New release: Version 3.8.4, 1st December 2013

Another stable version update.


Human readable changelog:

- More fixes to make usb communications more forgiving and robust which may improve reliability and speeds.
- Timeout overruns won't show unless you have verbose mode on now.
- Voltage displayed for BFL SC devices is the 2nd voltage which is allegedly more relevant.
- API stats for BFL SC devices now show a nonce and hardware error count per core.
- Json API commands should work again.
- More fixes for upcoming hashfast hardware.
- Lowmem mode has been extended to use USB sync transfers
- BXF devices should align better with other devices on the display.
- Devices will now initialise before trying to connect to pools.


Full changelog:

- Deprecate the usb usecps function and just split up transfers equal to the
maxpacketsize on usb1.1 devices.
- Retry sending after successfully clearing a pipe error.
- Drop logging of timeout overrun message to verbose level.
- Use a much longer callback timeout for USB writes on windows only as a last
resort since cancellations work so poorly.
- Use vcc2 in bflsc voltage displayed.
- Increment per core errors on false nonces in bflsc and add per core statistics
to api stats, removing debugging.
- Store a per-core nonce and hw error count for bflsc.
- Fix json parsing in api.c
- Add debugging to hfa driver for how many jobs are being sent.
- Shut down the hfa read thread if the device disappears.
- Add debug output saying what frame command is being sent in hfa driver.
- Revert "Disable USB stats which were not meant to be enabled by default and
add extra memory for a memory error when stats are enabled."
- Reset work restart flag in hfa driver since we may check for it again in
restart_wait.
- Add more op usb init errors for hfa driver.
- Perform basic displaying of hfa notices received.
- Add hfa op usb notice macros.
- Update hf protocol header.
- Use sync usb transfers in lowmem mode.
- Go back to allowing timeout errors on USB writes to be passed back to the
driver without removing the device in case the driver wishes to manage them.
- Initialise more values for the hfa data structures.
- A USB control error must be < 0
- Simplify USB NODEV error checking to success only for writes and control
transfers, and success and timeout for reads.
- libusb error IO should be fatal as well if it gets through usb read and write.
- Allow IO errors in usb reads/writes to be ignored up to retry max times.
- Use correct padding for bxf temperature display.
- Initialise devices before attempting to connect to pools to allow their thread
prepare function to be called before having to connect to pools.
- Add hidden hfa options to set hash clock, group ntime roll and pll bypass,
fixing frame sent on reset to include extra data.
- Relax the timeouts for the slower usb devices on linux.
- Add big endian hf protocol header to Makefile
- Check for correct big endian macro in hf_protocol
- Use an absolute timeout in hfa_get_header to cope with buffered usb reads
returning instantly confusing the 200ms counter.
- Update hfa_detect_one to use the new detect function API.

GPU and scrypt mining is no longer supported by cgminer and offtopic on this thread.

Quoted for posterity. Appreciated.

EDIT: I believe in bitcoin and I'm sorry but this is my thread for my software and I gave plenty of warning about GPU and scrypt being deprecated, and even helped and encouraged people to fork and take over the existing code. Amazing how quickly people turn to personal attacks when you don't support their cause (i.e. litecoin).

EDIT2: No I do not really appreciate being called names. It fucking hurts and my response was purely to try and deflect the pain.

Then this thread wouldn't belong here since this is the bitcoin mining software section, and it would be moved to altcoins, whereas I am maintaining cgminer as an ASIC bitcoin miner. Honestly, I suggest you go looking at the litecoinforum instead where they've forked the code and are developing on it.
https://forum.litecoin.net/index.php?board=3.0

Well oddly enough cgminer used to also have CPU mining code in it.
Even more oddly enough, no we don't support CPU mining any more and that is also off topic in here.

Lulz sounds like: "I want help and you must give it to me coz I say so you jerk"

Let me respond appropriately "Go fuck yourself"

I dont know if anyone reported this, but I left my computer on all night like i've been doing for weeks mining, then I woke up and I saw Malwaresbyte reporting cgminer-nogpu.exe as a trojan, specifically "Trojan.Bitminer". Im CPU mining with 3.7.2 so I was using cgminer.exe. Anyway, I disconnected and ran a complete scan, I detected Trojan.Winload on Winload-Loader-v2.2.1.zip which I guess is a false positive from Daz's Windows activator for Win7, and that's about it.
To be sure, I downloaded again http://ck.kolivas.org/apps/cgminer/3.7/cgminer-3.7.2-windows.7z scanned it and found again "Trojan.Bitminer" on cgminer-nogpu.exe.

Im asuming Malwarebytes has included that exe as a trojan because people has been installing it around to secretly mine with other people's computers. False positive?

Happy to oblige as always.  Test ran for just under 2 hours before a zombie (AMU16) appeared.  Did the usual thing with debug, then stopped the test.

Logfile here:
  https://dl.dropboxusercontent.com/u/44240170/logfile-3.8.4.txt

Help plz, if I have a laptop connected to several bfl asics, how do I open one window of cgminer for each?

Any chance of getting the Drillbit thumbs integrated? For some reason when I run 2 instances of cgminer, 1 for my BE's and 1 for DB thumbs, my laptop will crash. Not an issue with the laptop, it has been running for over a month with cgminer for the BE's and stratum proxy. 

Each use:
--usb BAS:1