4:
Using USB 2.0 ports directly on the motherboard, restarting CGMINER seems to fix the issue. Tried different USB ports, same issue. Can happen 20 minuter or hours after CGMINER.
here the same [2013-10-22 11:43:43] AMU11: Comms error (werr=-7 amt=0)
[2013-10-22 11:43:43] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:43] AMU 8 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:43] AMU8: Comms error (werr=-7 amt=0)
[2013-10-22 11:43:43] AMU 11 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:43] AMU11: Comms error (werr=-7 amt=0)
[2013-10-22 11:43:43] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:44] AMU 8 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:44] AMU8: Comms error (werr=-7 amt=0)
[2013-10-22 11:43:44] AMU 11 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:44] AMU11: Comms error (werr=-7 amt=0)
[2013-10-22 11:43:44] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:44] AMU 8 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 11:43:44] AMU8: Comms error (werr=-7 amt=0)
[2013-10-22 11:43:44] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
 |
|
|
|
|
|
|
|
|
|
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
greek_hephaestus
Full Member
Offline
Activity: 162
Merit: 100
Eloncoin.org - Mars, here we come!
|
|
October 22, 2013, 12:54:58 PM |
|
Helo CKolivas and Kano,
Do you have any plans to release a KnCMiner version of CGMiner?
Thank you very much for your contribution to Cryptocoin Mining!!!
KFC sent me a saturn unit for development. I am currently swamped with other work but will eventually get around to working on integrating support for it into master cgminer and hopefully improve on the driver too. Thank you Very much for your answer!!! |
|
|
|
|
c1010010
|
|
October 22, 2013, 12:58:33 PM |
|
Do you always plan to edit the subject line with the new version #?
Makes it easy to see if there was a change, or if there was just a comment on the thread.
|
|
|
|
-ck (OP)
Legendary
Offline
Activity: 4088
Merit: 1630
Ruu \o/
|
|
October 22, 2013, 01:11:07 PM |
|
...it is always edited with the current version.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel
2% Fee Solo mining at solo.ckpool.org
-ck
|
|
|
|
c1010010
|
|
October 22, 2013, 04:09:24 PM |
|
...it is always edited with the current version.
Thanks. Apparently my window is truncating it... I see that you are completely right. Thanks. |
|
|
|
mdude77
Legendary
Offline
Activity: 1540
Merit: 1001
|
|
October 22, 2013, 07:05:10 PM |
|
More 3.6.4 feedback: I moved all my USB erupters to another machine. I found cgminer doesn't see any of those plugged into my blue USB 3.0 ports. This is the same experience I had with my laptop. They work fine on the non blue ports. I'm also getting problems with devices zombifying, and the AMU write errors. I did not see any of these errors on the old PC, so it seems once again this is hardware specific. Last time it was AMU 7 that zombified. cgminer version 3.6.4 - Started: [2013-10-22 14:59:20]
-------------------------------------------------------------------------------
(5s):11.27G (avg):10.52Gh/s | A:436 R:2 HW:14 WU:162.2/m
