Cairnsmore1 - Quad XC6SLX150 Board

[Lethos]

I would love to test the bitstream, but virtual box usb has issues, most the time it refuses any usb connected devices, even though they are installed and work fine in Windows.
To get around a few steps, the furthest along I've got it using wget all the files via my server, since I can't get my usb pen mounted either.

It kinda worked before, now the keeps complaining about "USB device is busy with a previous request".
Tried many of the usual stuff suggested on the virtual box support site, but it's short lived or doesn't work.

Their needs to be an easier way, maybe a native program for linux, rather than having to go via virtualbox.
Surely xc3sprog would work in a normal linux environment right? Don't mind sending a few bitcoins to someone who can help me out.

Seems a waste for me to be helping towards the bounty, then can't actually use it.

Lethos,

are you trying to flash the SPI firmaware (to update it to controller rev. 1.3) or a bitstream from makomk?

For the latter, you don't need windows at all, but for the former right now you can only do it in windows.

spiccioli

I am trying to use makomk bitstream. I believe my boards were already updated to 1.3v. So how do I update to the latest bitstream without using virtualbox?

Lethos,

start your virtualbox instance and then scp xc3sprog to the linux pc.

Let's say that you have a user called "miner" on your linux box you issue this command from the virtualbox running image

Code:

scp ./xc3sprog miner@ip-of-your-linux-box:

example

Code:

scp ./xc3sprog spiccioli@192.168.1.100:

mind the colon at the end of the command.

This will ask miner's password and copy xc3sprog to the home dir of user miner on the linux box.

Then, from the linux box you use the same commands as if you were inside the virtualbox image.

Mind you that if you have multiple boards attached to the linux box you need to:

• unplug one board
• move SW1 switch 3 to OFF (start of programming)
• plug it again, this makes the board the "active" board for xc3sprog (1)
• issue a ./xc3sprog -c cm1 -j to see that the board is visible
• issue the xc3sprog command for each FPGA
• unplug the board again
• move SW1 switch 3 to ON
• move SW1 switch 1 to OFF (resets board) then after a few seconds to ON again
• when the yellow leds are on, plug the board again

(1) I don't know how to address one board between a few using xc3sprog if not using this series of plugging/unplugging to make it the "active" board.

I've used this method several times and it works with my ubuntu server 12.04.

SW2 and SW5 have all switches ON, SW3 and SW4 are ON-OFF-ON-ON
SW6 switch 1 is OFF

Tell me if it works.

spiccioli.


EDIT: the xc3sprog has to be run as sudo otherwise it cannot access the board you're trying to flash

Not having much luck doing it this way - Moving the file directly off the Virtualbox. I appreciate how much detail you went into explaining it though.

[1713496869]

1713496869

[Lethos]

I would love to test the bitstream, but virtual box usb has issues, most the time it refuses any usb connected devices, even though they are installed and work fine in Windows.
To get around a few steps, the furthest along I've got it using wget all the files via my server, since I can't get my usb pen mounted either.

It kinda worked before, now the keeps complaining about "USB device is busy with a previous request".
Tried many of the usual stuff suggested on the virtual box support site, but it's short lived or doesn't work.

Their needs to be an easier way, maybe a native program for linux, rather than having to go via virtualbox.
Surely xc3sprog would work in a normal linux environment right? Don't mind sending a few bitcoins to someone who can help me out.

Seems a waste for me to be helping towards the bounty, then can't actually use it.

You can get the program here.

sudo svn co https://xc3sprog.svn.sourceforge.net/svnroot/xc3sprog/trunk xc3sprog

You will need Cmake, libftdi1, libftdipp1 from your distributions repository. Then  do a cmake . make, and make install and you should be golden. I was never able to get the pre-built binary version to work so this is how i did it.

Going to give this a go, but with it getting late, I probably won't have any feedback till tomorrow.

[ebereon]

Update on latest bitstreams:
Total 40 FPGA's = 10 boards SN#400+. (my early board SN#0015 is not working at all)

Makomk's bitstreams:
11x 200Mh (still testing), 23x 190Mh, 1x 170Mh, 5x 160Mh. That means ~85% of all fpga's running @190Mh+ with <4% invalids (will update after a day or two)

Working bitstreams in percent:
200Mh oc = 27,5%
190Mh oc = 85%
180Mh oc = 85%
170Mh oc = 87,5%
160Mh = 100%
150Mh = 100%
140Mh = 100%


Update Glasswalker's bitstreams:
20x 175Mh

Working bitstreams in percent:
175Mh = 50% (runs only on 2 FPGA's per Board)
125Mh = 0% (not working at all)

eb

[misternoodle]

For giggles, I ended up flashing my USB controller again with 1.3.  I have the 190 OC bitstream on there already, after the controller reflash, I'm able to successfully hash using CGMiner and reporting around 10.2 U/M.  Thanks for the help and suggestions so far from everyone.

