Asic.to Custom Firmware - Optimize Performance & Efficiency S19/S21/T21

[taserz]

didn't work  

Sorry to hear that you board is probably in bad shape a reflow might fix it.

[dutchlincoln]

@Taserz: Is it possible to combine all your work and info in the first post by editing, so all the info is in one post?
More easy to find, and more clear to people that are not so familiar with flashing etcetera. Also,as some other people post their linkt and software in your post as well, which is confusing, it might prevent unwanted errors. (especially for non-natives...)

Not meaning to criticize, just giving my opinion.

Im looking for the best firmware to put on the S9 and S9i with 1600W psu's. Is it the one in your first post? Seems to be the one for 1800W as far as i understand. Thank you.

Keep on the great job you're doing!  

[taserz]

880 aka the 8.8v one was running very well for me at 800mhz.

[cryptacks]

Hi, noob question, Flash Fixed Frequency firmware which firmware is for T9+?

Can i use the one at Bitmain website but that one is Antminer-T9-PLUS-all-201711242319-autofreq-user-Update2UBI-NF.tar.gz

[taserz]

Hi, noob question, Flash Fixed Frequency firmware which firmware is for T9+?

Can i use the one at Bitmain website but that one is Antminer-T9-PLUS-all-201711242319-autofreq-user-Update2UBI-NF.tar.gz

Yes use the T9+ latest firmware for bmminer swap.

[cryptacks]

Yes use the T9+ latest firmware for bmminer swap.

i have flashed the T9+ latest firmware.. and have swap as per instruction above.. but it's not hashing.. below are the kernel log.

Code:

[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 3.14.0-xilinx-gb190cb0-dirty (lzq@armdev2) (gcc version 4.8.3 20140320 (prerelease) (Sourcery CodeBench Lite 2014.05-23) ) #57 SMP PREEMPT Fri Dec 9 14:49:22 CST 2016
[    0.000000] CPU: ARMv7 Processor [413fc090] revision 0 (ARMv7), cr=18c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT aliasing instruction cache
[    0.000000] Machine model: Xilinx Zynq
[    0.000000] cma: CMA: reserved 128 MiB at 16800000
[    0.000000] Memory policy: Data cache writealloc
[    0.000000] On node 0 totalpages: 126976
[    0.000000] free_area_init_node: node 0, pgdat c0740a40, node_mem_map debd8000
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] PERCPU: Embedded 8 pages/cpu @debc1000 s9088 r8192 d15488 u32768
[    0.000000] pcpu-alloc: s9088 r8192 d15488 u32768 alloc=8*4096
[    0.000000] pcpu-alloc: [0] 0 [0] 1 
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: noinitrd mem=496M console=ttyPS0,115200 root=ubi0:rootfs ubi.mtd=1 rootfstype=ubifs rw rootwait
[    0.000000] PID hash table entries: 2048 (order: 1, 8192 bytes)
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes)
[    0.000000] Memory: 364356K/507904K available (5032K kernel code, 283K rwdata, 1916K rodata, 204K init, 258K bss, 143548K reserved, 0K highmem)
[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0xffff0000 - 0xffff1000   (   4 kB)
[    0.000000]     fixmap  : 0xfff00000 - 0xfffe0000   ( 896 kB)
[    0.000000]     vmalloc : 0xdf800000 - 0xff000000   ( 504 MB)
[    0.000000]     lowmem  : 0xc0000000 - 0xdf000000   ( 496 MB)
[    0.000000]     pkmap   : 0xbfe00000 - 0xc0000000   (   2 MB)
[    0.000000]     modules : 0xbf000000 - 0xbfe00000   (  14 MB)
[    0.000000]       .text : 0xc0008000 - 0xc06d1374   (6949 kB)
[    0.000000]       .init : 0xc06d2000 - 0xc0705380   ( 205 kB)
[    0.000000]       .data : 0xc0706000 - 0xc074cf78   ( 284 kB)
[    0.000000]        .bss : 0xc074cf84 - 0xc078d9fc   ( 259 kB)
[    0.000000] Preemptible hierarchical RCU implementation.
[    0.000000] 	Dump stacks of tasks blocking RCU-preempt GP.
[    0.000000] 	RCU restricting CPUs from NR_CPUS=4 to nr_cpu_ids=2.
[    0.000000] RCU: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=2
[    0.000000] NR_IRQS:16 nr_irqs:16 16
[    0.000000] ps7-slcr mapped to df802000
[    0.000000] zynq_clock_init: clkc starts at df802100
[    0.000000] Zynq clock init
[    0.000015] sched_clock: 64 bits at 333MHz, resolution 3ns, wraps every 3298534883328ns
[    0.000311] ps7-ttc #0 at df804000, irq=43
[    0.000625] Console: colour dummy device 80x30
[    0.000657] Calibrating delay loop... 1332.01 BogoMIPS (lpj=6660096)
[    0.090213] pid_max: default: 32768 minimum: 301
[    0.090437] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes)
[    0.090457] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes)
[    0.092647] CPU: Testing write buffer coherency: ok
[    0.093000] CPU0: thread -1, cpu 0, socket 0, mpidr 80000000
[    0.093060] Setting up static identity map for 0x4c4af8 - 0x4c4b50
[    0.093282] L310 cache controller enabled
[    0.093301] l2x0: 8 ways, CACHE_ID 0x410000c8, AUX_CTRL 0x72760000, Cache size: 512 kB
[    0.171041] CPU1: Booted secondary processor
[    0.260231] CPU1: thread -1, cpu 1, socket 0, mpidr 80000001
[    0.260365] Brought up 2 CPUs
[    0.260384] SMP: Total of 2 processors activated.
[    0.260393] CPU: All CPU(s) started in SVC mode.
[    0.261067] devtmpfs: initialized
[    0.263509] VFP support v0.3: implementor 41 architecture 3 part 30 variant 9 rev 4
[    0.264754] regulator-dummy: no parameters
[    0.273345] NET: Registered protocol family 16
[    0.275626] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.277945] cpuidle: using governor ladder
[    0.277958] cpuidle: using governor menu
[    0.285425] syscon f8000000.ps7-slcr: regmap [mem 0xf8000000-0xf8000fff] registered