ST: 2 SS: 1 NB: 2 LW: 791 GF: 0 RF: 0
Connected to llllll diff 8 with stratum as user kkkkk
Block: 0000075bdc49ed00... Diff:268M Started: [15:00:47] Best share: 4.07K
-------------------------------------------------------------------------------
[P]ool management [S]ettings [D]isplay options [Q]uit
AMU 0: | 335.4M/326.9Mh/s | A:17 R:0 HW:2 WU: 11.3/m
AMU 1: | 335.7M/337.2Mh/s | A:40 R:0 HW:0 WU: 7.0/m
AMU 2: | 335.7M/332.3Mh/s | A: 8 R:2 HW:0 WU: 4.8/m
AMU 3: | 335.4M/325.6Mh/s | A:16 R:0 HW:0 WU: 2.9/m
AMU 4: | 335.6M/332.1Mh/s | A: 8 R:0 HW:1 WU: 4.8/m
AMU 5: | 335.4M/315.4Mh/s | A:24 R:0 HW:0 WU: 6.6/m
AMU 6: | 335.5M/324.8Mh/s | A:16 R:0 HW:0 WU: 3.7/m
AMU 7: | 335.7M/333.9Mh/s | A:12 R:0 HW:1 WU: 5.9/m
AMU 8: | 335.6M/336.4Mh/s | A: 8 R:0 HW:0 WU: 4.8/m
AMU 9: | 335.6M/334.5Mh/s | A: 8 R:0 HW:0 WU: 3.7/m
AMU 11: | 335.6M/336.7Mh/s | A:17 R:0 HW:1 WU: 6.6/m
AMU 12: | 335.6M/331.3Mh/s | A: 8 R:0 HW:1 WU: 2.2/m
AMU 13: | 335.2M/325.4Mh/s | A: 0 R:0 HW:0 WU: 3.7/m
AMU 14: | 335.8M/324.8Mh/s | A: 0 R:0 HW:1 WU: 4.0/m
AMU 15: | 335.4M/332.2Mh/s | A:32 R:0 HW:0 WU: 5.5/m
AMU 16: | 335.7M/332.6Mh/s | A:24 R:0 HW:0 WU: 4.4/m
AMU 17: | 335.2M/313.6Mh/s | A:16 R:0 HW:0 WU: 6.2/m
AMU 18: | 335.5M/315.4Mh/s | A: 1 R:0 HW:0 WU: 4.0/m
AMU 19: | 335.7M/320.8Mh/s | A: 0 R:0 HW:1 WU: 3.3/m
AMU 20: | 335.6M/334.4Mh/s | A: 0 R:0 HW:0 WU: 7.0/m
AMU 21: | 335.5M/325.1Mh/s | A: 8 R:0 HW:0 WU: 4.4/m
AMU 22: | 335.6M/335.8Mh/s | A:32 R:0 HW:0 WU: 3.3/m
AMU 23: | 335.4M/329.7Mh/s | A:16 R:0 HW:1 WU: 5.9/m
AMU 24: | 335.6M/315.4Mh/s | A: 1 R:0 HW:0 WU: 5.1/m
AMU 25: | 335.7M/324.9Mh/s | A:48 R:0 HW:1 WU: 6.6/m
AMU 26: | 335.7M/331.0Mh/s | A:32 R:0 HW:1 WU: 5.5/m
AMU 27: | 335.6M/320.1Mh/s | A: 9 R:0 HW:0 WU: 2.2/m
AMU 28: | 335.7M/329.1Mh/s | A: 1 R:0 HW:0 WU: 4.4/m
AMU 29: | 335.5M/333.1Mh/s | A: 0 R:0 HW:1 WU: 5.9/m
AMU 30: | 335.5M/324.5Mh/s | A:24 R:0 HW:0 WU: 6.6/m
AMU 31: | 335.4M/330.4Mh/s | A: 8 R:0 HW:0 WU: 5.1/m
AMU 32: | 335.8M/324.1Mh/s | A: 0 R:0 HW:2 WU: 4.4/m
AMU 10: | ZOMBIE/93.54Mh/s | A: 2 R:0 HW:0 WU: 1.8/m
-------------------------------------------------------------------------------
[2013-10-22 15:01:51] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 15:01:52] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 15:01:58] Accepted 1a733b3e Diff 9/8 AMU 21 pool 0
[2013-10-22 15:02:01] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 15:02:02] AMU 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
[2013-10-22 15:02:02] Accepted 08989b2a Diff 29/8 AMU 15 pool 0
M |
I mine at Kano's Pool because it pays the best and is completely transparent! Come join me! |
|
|
-ck (OP)
Legendary
Offline
Activity: 4088
Merit: 1630
Ruu \o/
|
|
October 22, 2013, 08:42:37 PM |
|
More 3.6.4 feedback:
I moved all my USB erupters to another machine. I found cgminer doesn't see any of those plugged into my blue USB 3.0 ports. This is the same experience I had with my laptop. They work fine on the non blue ports. I'm also getting problems with devices zombifying, and the AMU write errors. I did not see any of these errors on the old PC, so it seems once again this is hardware specific. Last time it was AMU 7 that zombified.
