OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.1

[phillipsjk]

Update: I have decided to try bfgminer instead. If it fails to work, I may install cgminer on a nearby Gnu/Linux machine and run a USB extension cable.

I was curious if anybody has been successful in compiling this for FreeBSD.
[spoiler]

Code:

$ ./configure --enable-blockerupter
checking build system type... amd64-unknown-freebsd9.2
checking host system type... amd64-unknown-freebsd9.2
checking target system type... amd64-unknown-freebsd9.2
checking for a BSD-compatible install... /usr/bin/install -c
checking whether build environment is sane... yes
checking for a thread-safe mkdir -p... ./install-sh -c -d
checking for gawk... no
checking for mawk... no
checking for nawk... nawk
checking whether make sets $(MAKE)... yes
checking whether make supports nested variables... yes
checking whether make supports nested variables... (cached) yes
checking for style of include used by make... GNU
checking for gcc... gcc
checking whether the C compiler works... yes
checking for C compiler default output file name... a.out
checking for suffix of executables... 
checking whether we are cross compiling... no
checking for suffix of object files... o
...
checking for windres... no
checking for inline... inline
checking operating system... configure: error: unsupported operating system
configure: error: ./configure failed for compat/libusb-1.0

It says the OS is unsupported. Not sure what kind of porting would be involved.

Code:

$ ./configure --enable-blockerupter --with-system-libusb
checking build system type... amd64-unknown-freebsd9.2
checking host system type... amd64-unknown-freebsd9.2
checking target system type... amd64-unknown-freebsd9.2
checking for a BSD-compatible install... /usr/bin/install -c
checking whether build environment is sane... yes
checking for a thread-safe mkdir -p... ./install-sh -c -d
checking for gawk... no
checking for mawk... no
checking for nawk... nawk
checking whether make sets $(MAKE)... yes
checking whether make supports nested variables... yes
checking whether make supports nested variables... (cached) yes
checking for style of include used by make... GNU
checking for gcc... gcc
checking whether the C compiler works... yes
checking for C compiler default output file name... a.out
checking for suffix of executables... 
checking whether we are cross compiling... no
...
checking for gcc __sync builtins... yes
checking for gcc __atomic builtins... no
checking that generated files are newer than configure... done
configure: creating ./config.status
config.status: creating jansson.pc
config.status: creating Makefile
config.status: creating src/Makefile
config.status: creating src/jansson_config.h
config.status: creating config.h
config.status: config.h is unchanged
config.status: executing depfiles commands
config.status: executing libtool commands



------------------------------------------------------------------------
cgminer 4.7.1
------------------------------------------------------------------------



Configuration Options Summary:

  libcurl(GBT+getwork).: Enabled: -L/usr/local/lib -lcurl  
  curses.TUI...........: FOUND: -lncurses

  Antminer.S1.Bitmain..: Disabled
  Antminer.S2.Bitmain..: Disabled
  Avalon.ASICs.........: Disabled
  Avalon2.ASICs........: Disabled
  BlackArrowMinion.ASIC: Disabled
  BlackArrow.ASICs.....: Disabled
  BFL.ASICs............: Disabled
  BitForce.FPGAs.......: Disabled
  BitFury.ASICs........: Disabled
  BlockErupter.ASICs...: Enabled
  Cointerra.ASICs......: Disabled
  Spond-sp10.ASICs.....: Disabled
  Spond-sp30.ASICs.....: Disabled
  Bitmine-A1.ASICs.....: Disabled
  Drillbit.BitFury.....: Disabled
  Hashfast.ASICs.......: Disabled
  Hashratio.ASICs......: Disabled
  Icarus.ASICs/FPGAs...: Disabled
  Klondike.ASICs.......: Disabled
  KnC.ASICs............: Disabled
  ModMiner.FPGAs.......: Disabled

Compilation............: make (or gmake)
  CPPFLAGS.............: 
  CFLAGS...............: -g -O2
  LDFLAGS..............:  -lpthread
  LDADD................: -ldl -L/usr/local/lib -lcurl   compat/jansson-2.6/src/.libs/libjansson.a -lpthread    -lm -lusb -lrt