[    0.286966] hw-breakpoint: found 5 (+1 reserved) breakpoint and 1 watchpoint registers.
[    0.286981] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.287100] zynq-ocm f800c000.ps7-ocmc: ZYNQ OCM pool: 256 KiB @ 0xdf880000
[    0.309227] bio: create slab <bio-0> at 0
[    0.310871] vgaarb: loaded
[    0.311790] SCSI subsystem initialized
[    0.312700] usbcore: registered new interface driver usbfs
[    0.312879] usbcore: registered new interface driver hub
[    0.313108] usbcore: registered new device driver usb
[    0.313636] media: Linux media interface: v0.10
[    0.313795] Linux video capture interface: v2.00
[    0.314033] pps_core: LinuxPPS API ver. 1 registered
[    0.314046] pps_core: Software ver. 5.3.6 - Copyright 2005-2007 Rodolfo Giometti <giometti@linux.it>
[    0.314175] PTP clock support registered
[    0.314538] EDAC MC: Ver: 3.0.0
[    0.315590] Advanced Linux Sound Architecture Driver Initialized.
[    0.318575] DMA-API: preallocated 4096 debug entries
[    0.318590] DMA-API: debugging enabled by kernel config
[    0.318681] Switched to clocksource arm_global_timer
[    0.339516] NET: Registered protocol family 2
[    0.340199] TCP established hash table entries: 4096 (order: 2, 16384 bytes)
[    0.340256] TCP bind hash table entries: 4096 (order: 3, 32768 bytes)
[    0.340342] TCP: Hash tables configured (established 4096 bind 4096)
[    0.340388] TCP: reno registered
[    0.340407] UDP hash table entries: 256 (order: 1, 8192 bytes)
[    0.340443] UDP-Lite hash table entries: 256 (order: 1, 8192 bytes)
[    0.340689] NET: Registered protocol family 1
[    0.341076] RPC: Registered named UNIX socket transport module.
[    0.341088] RPC: Registered udp transport module.
[    0.341096] RPC: Registered tcp transport module.
[    0.341104] RPC: Registered tcp NFSv4.1 backchannel transport module.
[    0.341118] PCI: CLS 0 bytes, default 64
[    0.341578] hw perfevents: enabled with ARMv7 Cortex-A9 PMU driver, 7 counters available
[    0.343652] futex hash table entries: 512 (order: 3, 32768 bytes)
[    0.345820] jffs2: version 2.2. (NAND) © 2001-2006 Red Hat, Inc.
[    0.346029] msgmni has been set to 967
[    0.346830] io scheduler noop registered
[    0.346843] io scheduler deadline registered
[    0.346885] io scheduler cfq registered (default)
[    0.356883] dma-pl330 f8003000.ps7-dma: Loaded driver for PL330 DMAC-2364208
[    0.356904] dma-pl330 f8003000.ps7-dma: 	DBUFF-128x8bytes Num_Chans-8 Num_Peri-4 Num_Events-16
[    0.483138] e0001000.serial: ttyPS0 at MMIO 0xe0001000 (irq = 82, base_baud = 3124999) is a xuartps
[    1.051344] console [ttyPS0] enabled
[    1.055639] xdevcfg f8007000.ps7-dev-cfg: ioremap 0xf8007000 to df866000
[    1.063304] [drm] Initialized drm 1.1.0 20060810
[    1.080398] brd: module loaded
[    1.089923] loop: module loaded
[    1.099420] e1000e: Intel(R) PRO/1000 Network Driver - 2.3.2-k
[    1.105176] e1000e: Copyright(c) 1999 - 2013 Intel Corporation.
[    1.113310] libphy: XEMACPS mii bus: probed
[    1.117676] ------------- phy_id = 0x3625e62
[    1.122469] xemacps e000b000.ps7-ethernet: pdev->id -1, baseaddr 0xe000b000, irq 54
[    1.131101] ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
[    1.137716] ehci-pci: EHCI PCI platform driver
[    1.144962] zynq-dr e0002000.ps7-usb: Unable to init USB phy, missing?
[    1.151820] usbcore: registered new interface driver usb-storage
[    1.158709] mousedev: PS/2 mouse device common for all mice
[    1.164823] i2c /dev entries driver
[    1.171712] zynq-edac f8006000.ps7-ddrc: ecc not enabled
[    1.177202] cpufreq_cpu0: failed to get cpu0 regulator: -19
[    1.183166] Xilinx Zynq CpuIdle Driver started
[    1.188033] sdhci: Secure Digital Host Controller Interface driver
[    1.194154] sdhci: Copyright(c) Pierre Ossman
[    1.198468] sdhci-pltfm: SDHCI platform and OF driver helper
