Quote from: restless on December 12, 2020, 01:21:05 PM

Quote from: PhoenixMiner on December 12, 2020, 12:27:18 PM

Quote from: restless on December 12, 2020, 11:48:21 AM

is there a fix for

Quote

Allocating buffers failed with: clCreateBuffer (-61).

?

We only see this error with Hawaii cards (R9 290/390). If you have other (newer) cards and still see the problem, please let us know the PhoenixMiner version, the AMD driver version, and the exact type of the cards. Also at least part of the log file will be helpful, including the error and one-two screens before that.

HiveOS latest build 181 10-12-2020, AMD drivers 19.50
Rig of 4 Radeons, first one 580 8GB (with display interface), others 470 4GB

command line - used "even"
-daglim 1 -eres 0 -rvram -1 -rmode 0

It started on 2nd december (epoch 379?) and here in the forum were other complaints from Polaris users.
log:

Code:

2020.12.12:15:43:47.097: main Phoenix Miner 5.3b Linux/gcc - Release build
2020.12.12:15:43:47.097: main Cmd line: -epool ru-eth.hiveon.net:4444 -ewal 0xe9.RX470 -epsw x -coin eth -cclock 1040 -cvddc 970 -mclock 1950 -tt 74 -daglim 0 -rmode 0 -wd 0 -eres 0 -rvram -1
2020.12.12:15:43:47.100: main Unable to enum CUDA GPUs: no CUDA-capable device is detected
2020.12.12:15:43:47.279: main Unknown OpenCL driver version! Hashrate and stale shares may suffer
2020.12.12:15:43:47.279: main OpenCL platform: OpenCL 2.1 AMD-APP (3004.6)
2020.12.12:15:43:47.279: main Available GPUs for mining:
2020.12.12:15:43:47.279: main GPU1: Radeon RX 580 Series (pcie 1), OpenCL 1.2, 8 GB VRAM, 36 CUs
2020.12.12:15:43:47.279: main GPU2: AMD Radeon (TM) RX 470 Graphics (pcie 2), OpenCL 1.2, 4 GB VRAM, 32 CUs
2020.12.12:15:43:47.279: main GPU3: AMD Radeon (TM) RX 470 Graphics (pcie 4), OpenCL 1.2, 4 GB VRAM, 32 CUs
2020.12.12:15:43:47.279: main GPU4: Radeon RX 570 Series (pcie 5), OpenCL 1.2, 4 GB VRAM, 32 CUs
2020.12.12:15:43:47.280: main AMD sysfs monitor initialized
2020.12.12:15:43:47.280: main Eth: Loading pools from epools.txt
2020.12.12:15:43:47.280: main Eth: the pool list contains 5 pools (1 from command-line)
2020.12.12:15:43:47.280: main Eth: primary pool: ru-eth.hiveon.net:4444
2020.12.12:15:43:47.280: main Starting GPU mining
2020.12.12:15:43:47.280: main GPU1: fan PWM control mode 2, min 0, max 255
2020.12.12:15:43:47.280: main GPU2: fan PWM control mode 2, min 0, max 255
2020.12.12:15:43:47.280: main GPU3: fan PWM control mode 2, min 0, max 255
2020.12.12:15:43:47.280: main GPU4: fan PWM control mode 2, min 0, max 255
2020.12.12:15:43:47.330: hwmc GPU1: set auto fan: 74C target temp (min fan 0, max fan 100)
2020.12.12:15:43:47.331: hwmc GPU2: set auto fan: 74C target temp (min fan 0, max fan 100)
2020.12.12:15:43:47.331: hwmc GPU3: set auto fan: 74C target temp (min fan 0, max fan 100)
2020.12.12:15:43:47.331: hwmc GPU4: set auto fan: 74C target temp (min fan 0, max fan 100)
2020.12.12:15:43:47.381: main Eth: Connecting to ethash pool ru-eth.hiveon.net:4444 (proto: EthProxy)
2020.12.12:15:43:47.382: main GPU1: 49C 21% 37W, GPU2: 27C 20% 19W, GPU3: 25C 24% 13W, GPU4: 29C 34% 36W
GPUs power: 105.6 W
2020.12.12:15:43:47.482: eths Eth: Connected to ethash pool ru-eth.hiveon.net:4444 (185.22.235.97)
2020.12.12:15:43:47.482: eths Eth: Send: {"id":1,"jsonrpc":"2.0","method":"eth_submitLogin","worker":"eth1.0","params":["0xe9.RX470","x"]}

2020.12.12:15:43:47.553: eths Eth: Received: {"id":1,"jsonrpc":"2.0","result":true}
2020.12.12:15:43:47.553: eths Eth: Send: {"id":5,"jsonrpc":"2.0","method":"eth_getWork","params":[]}
...