Installation...........: make install (as root if needed, with 'su' or 'sudo')
  prefix...............: /usr/local

$ make
make  all-recursive
Making all in lib
  GEN      arg-nonnull.h
  GEN      c++defs.h
  GEN      warn-on-use.h
  GEN      signal.h
  GEN      stdint.h
  GEN      string.h
make  all-recursive
  CC       dummy.o
  CC       memmem.o
  AR       libgnu.a
Making all in compat
Making all in jansson-2.6
"Makefile", line 860: Need an operator
make: fatal errors encountered -- cannot continue
*** [all-recursive] Error code 1

Stop in /usr/home/P2Pool/cgminer/compat.
*** [all-recursive] Error code 1

Stop in /usr/home/P2Pool/cgminer.
*** [all] Error code 1

Stop in /usr/home/P2Pool/cgminer.

If the full configure output is required, I can redirect it to a file.

Edit: forgot that "make clean" also fails:

Code:

$ make clean
Making clean in lib
test -z "" || rm -f 
rm -rf .libs _libs
test -z "libgnu.a" || rm -f libgnu.a
test -z "" || rm -f 
rm -f *.o
test -z "core *.stackdump arg-nonnull.h arg-nonnull.h-t  c++defs.h c++defs.h-t signal.h signal.h-t stddef.h stddef.h-t  stdint.h stt
rm -f *.lo
Making clean in compat
Making clean in jansson-2.6
"Makefile", line 860: Need an operator
make: fatal errors encountered -- cannot continue
*** [clean-recursive] Error code 1

Stop in /usr/home/P2Pool/cgminer/compat.
*** [clean-recursive] Error code 1

Stop in /usr/home/P2Pool/cgminer.

Edit2: dug up this advice...testing (rerunning autogen.sh)

It's been a lot of years since I ran FreeBSD, but as I recall, the default 'make' command they use is BSD-specific and not compatible in all ways with the Gnu make that comes with GCC, so you typically have to tell it to use gmake explicitly when using some Gnu make makefiles.  I think if you type 'MAKE=gmake; gmake' it will probably work.

gmake failed as well. (./configure still failed with the same error as above without system-usb)

Code:

gmake
gmake  all-recursive
gmake[1]: Entering directory `/usr/home/P2Pool/cgminer'
Making all in lib
gmake[2]: Entering directory `/usr/home/P2Pool/cgminer/lib'
  GEN      arg-nonnull.h
  GEN      c++defs.h
  GEN      warn-on-use.h
  GEN      signal.h
  GEN      stdint.h
  GEN      string.h
gmake  all-recursive
gmake[3]: Entering directory `/usr/home/P2Pool/cgminer/lib'
gmake[4]: Entering directory `/usr/home/P2Pool/cgminer/lib'
  CC       dummy.o
  CC       memmem.o
  AR       libgnu.a
gmake[4]: Leaving directory `/usr/home/P2Pool/cgminer/lib'
gmake[3]: Leaving directory `/usr/home/P2Pool/cgminer/lib'
gmake[2]: Leaving directory `/usr/home/P2Pool/cgminer/lib'
Making all in compat
gmake[2]: Entering directory `/usr/home/P2Pool/cgminer/compat'
Making all in jansson-2.6
gmake[3]: Entering directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6'
gmake  all-recursive
gmake[4]: Entering directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6'
Making all in src
gmake[5]: Entering directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6/src'
  CC       dump.lo
  CC       error.lo
  CC       hashtable.lo
  CC       load.lo
  CC       memory.lo
  CC       pack_unpack.lo
  CC       strbuffer.lo
  CC       strconv.lo
  CC       value.lo
  CCLD     libjansson.la
gmake[5]: Leaving directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6/src'
gmake[5]: Entering directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6'
gmake[5]: Leaving directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6'
gmake[4]: Leaving directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6'
gmake[3]: Leaving directory `/usr/home/P2Pool/cgminer/compat/jansson-2.6'
gmake[3]: Entering directory `/usr/home/P2Pool/cgminer/compat'
gmake[3]: Nothing to be done for `all-am'.
gmake[3]: Leaving directory `/usr/home/P2Pool/cgminer/compat'
gmake[2]: Leaving directory `/usr/home/P2Pool/cgminer/compat'
Making all in ccan
gmake[2]: Entering directory `/usr/home/P2Pool/cgminer/ccan'
  CC       opt/libccan_a-helpers.o
  CC       opt/libccan_a-opt.o
  CC       opt/libccan_a-parse.o
  CC       opt/libccan_a-usage.o
  AR       libccan.a
gmake[2]: Leaving directory `/usr/home/P2Pool/cgminer/ccan'
gmake[2]: Entering directory `/usr/home/P2Pool/cgminer'
  CC       cgminer-cgminer.o
In file included from ./sha2.h:36,
                 from cgminer.c:52:
./miner.h:50:1: warning: "alloca" redefined
In file included from cgminer.c:19:
/usr/include/stdlib.h:246:1: warning: this is the location of the previous definition
gcc: -lpthread: linker input file unused because linking not done
  CC       cgminer-util.o
In file included from util.c:42:
miner.h:50:1: warning: "alloca" redefined
In file included from util.c:14:
/usr/include/stdlib.h:246:1: warning: this is the location of the previous definition
gcc: -lpthread: linker input file unused because linking not done
  CC       cgminer-sha2.o
In file included from sha2.h:36,
                 from sha2.c:37:
miner.h:50:1: warning: "alloca" redefined
In file included from ./compat/jansson-2.6/src/jansson.h:12,
                 from miner.h:10,
                 from sha2.h:36,
                 from sha2.c:37:
/usr/include/stdlib.h:246:1: warning: this is the location of the previous definition
gcc: -lpthread: linker input file unused because linking not done
  CC       cgminer-api.o
In file included from api.c:25:
miner.h:50:1: warning: "alloca" redefined
In file included from api.c:16:
/usr/include/stdlib.h:246:1: warning: this is the location of the previous definition
api.c: In function 'mcast':
api.c:4717: error: 'IPV6_ADD_MEMBERSHIP' undeclared (first use in this function)
api.c:4717: error: (Each undeclared identifier is reported only once
api.c:4717: error: for each function it appears in.)
gmake[2]: *** [cgminer-api.o] Error 1
gmake[2]: Leaving directory `/usr/home/P2Pool/cgminer'
gmake[1]: *** [all-recursive] Error 1
gmake[1]: Leaving directory `/usr/home/P2Pool/cgminer'
gmake: *** [all] Error 2

Edit: More hints...

Hi,

I try to delete -ldl switch to Makefile.
But now i have an other problem during compilation on the same object:

Code:

SNIP!

you have an idea ?

It looks like compiling cgminer on FreeBSD is not going to be a trivial task.  I would be glad to help you through it, but I haven't had a FreeBSD box in some years, so I really can't.  I think your options are these:

1) There may actually be a FreeBSD port for CGMiner that already exists.  If so, that is clearly the best way to build it.

2) If not, download the Linux binary and run it.  The FreeBSD Linux emulation is very good, and I'd be surprised if it didn't work just fine.

3) Get onto the FreeBSD ports mailing lists/forums and ask around.  I would be surprised if no FreeBSD users are using cgminer, and some of them can probably help you get it compiled.  Maybe you can inspire someone to create a port and add it to the ports tree.

Point 1 appears to be a "no". There is a cpuminer in /usr/ports/net-p2p/cpuminer
find was not accepting wildcards, so did not check other categories. (The ports list on the website appears
to miss some: like bitcoin-daemon.)

I suppose I should try point 2.
Edit3:

Code:

$ ./cgminer -o 192.168.243.2:9332
ELF binary type "0" not known.
./cgminer: Exec format error

--looks like I actually need to install Linux binary compatibility.
--Appears to be fedora 10
Tried without installing Linux base-system (based on these instructions).