M
Thanks M. It seems the massive rewrite I did to fix random timeouts fixed something intrinsic in libusb's own timeouts, but USB 3 slot vs USB1.1 device related issues are different again. I do know that the WinUSB driver itself is not complete with USB3 support and some hardware the devices won't even show up, so there is still hope. I will keep chipping away at whatever other internals I can on cgminer and hopefully newer versions of WinUSB come out that get better (the version on my site bundled with zadig is still the latest version). |
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel
2% Fee Solo mining at solo.ckpool.org
-ck
|
|
|
mdude77
Legendary
Offline
Activity: 1540
Merit: 1001
|
|
October 22, 2013, 10:05:42 PM |
|
More 3.6.4 feedback:
I moved all my USB erupters to another machine. I found cgminer doesn't see any of those plugged into my blue USB 3.0 ports. This is the same experience I had with my laptop. They work fine on the non blue ports. I'm also getting problems with devices zombifying, and the AMU write errors. I did not see any of these errors on the old PC, so it seems once again this is hardware specific. Last time it was AMU 7 that zombified.
M
Thanks M. It seems the massive rewrite I did to fix random timeouts fixed something intrinsic in libusb's own timeouts, but USB 3 slot vs USB1.1 device related issues are different again. I do know that the WinUSB driver itself is not complete with USB3 support and some hardware the devices won't even show up, so there is still hope. I will keep chipping away at whatever other internals I can on cgminer and hopefully newer versions of WinUSB come out that get better (the version on my site bundled with zadig is still the latest version). One additional note: the cgminer-nogpu app hasn't been able to end properly for me for ages. If I press Q, it starts the shutdown routine, I see thread errors all over the place, and it doesn't end. Closing the box has always ended it however. Until now. Whatever was going wrong under the hood is severe enough I can't get it to end. Closing it doesn't work, q doesn't work, end task doesn't work. All the lights came back on my erupters, so it did stop mining. It just won't end. I fear this app will be with me until my next reboot.  Just a reminder, this is win7x64 with all updates. M |
I mine at Kano's Pool because it pays the best and is completely transparent! Come join me! |
|
|
|
Askit2
|
|
October 22, 2013, 10:11:14 PM |
|
snip
Well that definitely sounds better than it was. The only other thing I've seen suggested is using a good powered USB supply to the RPi. Perhaps a shorter, thicker cable from a USB3 slot if running off a PC, or a well powered hub if not. My Usb cable was the shortest I could find It looks like 12" it was a monster brand one. Supply is rated at 1.5A and all devices attached to raspberry pi have hubs (they are usb3...) and are powered off a 12V supply (they have an internal switch mode circuit) and will deliver 1.5A per port and only 4 ports so no internal hub on hub action. It shouldn't be a lack of power. Then again I can't be sure it isn't somehow related to power. It could be USB 3 related but I have used these hubs since I started getting asics. My only asics are BFL and don't seem to draw power from the USB. I am expecting this will be like it used to be stable 24/7 for months at a time. Thank You for all that you do. The other thing to consider is Arch rather than Raspbian. The Arch USB network driver seems to be more reliable than the Raspbian one. As mentioned before, the network device on the RPi is actually USB. The basics of RPi Arch setup: http://www.kano-kun.net/?p=87I apparently have the wrong version of arch. Well not version as much as wrong image dated file. Mine was 2013-7-22. The partition is kinda messing me up. I know I can't partition as you did with 3 partitions. [root@alarmpi ~]# fdisk -ls
Disk /dev/mmcblk0: 15.9 GB, 15931539456 bytes, 31116288 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk label type: dos
Disk identifier: 0x00057540
Device Boot Start End Blocks Id System
/dev/mmcblk0p1 2048 186367 92160 c W95 FAT32 (LBA)
/dev/mmcblk0p2 186368 3667967 1740800 5 Extended
/dev/mmcblk0p5 188416 3667967 1739776 83 Linux
[root@alarmpi ~]#