2020.12.12:15:43:47.629: GPU1 GPU1: Starting up... (0)
2020.12.12:15:43:47.630: GPU1 GPU1: Generating ethash light cache for epoch #381
2020.12.12:15:43:47.630: GPU2 GPU2: Starting up... (0)
2020.12.12:15:43:47.630: GPU3 GPU3: Starting up... (0)
2020.12.12:15:43:47.630: GPU4 GPU4: Starting up... (0)
2020.12.12:15:43:49.078: eths Eth: Received:
2020.12.12:15:43:49.078: eths Eth: New job #ae9927c8 from diff: 5000MH
2020.12.12:15:43:51.029: GPU1 Light cache generated in 3.4 s (18.7 MB/s)
2020.12.12:15:43:51.164: GPU1 GPU1: Free VRAM: 7.960 GB; used: 0.019 GB
2020.12.12:15:43:51.172: GPU1 GPU1: Allocating DAG for epoch #381 (3.98) GB
2020.12.12:15:43:51.173: GPU1 GPU1: Allocating buffers failed with: clCreateBuffer (-61).
2020.12.12:15:43:51.173: GPU1 Fatal error detected. Restarting.
2020.12.12:15:43:51.300: GPU2 GPU2: Free VRAM: 3.968 GB; used: 0.019 GB
2020.12.12:15:43:51.308: GPU2 GPU2: Allocating DAG for epoch #381 (3.98) GB
2020.12.12:15:43:51.308: GPU2 GPU2: Allocating buffers failed with: clCreateBuffer (-61).
2020.12.12:15:43:51.435: GPU3 GPU3: Free VRAM: 3.968 GB; used: 0.019 GB
2020.12.12:15:43:51.438: GPU3 GPU3: Allocating DAG for epoch #381 (3.98) GB
2020.12.12:15:43:51.439: GPU3 GPU3: Allocating buffers failed with: clCreateBuffer (-61).
2020.12.12:15:43:51.582: GPU4 GPU4: Free VRAM: 3.968 GB; used: 0.019 GB
2020.12.12:15:43:51.582: GPU4 GPU4: Allocating DAG for epoch #381 (3.98) GB
2020.12.12:15:43:51.582: GPU4 GPU4: Allocating buffers failed with: clCreateBuffer (-61).

Thank you for the amazing bug report! This is exactly how it is done

The problem is here: Unknown OpenCL driver version! Hashrate and stale shares may suffer - this means that the miner can't recognize the driver and is forced to use the generic kernels, which doesn't support most DAG configs. The quick fix is to upgrade to driver 20.10-1048554 or downgrade to driver 19.50-967956. We suspect that you are using driver 19.50-1011208, which is not supported because somehow it "fell through the cracks" and was never even tested. We will add support for it in the 5.4c for sure.

Quote from: yungplayz on Today at 01:06:11 AM

I'm bashing my head against the wall, guys. Like what am I doing wrong?
The miner just won't start, instead, that's what I'm getting:

Code:

C:\Users\Admin\Downloads\PhoenixMiner_5.3b_Windows>PhoenixMiner.exe -nvidia -pool ssl://eu1.ethermine.org:5555 -pool2 ssl://us1.ethermine.org:5555 -wal 0xAA478d95b2Ff426712fC5003404D0B6f4CEaBd4C.Rig0 -coin etc
Phoenix Miner 5.3b Windows/msvc - Release build
-----------------------------------------------
CUDA version: 11.0, CUDA runtime: 8.0
Available GPUs for mining:
GPU1: GeForce GTX 1070 (pcie 9), CUDA cap. 6.1, 8 GB VRAM, 15 CUs
GPU2: GeForce GTX 1070 (pcie 10), CUDA cap. 6.1, 8 GB VRAM, 15 CUs
C:\Users\Admin\Downloads\PhoenixMiner_5.3b_Windows>

Honestly tried googling, found nothing.
Win10 2004, latest NVidia drivers (downloaded and installed today)

The problem is with the drivers - here is how to fix it.

Quote from: adaseb on Today at 04:46:41 AM

Quote from: Nasareth on December 12, 2020, 07:22:42 PM

Hi all I am New in this Forum Very Helpfull Forum Sry for my english it is not Good.
Dear PhoenixMiner Need your Help I have one Ring With 4 Gpu 2 RX 570 4GB & 2 RX 570 8 GB.
2 gpu Rx 570 4 gb before some days have 27-28MHs Now have 19Mhs Already Check
With PhoenixMiner_5.4b , PhoenixMiner_5.2a , PhoenixMiner_5.2b , PhoenixMiner_5.2c , PhoenixMiner_5.2e With AMD driver 20.5.1 All Have Enable Compute Mode

My Line is This Have Something Wrong ?
PhoenixMiner.exe -pool ssl://eth-eu1.nanopool.org:9433 -wal (HIDE) -amd -gt 15 -amd -rmode 2 -clkernel 1,3,3,1 -rxboost 1 -eres 0,2,2,0 -dagrestart 1 -daglim 1 -rvram -1,384,384,-1 -tt 53 -tstop 75

Already Use PhoenixMiner_5.2e With AMD driver 20.5.1
Eth speed: 101.053 MH/s, shares: 1/0/0, time: 0:01
main GPUs: 1: 19.430 MH/s (0) 2: 30.038 MH/s (0) 3: 32.162 MH/s (0) 4: 19.423 MH/s (1)
My Results Have any opinion Please For Fix 2 Cards Again to go to 27-28 Mhs?
Thanks A lot Guys Have A nice to meet you all...

It’s because your 4GB are reaching the dag limit on ETH. Basically you have 2 more days on epoch 381 and when epoch 382 will start they will no longer be able to mine at all. There is another miner that if you search lets you mine for another month or two but I don’t want to mention in this thread since it’s not phoenixminer specific.

I would just however remove those 2 GPUs from the rig and sell them on Craigslist. Then add some money and buy an 8gb version and you will get your 30mhs speed back. Nothing else to do. You can mine etc at half the profit however I don’t think you want to do that.

PhoenixMiner will also work after epoch 381 but with speed decreasing with each DAG epoch - we also support "zombie" mode, we just do not call it this way (you need PhoenixMiner 5.2e or later for it to work). There is no "magic" that will keep 4GB cards working with the same hashrate as simply the DAG is becoming too big. Depending on the hashrate loss, you may be better of mining ETC with the 4GB AMD GPUs.

IMPORTANT! The latest Nvidia drivers for Windows 10 460.79 cause crash when PhoenixMiner starts. The problem is caused by a corrupted DLL file that is installed by the driver. You can see how to fix this problem in this post (it is an easy and quick fix).