Code:

root@casey:~ # kldload linux
root@casey:~ # kldstat
Id Refs Address            Size     Name
 1   11 0xffffffff80200000 15b9440  kernel
 2    1 0xffffffff81812000 10d97    geom_eli.ko
 3    1 0xffffffff81823000 1efc1    crypto.ko
 4    1 0xffffffff81842000 57bcf    linux.ko
root@casey:~ # su P2Pool
$ cd
$ cd cgminer-4.7.1-x86_64-built
$ ./cgminer -o 192.168.243.2:9332 
ELF binary type "0" not known.
./cgminer: Exec format error
$ brandelf -t Linux cgminer 
$ ./cgminer -o 192.168.243.2:9332
ELF binary type "3" not known.
./cgminer: Exec format error

Did not appear to work even installing the fedora package:

Code:

root@casey:~ # pkg install linux_base-f10-10_7
Updating FreeBSD repository catalogue...
FreeBSD repository is up-to-date.
All repositories are up-to-date.
The following 1 packages will be affected (of 0 checked):

New packages to be INSTALLED:
        linux_base-f10: 10_7

The process will require 134 MB more space.
16 MB to be downloaded.

Proceed with this action? [y/N]: y
Fetching linux_base-f10-10_7.txz: 100%   16 MB   2.9M/s    00:06    
Checking integrity... done (0 conflicting)
[1/1] Installing linux_base-f10-10_7: 100%

+++ Some programs may need linprocfs, please add it to /etc/fstab! +++

Running linux ldconfig...
root@casey:~ # su P2Pool
$ cd
$ cd cgminer-4.7.1-x86_64-built
$ ./cgminer -o 192.168.243.2:9332
ELF binary type "3" not known.
./cgminer: Exec format error
$ brandelf -t Linux cgminer 
$ ./cgminer -o 192.168.243.2:9332
ELF binary type "3" not known.
./cgminer: Exec format error

Edit4: does the 2.9.1 changelog mean anything?

- Changes to build prototypes to support building on FreeBSD 9.1-RC2 amd64

YAEdit: I wonder if the Linux binary compatibility only works for 32 bit binaries.

Base set of packages needed in Linux mode for i386/amd64 (Linux Fedora 10)

..so should work with amd64
[/spoiler]

[1714932580]

1714932580

[1714932580]

1714932580

[-ck]

Edit4: does the 2.9.1 changelog mean anything?

- Changes to build prototypes to support building on FreeBSD 9.1-RC2 amd64

That should be the most telling quote for you, no? It's the last time anyone pushed any changes for 'bsd support into cgminer and it has since moved further and further away from being code compatible, so unless you or someone else is interested in sitting down and doing patches for it as some kind of (large) project, it has no chance of compiling. I'm not even sure of the status of freebsd support in libusb which is pivotal to running most modern mining equipment.

[-ck]

Okay here goes. Even though I don't have the U3 yet, bitmainUSA let me work on one remotely in the interim so I've merged my own U3 driver into the main cgminer git master and it is now available for testing.

Note the following changes:
You do not need to specify ANY command line parameters to run the U3 (no weird options, timeouts or frequency or voltage required) for it to run at default voltage and frequency.
The default frequency is set to 225Mhz
The default voltage is set to 750mV

In addition, the following parameters are recognised, though the defaults are obviously safest.

Code:

--au3-freq <arg>    Set AntminerU3 frequency in MHz, range 100-250 (default: 225.0)
--au3-volt <arg>    Set AntminerU3 voltage in mv, range 725-850, 0 to not set (default: 750)

In addition, I have a test windows binary:
http://ck.kolivas.org/apps/cgminer/temp/cgminer.exe

Of course this is all new code but so far it's looking stable, but I'm unable to test the windows binary myself.

[GrrOucH]

It does work. Picks up the U3 just fine. I think I have a bad us though. With the defaults you have it will only hash at about 43gh. I have not tried to play with the volt/freq settings because I would like to think the defaults should wrk. Thanks for getting it working without all those nasty settings.