I think but can't be sure that /dev/mmcblk0p5 is the one I want to delete and recreate. I am a little confused by the extended and Linux partitions using the same space. It looks wrong to me to have two partitions sharing space. |
|
|
|
kano
Legendary
Offline
Activity: 4466
Merit: 1798
Linux since 1997 RedHat 4
|
|
October 22, 2013, 10:44:28 PM |
|
I apparently have the wrong version of arch. Well not version as much as wrong image dated file. Mine was 2013-7-22. The partition is kinda messing me up. I know I can't partition as you did with 3 partitions. [root@alarmpi ~]# fdisk -ls
Disk /dev/mmcblk0: 15.9 GB, 15931539456 bytes, 31116288 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk label type: dos
Disk identifier: 0x00057540
Device Boot Start End Blocks Id System
/dev/mmcblk0p1 2048 186367 92160 c W95 FAT32 (LBA)
/dev/mmcblk0p2 186368 3667967 1740800 5 Extended
/dev/mmcblk0p5 188416 3667967 1739776 83 Linux
[root@alarmpi ~]#
I think but can't be sure that /dev/mmcblk0p5 is the one I want to delete and recreate. I am a little confused by the extended and Linux partitions using the same space. It looks wrong to me to have two partitions sharing space. Yeah I never updated that did I  I actually have one of these sitting on an SD card on an RPi next to me (switched off) that I've not got around to resizing. p5 is inside p2 Let me think ... I'm pretty sure it is: so you need to remember both 186368 and 188416 delete them both create a new extended p2 from 186368 to the end then create a new Linux partition p5 from 188416 to the end then continue as before. Edit: no I should say, resize2fs /dev/mmcblk0p5 Let me know how it goes then I'll add it to the blog. |
|
|
|
-ck (OP)
Legendary
Offline
Activity: 4088
Merit: 1630
Ruu \o/
|
|
October 22, 2013, 11:01:45 PM |
|
To the various windows victims valued testers out there, I have uploaded new experimental binaries that now use windows' own low level clocks and timers to support higher resolution timeouts and hopefully be more reliable than using mingw's wrappers. I also found a bug in the windows code while working on these. For people with timeout issues on windows, please grab one of the exes from this directory and put it into your current cgminer directory: http://ck.kolivas.org/apps/cgminer/temp/ |
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel
2% Fee Solo mining at solo.ckpool.org
-ck
|
|
|
|
techman05
|
|
October 22, 2013, 11:37:17 PM
Last edit: October 23, 2013, 12:25:56 AM by techman05 |
|
Hey short question if you get a chance. Did a fresh install of cgminer on a raspberry pi and its not noticing my devices. As far as I know I did the enable icarus command.
Is there a way to refresh and enable without a clean install? Just to make sure I ran the enable right. I didn't get any errors when I ran the command only one I got was that there were no gpus and that was during the autogen.sh
Thanks
...edit...
Got lucky I don't think I fudged in my cleaning. I did a make clean , set the configure file again ,make clean install and I'm good now.
Somehow it didn't stick but now its going till the pi decides otherwise. Last time I tried this I got about 4 hrs till it froze and stopped mining (not your fault the pi is weird)
.....edit2....
crap did 1 minute and 10 second. A new record [for me].
|
|
|
|
|
techman05
|
|
October 23, 2013, 12:10:59 AM |
|
To the various windows victims valued testers out there, I have uploaded new experimental binaries that now use windows' own low level clocks and timers to support higher resolution timeouts and hopefully be more reliable than using mingw's wrappers. I also found a bug in the windows code while working on these. For people with timeout issues on windows, please grab one of the exes from this directory and put it into your current cgminer directory: http://ck.kolivas.org/apps/cgminer/temp/My only crap out is when I unplug my ethernet plug and it take 2hrs to kick back in to see my wireless as another option (either that or cgminer got lonely). As soon as I walked out for errans it stopped working and started as soon as I came back. Your victims thank you for cgminer being a cuddle bug. |
|
|
|
|
xzempt
|
|
October 23, 2013, 12:31:56 AM |
|
 why doesnt cgminer support asicminer blades?
