AOCLBF (was Phoenix Rising): Phoenix/POCLBM GUI With Built-In Overclock Tool

[dishwara]

I need some feedback on how well my temperature monitor is working (AMD/ATI GPUs only).
http://www.mediafire.com/?oyd7tj6ezvebaiz
Does it list both the slot the GPUs are plugged in and the temperature correctly?
Also, is Device 0 the one in first slot, Device 1 the GPU in the next-filled slot, etc.?
What I mean is, for example, I'm using slot 0 and 3, and Device 0 is the one in slot 0, Device 1 is the one in slot 3. Is the Device Number always incremented like that?

It displays wrongly. I have 3 cards in pcie_1, pcie_2 & pcie_3 slot thats Device 0,1 & 2, but temperature monitor shows as Device 0 temperature, Device 3 temperature & Device 6 temperature.
Only device number is wrong. But shows temperature correctly for each device.

[1713891200]

1713891200

[lvlrdka22]

About the restart bug thing, try this: http://www.mediafire.com/?yy43ik6nr58a85t

I need some feedback on how well my temperature monitor is working (AMD/ATI GPUs only).
http://www.mediafire.com/?oyd7tj6ezvebaiz
Does it list both the slot the GPUs are plugged in and the temperature correctly?
Also, is Device 0 the one in first slot, Device 1 the GPU in the next-filled slot, etc.?
What I mean is, for example, I'm using slot 0 and 3, and Device 0 is the one in slot 0, Device 1 is the one in slot 3. Is the Device Number always incremented like that?

It displays wrongly. I have 3 cards in pcie_1, pcie_2 & pcie_3 slot thats Device 0,1 & 2, but temperature monitor shows as Device 0 temperature, Device 3 temperature & Device 6 temperature.
Only device number is wrong. But shows temperature correctly for each device.

Does it at least list the device numbers' order correctly? As in, Device 0 is Device 0 on the monitor, Device 1 is 3 on monitor, 2 is 6, etc.

I need some feedback on how well my temperature monitor is working (AMD/ATI GPUs only).
http://www.mediafire.com/?oyd7tj6ezvebaiz
Does it list both the slot the GPUs are plugged in and the temperature correctly?
Also, is Device 0 the one in first slot, Device 1 the GPU in the next-filled slot, etc.?
What I mean is, for example, I'm using slot 0 and 3, and Device 0 is the one in slot 0, Device 1 is the one in slot 3. Is the Device Number always incremented like that?

Tested in 3 rigs:

1x 6990 - ok - 2 devices detected - Device 0 and 1 worked properly;
1x 5970 - just onde device detected - Device 0
2x 5970 - just 2 devices detected - Device 0 and 3

Hm, mind trying this with the 5970s?
http://www.mediafire.com/?6fwbxckivpohml2

Same result, lvlrdka22.

Bleh.
By the way, does the 5970 happen to only show 1 temperature for both cores?
The 6990 shows 2 temperatures, 1 per core, right?

[dikidera]

Then make the fields bigger lol. I am not some ant to try and read that. Also, why write 5Xxx+?? Just write BFI_INT so there are no misunderstandings then.
And also, if it stays on top of every program(the osd) then how the hell can i watch a movie without that thing??

Is that your complaint about GUIMiner as well? They're the same size.
5xxx+ because not everyone understand what BFI_INT means. And I did say that they're the same thing on the first post. Do you just hit "Download" and double click when finished?
Drag it off screen then.

Most of the time.

Guiminer unlike this one has very readable text, and phoenixminer is coming for the phoenix which should allow even smoother mining. My advice is to add the speed on the gui itself.

[Peao]

I do not know, but is it essential to your plans to have the temperature information of the two GPUs?
Perhaps one measure is enough, because the card is the same and there is no big difference between the temperature of the two processors of a 5970 ...

[lvlrdka22]

Bleh.
By the way, does the 5970 happen to only show 1 temperature for both cores?
The 6990 shows 2 temperatures, 1 per core, right?

5970 - 1 temp. But I do not know if it is the average of the two cores or it's the temperature of just one core.
6990 - yes - 2 temps, 1 per core.

Hm, do you happen to have GPUZ/Afterburner? They should list 2 GPUs for the 5970. If you just mine on one core, do they show different temperatures for the cores?

I do not know, but is it essential to your plans to have the temperature information of the two GPUs?
Perhaps one measure is enough, because the card is the same and there is no big difference between the temperature of the two processors of a 5970 ...

There should be though, if you mine on 1 core and play games/idle on another, which is what I'm trying to avoid   (mining on one core with a preset shutdown temperature, but the monitor reads temperature for the "idle" core).
Does this list two temperatures for the 5970? http://www.mediafire.com/?2rdbd97anpia8iw
Ignore the 10 devices thing, I'm having it list everything.
If not, I give up for now and probably will just throw a hacky fix that forces people to find out which "slot" each of the devices are on.

[dishwara]

Afterburner i am using to read temp now.

The order of device number is also different.
GPU1 temp in Afterburner = Device 6 temp
GPU2 temp in Afterburner = Device 0 temp
GPU3 temp in Afterburner = Device 3 temp

[lvlrdka22]

Afterburner i am using to read temp now.

The order of device number is also different.
GPU1 temp in Afterburner = Device 6 temp
GPU2 temp in Afterburner = Device 0 temp
GPU3 temp in Afterburner = Device 3 temp

Still the same results in both rigs...  

1x5970 just device 0
2x5970 devices 0 and 3

If you don't mind, can you mine on one core and have the other core sitting idle for the 5970, and monitor temperatures using Afterburner or GPUZ (open 2 instances of it, the drop down menu on the bottom left should list 2 5900 series, one for each core)? Do they show different temperatures?