[-ck]

It does work. Picks up the U3 just fine. I think I have a bad us though. With the defaults you have it will only hash at about 43gh. I have not tried to play with the volt/freq settings because I would like to think the defaults should wrk. Thanks for getting it working without all those nasty settings.

Try the latest git or binary if you haven't updated recently. The last update was only about 12 hours ago.

[GrrOucH]

what is the --enable flag for the u3?

[-ck]

what is the --enable flag for the u3?

--enable-icarus
though you only use that when building your own binary...

[-ck]

New release: Version 4.8.0, 25th November 2014

New driver and driver updates so a stable update despite the version bump.


Human readable changelog:

- Fix for autogen.sh on OSX not liking -f option with readlink.
- Added an --enable-forcecombo option to allow building of driver combinations with drivers that are normally built standalone.
- Avalon2 updates to try to detect failing devices and reset them if possible.
- Avalon2 will no longer loop forever on a failed usb write.
- Displays for devices with a number in their name will no longer be confused by having a space before their id (eg: AV2 1 instead of AV21)
- Fixes for understanding share errors from pools that have a non-standard stratum response (i.e. Ghash) which would otherwise have obscured rejects that were occuring for duplicates, etc.
- Device driver for Antminer U3s:
-- These require the --enable-icarus option when building a binary and appear as "AU3"
-- Unlike the bitmain driver, these need no options to work, having sane defaults and accepting the extra options for changing frequency and voltage with the following new options:

Code:

--au3-freq <arg>    Set AntminerU3 frequency in MHz, range 100-250 (default: 225.0)
--au3-volt <arg>    Set AntminerU3 voltage in mv, range 725-850, 0 to not set (default: 750)

NOTE: These have a tendency of dying if you leave them running at high voltages either at the PSU and/or the device so NOT recommended!
- The display will show the voltage and frequency configured for antminer U3 and the frequency for U1/2.
- Numerous improvements to tell apart Icarus devices more reliably (AMU, ANU, AU3, LIN, LIR).
- Fix for unprefixed IPv6 entries in the api allow command.
- Fix for misspelt temperature in spondoolies driver.
- Rockminer driver updates for better support of New R-Box and RK Box:
-- Filter out the bulk of the duplicate shares. This will make the hashrate appear lower, but no pool would have been honouring them as valid shares anyway.
-- Substantially lower hardware errors.
-- Lower CPU usage.
-- Slightly higher effective hashrate.

Full changelog:

- Allow forcing of building driver combinations with --enable-forcecombo
- Put spaces between name and id in avalon2 and icarus
- Relax detection of a failing ava2 to more than 1 minute and perform the test
after polling for results
- Cap maximum diff on ava2 in order to still get shares
- Put space between device name and id to prevent device names with numbers in
them confusing the display
- USB write errors are always fatal so they should be treated as such on ava2
- Issue a usb reset for ava2 that is not returning valid shares and then drop it
if it persists for over a minute
- Process share results without a result value
- Damp out hashrate displayed for antminer USBs
- Add voltage and speed where relevant to antminer USBs
- Don't estimate time on any antminer usb during a timeout
- Return icarus nonce ok only when the nonce size matches the device or more
- Don't discard old workids until we cycle back to them on antusb and look for
more nonces in the buffer
- Adjust ant usb timing for queued work
- Use a cyclical list for the ant queued work
- Mask and limit workid for antusb and dont clear buffer
- Check the nonce on the worked item, not the submitted work
- Skip over unfinished work that we can't free in ant usb
- Use a workid and array if possible for the small ant usb work queue
- Create an array for antworks for antminer usb devices
- On U3 calculate hashrate purely on shares, not timeouts
- Add switches for AU3
- Adjust icarus wait timeout according to device
- Differentiate U3 from U1/2 as a separate driver with different parameters and
adjust timing accordingly
- Skip ANUs detected in rock detect
- Try U3 after trying other icarus options
- Add rudimentary ANU voltage setting support for U3
- Fix ignoring unprefixed v6 address in api allow list
- Fix minor typos in Spondoolies SP10 and SP30 drivers
- Implement a basic rock_flush function to discard the base work we are rolling
work from.
- Task_no for rockminer from the nonce bin should simply be masked
- Change rbox default correction times to 5 in a revised frequency order
- Change default frequency on T1 to 330
- Reinstate last received check and resend in rockminer, being more lenient at 2
seconds to allow for dither errors at 1
- Roll work for the rbox when possible