|
|
|
|
|
|
techman05
|
|
October 23, 2013, 12:35:50 AM |
|
Did you compile and enable it if your using it on linux?  |
|
|
|
-ck (OP)
Legendary
Offline
Activity: 4088
Merit: 1630
Ruu \o/
|
|
October 23, 2013, 12:37:54 AM |
|
why doesnt cgminer support asicminer blades? Blades have their own inbuilt ancient mining software and there is no other way to mine on them. All they do is support getwork protocol to a pool/server. If you're talking about cgminer not pretending to be pool software supporting an ancient outdated mining protocol, you're correct, it doesn't. |
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel
2% Fee Solo mining at solo.ckpool.org
-ck
|
|
|
aigeezer
Legendary
Offline
Activity: 1450
Merit: 1013
Cryptanalyst castrated by his government, 1952
|
|
October 23, 2013, 12:52:35 AM |
|
To the various windows victims valued testers out there, I have uploaded new experimental binaries that now use windows' own low level clocks and timers to support higher resolution timeouts and hopefully be more reliable than using mingw's wrappers. I also found a bug in the windows code while working on these. For people with timeout issues on windows, please grab one of the exes from this directory and put it into your current cgminer directory: http://ck.kolivas.org/apps/cgminer/temp/Trying it now. No problems in first two minutes. Thanks for working on it! |
|
|
|
|
|
techman05
|
|
October 23, 2013, 01:20:00 AM |
|
what does higher resolution timeouts mean? Does that mean that cgminer has more time to work itself out if something goes wrong with a device hardware or in cgminer software?
Thanks
|
|
|
|
-ck (OP)
Legendary
Offline
Activity: 4088
Merit: 1630
Ruu \o/
|
|
October 23, 2013, 01:32:07 AM |
|
what does higher resolution timeouts mean? Does that mean that cgminer has more time to work itself out if something goes wrong with a device hardware or in cgminer software?
Thanks
Whenever we wait for something in code we tell it to sleep. By default in windows, the sleep time can be accurate to only +/-15ms best case scenario (and is often worse). This may or may not be relevant depending on how critical the timing is on the device. With the experimental changes I've been able to bring it down to +/-0.1μs. Also because there are less layers of code between cgminer and the operating system now, there are less potential points of failure in the code. |
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel
2% Fee Solo mining at solo.ckpool.org
-ck
|
|
|
|
Askit2
|
|
October 23, 2013, 01:57:22 AM |
|
I apparently have the wrong version of arch. Well not version as much as wrong image dated file. Mine was 2013-7-22. The partition is kinda messing me up. I know I can't partition as you did with 3 partitions. [root@alarmpi ~]# fdisk -ls
Disk /dev/mmcblk0: 15.9 GB, 15931539456 bytes, 31116288 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk label type: dos
Disk identifier: 0x00057540
Device Boot Start End Blocks Id System
/dev/mmcblk0p1 2048 186367 92160 c W95 FAT32 (LBA)
/dev/mmcblk0p2 186368 3667967 1740800 5 Extended
/dev/mmcblk0p5 188416 3667967 1739776 83 Linux
[root@alarmpi ~]#
I think but can't be sure that /dev/mmcblk0p5 is the one I want to delete and recreate. I am a little confused by the extended and Linux partitions using the same space. It looks wrong to me to have two partitions sharing space. Yeah I never updated that did I I actually have one of these sitting on an SD card on an RPi next to me (switched off) that I've not got around to resizing. p5 is inside p2 Let me think ... I'm pretty sure it is: so you need to remember both 186368 and 188416 delete them both create a new extended p2 from 186368 to the end then create a new Linux partition p5 from 188416 to the end then continue as before. Edit: no I should say, resize2fs /dev/mmcblk0p5 Let me know how it goes then I'll add it to the blog. It worked great. I didn't go to the end. I wanted to make a swap partition but it is incomplete. Both deleted, Recreated one at a time, extended to the calculated last sector then the Linux partition to the end of the extended partition. Restart, no FS crash, and resize2fs /dev/mmcblk0p5 seems to have worked. Shows up as 6% used on the root partition. One small thing. I had raspbian some semblance of running and I could use all 17 devices I have. Granted it lasted 1 hour to 3 minutes between crashes. On arch using all the same hardware I can only detect 13. The other 4 have been put on my dvr for now. I have hub A attatched to the lower port on raspberry pi, Hub B on hub A port 1, Hub C on hub A port 3 one port doesn't work here. Hub D is on the upper port of the raspberry pi, Hub E was hooked up to port 1 on Hub D. I have a keyboard also hooked to hub D. Everything hooked up to hub D works except if I add either the last hub or another jalapeno by itself it will not detect. Nothing comes up, no errors, no found devices, nothing. I am curious does arch have less USB devices it can handle? I don't mind as I plan on running 8 per pi but I have to buy a couple more first. So far arch seems much more stable. Less GUI but much more stable. |
|
|
|
|