[kano]

I've written an Icarus change for cgminer that will support 3 new options:
baud rate (115200 or 57600) work divisor (1, 2, 4 or 8 ) and number of FPGA
This should even work with the old setup with only 1 of 2 FPGA working
https://github.com/ckolivas/cgminer/pull/283
Anyone interested come visit #cgminer as usual ... tomorrow ...

[ebereon]

...
Running at 57600 may alleviate USB noise issues.   Are the people running MPBMM using 57.6?  Does it impact overall hash rates?
...

What you mean with USB noise issues? Never heard of that.

The baud rate depends on what controller version you use, with rev. 1.3 you need to use 115200 baud, with glasswalker2 rev. 1.0 you need to use 57600 baud.

eb

[Lethos]

I would love to test the bitstream, but virtual box usb has issues, most the time it refuses any usb connected devices, even though they are installed and work fine in Windows.
To get around a few steps, the furthest along I've got it using wget all the files via my server, since I can't get my usb pen mounted either.

It kinda worked before, now the keeps complaining about "USB device is busy with a previous request".
Tried many of the usual stuff suggested on the virtual box support site, but it's short lived or doesn't work.

Their needs to be an easier way, maybe a native program for linux, rather than having to go via virtualbox.
Surely xc3sprog would work in a normal linux environment right? Don't mind sending a few bitcoins to someone who can help me out.

Seems a waste for me to be helping towards the bounty, then can't actually use it.

You can get the program here.

sudo svn co https://xc3sprog.svn.sourceforge.net/svnroot/xc3sprog/trunk xc3sprog

You will need Cmake, libftdi1, libftdipp1 from your distributions repository. Then  do a cmake . make, and make install and you should be golden. I was never able to get the pre-built binary version to work so this is how i did it.

Going to give this a go, but with it getting late, I probably won't have any feedback till tomorrow.

Thanks, managed to get it working I beleive, assuming the update of a new bitstream goes okay, I'll send a little tip your way.

[spiccioli]

I've built cgminer 2.6.1 for linux 32bit (works ok on my ubuntu 12.04 64bit) with OpenCL/ADL disabled and just icarus support enabled.

If you trust my build you can get it from:

http://p2pool.soon.it/cgminer/cgminer-2.6.1

Download it, issue a

Code:

chmod 755 cgminer-2.6.1

to make it executable and it's ready to go.

spiccioli

[ShadesOfMarble]

my early board SN#0015 is not working at all

Are you sure this board isn't broken? I can run 150 Mhz just fine on my S/N 26 board.
Where are the other pre-S/N-50 board users? I would love to hear some more experiences so maybe we can get a better image on how "bad" pre-50 "really" are...

[ebereon]

my early board SN#0015 is not working at all

Are you sure this board isn't broken? I can run 150 Mhz just fine on my S/N 26 board.
Where are the other pre-S/N-50 board users? I would love to hear some more experiences so maybe we can get a better image on how "bad" pre-50 "really" are...

It's waiting for pickup from enterpoint. I got it already replaced. So yes it seems broken, but i can flash forth and back and switching what i want and led also working, but it's not hashing. Only shipping bitstream i think was working on 2 fpga's.

[norulezapply]

my early board SN#0015 is not working at all

Are you sure this board isn't broken? I can run 150 Mhz just fine on my S/N 26 board.
Where are the other pre-S/N-50 board users? I would love to hear some more experiences so maybe we can get a better image on how "bad" pre-50 "really" are...

Mine are being replaced as we speak. Couldn't run twin_test on them..

[Doff]

I got board 62-0013 hashing at about 120mh at 1.9U with makomk 160.bit, nothing else works on it including the original shipping test bitstream.

I should also say it only works on one pair.

[Isokivi]

Where Im at now:

Makomk's 190mhs bitstream on all boards, two fgpa pairs only working at half the expected rate. For some ppl identifying which pair is failing has been as easy as unplugging the usb and checking what led stays on. For me they all stay on, so in order to flash a functioning bitstream to all fgpa-pairs I'd have to do a fairly annoying amount of flashing and seeing what works. Does anyone have any toughts on how I could identify the failing cores ?