[hurricandave]

Last week you posted a link on the Rockminer thread for an experimental windows binary. When I tried it out, I found that my 6 units(New R-Box) were about 2-3 C warmer than when using the Rockminer fork. Over this past weekend I experimented with running one of my New R-Box on a RasPi.  Since this gave me an opportunity to view the Rockminer browser interface for the first time I noticed it included a Fan speed selector, Full or Auto. It then occurred to me that when using the Rockminer Fork the <arg->  --rmu-auto 320, in this case the "auto", may have been pointed at the Fan speed? Just guessing really. BUT. If this is the case, should your main binary include this fan declaration as well? Maybe its already built in and I just haven't realized what the proper way/place to invoke it would be? So far on 4.8.0 I'm up only 1-2 C per unit over normal but its hard to tell if the fans are running as hard as they usually do. Because I live 50 feet from a busy highway and the drone from outside can make it hard to discern small frequency changes like fan RPM.

[Perlover]

Miner: HashCoins Apollo
I regulary get a segmentation fault
I tryied with included into cgminer 'libusb' and with system libusb - same result
This binary was compiled as ' ./configure --enable-hashfast --with-system-libusb'
I turned on a core dumping and i got:

Code:

pi@raspberrypi ~ $ gdb -c ./core.26546 /usr/local/bin/cgminer
GNU gdb (GDB) 7.4.1-debian
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-linux-gnueabihf".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/local/bin/cgminer...done.
[New LWP 26599]
[New LWP 26548]
[New LWP 26559]
[New LWP 26560]
[New LWP 26561]
[New LWP 26562]
[New LWP 26558]
[New LWP 26557]
[New LWP 26555]
[New LWP 26567]
[New LWP 26556]
[New LWP 26563]
[New LWP 26564]
[New LWP 26546]
[New LWP 26597]
[New LWP 26598]
[New LWP 26566]
[New LWP 26547]
[New LWP 26600]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/arm-linux-gnueabihf/libthread_db.so.1".
Core was generated by `cgminer -c cgminer.conf --hfa-hash-clock 700'.
Program terminated with signal 11, Segmentation fault.
#0  0x00041b58 in usb_perform_transfer (cgpu=0x20193d8, usbdev=<optimized out>, intinfo=<optimized out>, epinfo=<optimized out>, data=0x0, length=0, transferred=0x0, timeout=0, mode=0,
    cmd=C_REJECTED, seq=0, cancellable=false, tt=false) at usbutils.c:3061
3061            else if ((endpoint & LIBUSB_ENDPOINT_DIR_MASK) == LIBUSB_ENDPOINT_IN && *transferred)
(gdb) bt
#0  0x00041b58 in usb_perform_transfer (cgpu=0x20193d8, usbdev=<optimized out>, intinfo=<optimized out>, epinfo=<optimized out>, data=0x0, length=0, transferred=0x0, timeout=0, mode=0,
    cmd=C_REJECTED, seq=0, cancellable=false, tt=false) at usbutils.c:3061
#1  0x00000000 in ?? ()
(gdb)

P.S. I looked sources and i think that problem may be in my hardware (memory for example)... The static function 'usb_perform_transfer' cannot be called with zero-valued parameters :-/

[-ck]

Miner: HashCoins Apollo

I don't know what a hashcoins apollo is nor what driver it uses. As far as I'm aware there is no official support for it in mainline cgminer so you're likely using a fork made by them which means you have to ask the manufacturer for support for their fork.