Because of dishwara, I think I'll have to have the user figure out which device is which GPU with an external script .

By the way, did this fix the restart bug?

About the restart bug thing, try this: http://www.mediafire.com/?yy43ik6nr58a85t

[Peao]

Hmm, was a little difficult, because even with only one core working, the temperature of the stopped core closely followed the temperature of the core that was @ 100%.
But I saw that appear different temperatures for each core (about 2 Celsius), and "Device Finder" catch one of them.

[lvlrdka22]

I suppose that'll be enough as a temporary solution  .

[FRanz33]

Nope it didnt. tried it on two machines with two 6970's in each. just checked one box was going the other hung up

EDIT: if your having trouble reproducing the bug just hit stop on the console and alot of time the box will pop up and sometimes itll just stay there and hang. its something to do with how phoenix process is shut down

[lvlrdka22]

Hm, try this (temperature monitor is still not implemented yet, ignore it for now): http://www.mediafire.com/?u1tmtamjekr8b9a
DO NOT USE THE 64-BIT. It doesn't kill phoenix processes correctly.
Switched from using ctrl+break to directly killing the process. Should not have the hanging bug anymore. Let me know if there are any side effects.

[FRanz33]

OK will try. ur pretty quick on this man

[FRanz33]

ooo yea that kill is much cleaner than the last version...lets see how it does overnight

Edit: nope. Now when i hit the stop button it doesn't kill the process it just run in background. I dont know if it affects the auto restart but the manual stop and start is botched now. i have to go into task man. to kill it myself

Editx2: i can now confirm even on restart the process doent kill at all.. ever 30 sec another phoenix instance and piles up until computer crashes. take this version down asap

[FRanz33]

Maybe theres away to just kill phoenix then kill the window that pops up automatically?

[lvlrdka22]

NEVERMIND, apparently the new kill process function I stole has broken compatibility with 64-bit.
From now on, only 32-bit binaries  .
Besides, AMD's DLLs for monitoring GPU temperature is 32-bit.

If the hang is finally fixed, I might be able to get temperature monitoring up (but that requires a bit of extra work on the user side, to figure out which devices are which).

[FRanz33]

NEVERMIND, apparently the new kill process function I stole has broken compatibility with 64-bit.
From now on, only 32-bit binaries  .
Besides, AMD's DLLs for monitoring GPU temperature is 32-bit.

If the hang is finally fixed, I might be able to get temperature monitoring up (but that requires a bit of extra work on the user side, to figure out which devices are which).

ur right. 32bit works...im going to try all my unit tonight. Ill report back in the morning

[dishwara]

I tired with 1.32 new now for hang testing.
I killed one phoenix.exe in task manager & it was restarted correctly, but with 4-5 retries of showing cmd windows restarting.
I did it only one time, so not sure.

Besides, yesterday night, i got strange problem.
I used 1.32 which u gave before, i never used your GUI, actually i was not aware a GUI for phoenix with phatk exists.
1.32 gave restart hang. So, i tried with 1.31 & at the same time i updated windows 7 32 bit to sp1.
After installing sp1, even 1.31 gave hang problem. So i uninstalled sp1 & then 1.31 worked fine by correctly restarting.

I think the new sp1 interfering with PR.
So, ask others also to uninstall sp1, if they have upgraded to sp1 & try with 1.32

EDIT:

Just know one miner crashed itself & after 4-5 restarts it restarted.
But it seems miner crashes itself with new version.
Within 10 minutes, 2 miners crashed.

So going back to 1.31 as my SH won't happy if i mine less.

[lvlrdka22]

I tired with 1.32 new now for hang testing.
I killed one phoenix.exe in task manager & it was restarted correctly, but with 4-5 retries of showing cmd windows restarting.
I did it only one time, so not sure.

Besides, yesterday night, i got strange problem.
I used 1.32 which u gave before, i never used your GUI, actually i was not aware a GUI for phoenix with phatk exists.
1.32 gave restart hang. So, i tried with 1.31 & at the same time i updated windows 7 32 bit to sp1.
After installing sp1, even 1.31 gave hang problem. So i uninstalled sp1 & then 1.31 worked fine by correctly restarting.

I think the new sp1 interfering with PR.
So, ask others also to uninstall sp1, if they have upgraded to sp1 & try with 1.32

EDIT:

Just know one miner crashed itself & after 4-5 restarts it restarted.
But it seems miner crashes itself with new version.
Within 10 minutes, 2 miners crashed.

So going back to 1.31 as my SH won't happy if i mine less.

Did you try the 32-bit in this version? I've had it running for 4 hours now, no hangs so far. Only a few restarts, but that's because of my flaky net (need to replace router and modem).
Only one rig, and using only W7 Ultimate with no SP, so can't really test anything for bugs, sorry .

Hm, try this (temperature monitor is still not implemented yet, ignore it for now): http://www.mediafire.com/?u1tmtamjekr8b9a
DO NOT USE THE 64-BIT. It doesn't kill phoenix processes correctly.
Switched from using ctrl+break to directly killing the process. Should not have the hanging bug anymore. Let me know if there are any side effects.

[dishwara]

I have windows 7 Ultimate, 32 bit only. I tested & ran with this file, Phoenix Rising 1.32 x32.exe.
The other file, Phoenix Rising 1.32.exe, gives me the following error,
"The version of this file is not compatible with the version of Windows you're running. Check your Computer's information..............................."

[lvlrdka22]

1.4! Let me know any and all bugs, as usual.

I have windows 7 Ultimate, 32 bit only. I tested & ran with this file, Phoenix Rising 1.32 x32.exe.
The other file, Phoenix Rising 1.32.exe, gives me the following error,
"The version of this file is not compatible with the version of Windows you're running. Check your Computer's information..............................."

Did that version above your post still have the hang bug?