[Doff]

Where Im at now:

Makomk's 190mhs bitstream on all boards, two fgpa pairs only working at half the expected rate. For some ppl identifying which pair is failing has been as easy as unplugging the usb and checking what led stays on. For me they all stay on, so in order to flash a functioning bitstream to all fgpa-pairs I'd have to do a fairly annoying amount of flashing and seeing what works. Does anyone have any toughts on how I could identify the failing cores ?

Are you using Linux or windows? Also I know this is a broken record but try different USB cables, Also try a good Powered USB hub. The powered USB hub allowed Makomk's bitstream to actually work on my board 62-177. I also use the powered USB to update the controller just in case its providing a less noisy update for some reason.

[Isokivi]

Where Im at now:

Makomk's 190mhs bitstream on all boards, two fgpa pairs only working at half the expected rate. For some ppl identifying which pair is failing has been as easy as unplugging the usb and checking what led stays on. For me they all stay on, so in order to flash a functioning bitstream to all fgpa-pairs I'd have to do a fairly annoying amount of flashing and seeing what works. Does anyone have any toughts on how I could identify the failing cores ?

Are you using Linux or windows? Also I know this is a broken record but try different USB cables, Also try a good Powered USB hub. The powered USB hub allowed Makomk's bitstream to actually work on my board 62-177. I also use the powered USB to update the controller just in case its providing a less noisy update for some reason.

Windows 7 64-bit, unpowered hub (only one working for me!), usb3 port, cgminer 2.6.1. Sadly I have a fairly limited amount of suitable-looking usb-cables lying around and after buying 2 hubs and contributing to the bitstream bounty Im actually pushing my RoI further. Does anyone know of a method suitable for testing usb-cables ?

[Isokivi]

Copypasta from the bounty thread:

I would once again like to urge more people who own cm1's to contribute to this bounty, you know there are a handfull of brillian individuals workin very hard at the moment to provide us a more efficient tool for earning. Most of you are allready benefitting from this. Im not saying it's only because of this bounty.. but Im fairly sure the pace things have been progressing for the last week or so has much to do with this bounty and as any well-informed virtual miner knows time is everything in this game.

https://bitcointalk.org/index.php?topic=94317.0

[Lethos]

I've built cgminer 2.6.1 for linux 32bit (works ok on my ubuntu 12.04 64bit) with OpenCL/ADL disabled and just icarus support enabled.

If you trust my build you can get it from:

http://p2pool.soon.it/cgminer/cgminer-2.6.1

Download it, issue a

Code:

chmod 755 cgminer-2.6.1

to make it executable and it's ready to go.

spiccioli

I didn't notice any major improvements over 2.5, other then the addition of litcoin mining. Which while interesting development, doesn't concern me much.
In your opinion is it worth upgrade 2.6, if you got 2.5 works fine.

[Doff]

I'm sure there is a tool out there, but I'm just using trial and error like most everyone else. I know zefir had explained to me that you can remove one of the wire's within the USB cables to make it so they wont draw power which is what the Powered USB fixes, but he would be able to explain in better detail how that's. done.

[Doff]

I've built cgminer 2.6.1 for linux 32bit (works ok on my ubuntu 12.04 64bit) with OpenCL/ADL disabled and just icarus support enabled.

If you trust my build you can get it from:

http://p2pool.soon.it/cgminer/cgminer-2.6.1

Download it, issue a

Code:

chmod 755 cgminer-2.6.1

to make it executable and it's ready to go.

spiccioli

I didn't notice any major improvements over 2.5, other then the addition of litcoin mining. Which while interesting development, doesn't concern me much.
In your opinion is it worth upgrade 2.6, if you got 2.5 works fine.

2.6.0 will allow anything using the Glasswalker controller to work since it has the 56k baud rate built in, whereas 2.5 does not have that yet. Since we aren't using the Glasswalker bitstreams currently it doesn't do much over 2.5.

[daemonic]

my early board SN#0015 is not working at all

Are you sure this board isn't broken? I can run 150 Mhz just fine on my S/N 26 board.
Where are the other pre-S/N-50 board users? I would love to hear some more experiences so maybe we can get a better image on how "bad" pre-50 "really" are...

My board #18 is back with Enterpoint, waiting to receive my replacement