[Taugeran]

Miner: HashCoins Apollo

I don't know what a hashcoins apollo is nor what driver it uses. As far as I'm aware there is no official support for it in mainline cgminer so you're likely using a fork made by them which means you have to ask the manufacturer for support for their fork.

They're another hashfast 3rd party mfr

[-ck]

They're another hashfast 3rd party mfr

Thanks I wasn't aware.

P.S. I looked sources and i think that problem may be in my hardware (memory for example)... The static function 'usb_perform_transfer' cannot be called with zero-valued parameters :-/

Well yeah there should be no reason for the code to be trying to request a usb transfer of nothing.

[CCW]

Did anybody get the latest cgminer working with the antminer U3 (on raspberry) to run longer than 1 hour?

after max. an hour the cgminer stops working (auto exit) or the miner becomes zombie

My console is also displaying a lot of the LIBUSB_ERROR_IO. Any ideas?

I used the latest version of cgminer with the default configurations

[-ck]

Did anybody get the latest cgminer working with the antminer U3 (on raspberry) to run longer than 1 hour?

after max. an hour the cgminer stops working (auto exit) or the miner becomes zombie

My console is also displaying a lot of the LIBUSB_ERROR_IO. Any ideas?

I used the latest version of cgminer with the default configurations

This is just indicative of the crap hardware that the U3 really is I'm afraid. Keep turning the frequency down and you might get lucky and find a stable frequency to run it at. Make sure to use a high quality shielded USB cable (I found that helped their stability compared to the default cable too). Don't overvolt them it will destroy the PSU. Don't put them on a usb3 port, don't use a powered hub, don't put too many other devices on the same hub port. Don't breathe on them, and make sure they're facing west...

[Taugeran]

Lol

[Perlover]

Hi,

Other device Raspberry Pi B+ (i bought other device and other micro sd card)
I installed new OS for new RasPi
The miner is same
And again segmentation fault in usb_perform_transfer
May be some compile options are not fine for gcc?
I think a hardware is not problem now...

Code:

pi@raspberrypi ~ $ gdb -c core.3179 /usr/local/bin/cgminer
GNU gdb (GDB) 7.4.1-debian
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-linux-gnueabihf".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/local/bin/cgminer...done.
[New LWP 3203]
[New LWP 3179]
[New LWP 3181]
[New LWP 3188]
[New LWP 3189]
[New LWP 3190]
[New LWP 3191]
[New LWP 3192]
[New LWP 3193]
[New LWP 3194]
[New LWP 3195]
[New LWP 3196]
[New LWP 3199]
[New LWP 3200]
[New LWP 3180]
[New LWP 3206]
[New LWP 3182]
[New LWP 3205]
[New LWP 3197]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/arm-linux-gnueabihf/libthread_db.so.1".
Core was generated by `cgminer -c cgminer.conf --hfa-hash-clock 700'.
Program terminated with signal 11, Segmentation fault.
#0  __list_del (next=0x86aa, prev=0xad5fc380) at elist.h:78
78		next->prev = prev;
(gdb) bt
#0  __list_del (next=0x86aa, prev=0xad5fc380) at elist.h:78
#1  list_del (entry=0xaf5fc380) at elist.h:89
#2  complete_usb_transfer (ut=0xaf5fc368) at usbutils.c:2864
#3  0x00041af8 in usb_perform_transfer (cgpu=0x1e8970, usbdev=<optimized out>, intinfo=<optimized out>, epinfo=<optimized out>, data=0x0, length=0, 
    transferred=0x0, timeout=0, mode=0, cmd=C_REJECTED, seq=0, cancellable=false, tt=false) at usbutils.c:3028
#4  0x00000000 in ?? ()

[-ck]

Hi,

Other device Raspberry Pi B+ (i bought other device and other micro sd card)
I installed new OS for new RasPi
The miner is same
And again segmentation fault in usb_perform_transfer
May be some compile options are not fine for gcc?
I think a hardware is not problem now...

Code:

pi@raspberrypi ~ $ gdb -c core.3179 /usr/local/bin/cgminer
GNU gdb (GDB) 7.4.1-debian
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-linux-gnueabihf".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/local/bin/cgminer...done.
[New LWP 3203]
[New LWP 3179]
[New LWP 3181]
[New LWP 3188]
[New LWP 3189]
[New LWP 3190]
[New LWP 3191]
[New LWP 3192]
[New LWP 3193]
[New LWP 3194]
[New LWP 3195]
[New LWP 3196]
[New LWP 3199]
[New LWP 3200]
[New LWP 3180]
[New LWP 3206]
[New LWP 3182]
[New LWP 3205]
[New LWP 3197]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/arm-linux-gnueabihf/libthread_db.so.1".
Core was generated by `cgminer -c cgminer.conf --hfa-hash-clock 700'.
Program terminated with signal 11, Segmentation fault.
#0  __list_del (next=0x86aa, prev=0xad5fc380) at elist.h:78
78		next->prev = prev;
(gdb) bt
#0  __list_del (next=0x86aa, prev=0xad5fc380) at elist.h:78
#1  list_del (entry=0xaf5fc380) at elist.h:89
#2  complete_usb_transfer (ut=0xaf5fc368) at usbutils.c:2864
#3  0x00041af8 in usb_perform_transfer (cgpu=0x1e8970, usbdev=<optimized out>, intinfo=<optimized out>, epinfo=<optimized out>, data=0x0, length=0, 
    transferred=0x0, timeout=0, mode=0, cmd=C_REJECTED, seq=0, cancellable=false, tt=false) at usbutils.c:3028
#4  0x00000000 in ?? ()

Try this one line change in usbutils.c line 893:
Change:

Code:

#define DO_USB_STATS 1

to:

Code:

#define DO_USB_STATS 0

[vayvanne]

[2014-11-29 01:45:03] Started cgminer 4.8.0
[2014-11-29 01:45:03] Loaded configuration file cgminer.conf
[2014-11-29 01:45:07] ANU 0: Found at 3:1
[2014-11-29 01:45:08] Probing for an alive pool
[2014-11-29 01:45:08] Testing pool http://localhost:8332
[2014-11-29 01:45:10] HTTP request failed: Failed connect to localhost:8332; No error
[2014-11-29 01:45:11] HTTP request failed: Failed connect to localhost:8332; No error
[2014-11-29 01:45:13] Failed to connect to stratum on localhost:8332
[2014-11-29 01:45:18] Waiting for work to be available from pools.

The above is while the bitcoind is still running DB check. After 5-10 sec of waiting it clears the screen and just exits after any key pressed. The subject is cgminer 4.8.0-windows. One more issue, it takes 250MHz as default frequency for antminer U2, while most stable is 200Mhz. I have to set "anu-freq" : "200.0" in config for stable hashing.

[-ck]

[2014-11-29 01:45:03] Started cgminer 4.8.0
[2014-11-29 01:45:03] Loaded configuration file cgminer.conf
[2014-11-29 01:45:07] ANU 0: Found at 3:1
[2014-11-29 01:45:08] Probing for an alive pool
[2014-11-29 01:45:08] Testing pool http://localhost:8332
[2014-11-29 01:45:10] HTTP request failed: Failed connect to localhost:8332; No error
[2014-11-29 01:45:11] HTTP request failed: Failed connect to localhost:8332; No error
[2014-11-29 01:45:13] Failed to connect to stratum on localhost:8332
[2014-11-29 01:45:18] Waiting for work to be available from pools.

The above is while the bitcoind is still running DB check. After 5-10 sec of waiting it clears the screen and just exits after any key pressed. The subject is cgminer 4.8.0-windows. One more issue, it takes 250MHz as default frequency for antminer U2, while most stable is 200Mhz. I have to set "anu-freq" : "200.0" in config for stable hashing.

I'm not sure what you're seeing as an issue in either of those things. Of course it can't connect to the only pool you've provided if it's dead. Wait till it's ready without pressing a key and it will work. Second, yours may be stable at 200, but they are usually stable at 250.