[    1.205304] mmc0: no vqmmc regulator found
[    1.209347] mmc0: no vmmc regulator found
[    1.248710] mmc0: SDHCI controller on e0100000.ps7-sdio [e0100000.ps7-sdio] using ADMA
[    1.257398] usbcore: registered new interface driver usbhid
[    1.262908] usbhid: USB HID core driver
[    1.267640] nand: device found, Manufacturer ID: 0x2c, Chip ID: 0xda
[    1.273929] nand: Micron MT29F2G08ABAEAWP
[    1.277902] nand: 256MiB, SLC, page size: 2048, OOB size: 64
[    1.283849] Bad block table found at page 131008, version 0x01
[    1.290090] Bad block table found at page 130944, version 0x01
[    1.296062] nand_read_bbt: bad block at 0x000000140000
[    1.301210] 3 ofpart partitions found on MTD device pl353-nand
[    1.306961] Creating 3 MTD partitions on "pl353-nand":
[    1.312109] 0x000000000000-0x000002000000 : "BOOT.bin-env-dts-kernel"
[    1.320183] 0x000002000000-0x00000b000000 : "angstram-rootfs"
[    1.327476] 0x00000b000000-0x000010000000 : "upgrade-rootfs"
[    1.336515] TCP: cubic registered
[    1.339786] NET: Registered protocol family 17
[    1.344433] Registering SWP/SWPB emulation handler
[    1.350386] regulator-dummy: disabling
[    1.354724] UBI: attaching mtd1 to ubi0
[    1.881695] UBI: scanning is finished
[    1.893401] UBI: attached mtd1 (name "angstram-rootfs", size 144 MiB) to ubi0
[    1.900486] UBI: PEB size: 131072 bytes (128 KiB), LEB size: 126976 bytes
[    1.907227] UBI: min./max. I/O unit sizes: 2048/2048, sub-page size 2048
[    1.913930] UBI: VID header offset: 2048 (aligned 2048), data offset: 4096
[    1.920788] UBI: good PEBs: 1152, bad PEBs: 0, corrupted PEBs: 0
[    1.926757] UBI: user volume: 1, internal volumes: 1, max. volumes count: 128
[    1.933889] UBI: max/mean erase counter: 2/0, WL threshold: 4096, image sequence number: 862493722
[    1.942832] UBI: available PEBs: 0, total reserved PEBs: 1152, PEBs reserved for bad PEB handling: 40
[    1.952048] UBI: background thread "ubi_bgt0d" started, PID 1080
[    1.952053] drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
[    1.956072] ALSA device list:
[    1.956076]   No soundcards found.
[    1.972534] UBIFS: background thread "ubifs_bgt0_0" started, PID 1082
[    2.021769] UBIFS: mounted UBI device 0, volume 0, name "rootfs"
[    2.027692] UBIFS: LEB size: 126976 bytes (124 KiB), min./max. I/O unit sizes: 2048 bytes/2048 bytes
[    2.036835] UBIFS: FS size: 128626688 bytes (122 MiB, 1013 LEBs), journal size 9023488 bytes (8 MiB, 72 LEBs)
[    2.046721] UBIFS: reserved for root: 0 bytes (0 KiB)
[    2.051841] UBIFS: media format: w4/r0 (latest is w4/r0), UUID 2A24A8E1-2021-4A54-8A4E-D8E294AE00E5, small LPT model
[    2.063336] VFS: Mounted root (ubifs filesystem) on device 0:11.
[    2.070320] devtmpfs: mounted
[    2.073431] Freeing unused kernel memory: 204K (c06d2000 - c0705000)
[    2.918380] random: dd urandom read with 0 bits of entropy available
[    3.318717] 
[    3.318717] bcm54xx_config_init
[    3.928726] 
[    3.928726] bcm54xx_config_init
[    7.929556] xemacps e000b000.ps7-ethernet: Set clk to 24999999 Hz
[    7.935647] xemacps e000b000.ps7-ethernet: link up (100/FULL)
[   26.444523] In axi fpga driver!
[   26.447594] request_mem_region OK!
[   26.451014] AXI fpga dev virtual address is 0xdf9fc000
[   26.456092] *base_vir_addr = 0xc510
[   26.471421] In fpga mem driver!
[   26.474494] request_mem_region OK!
[   26.478057] fpga mem virtual address is 0xe2000000
[   27.268697] 
[   27.268697] bcm54xx_config_init
[   27.898696] 
[   27.898696] bcm54xx_config_init
[   31.899577] xemacps e000b000.ps7-ethernet: Set clk to 24999999 Hz
[   31.905594] xemacps e000b000.ps7-ethernet: link up (100/FULL)
[  215.875869] random: nonblocking pool is initialized
Detect 512MB control board of XILINX
Miner Type = T9+
set_reset_allhashboard = 0x0000ffff
set_reset_allhashboard = 0x00000000
set_reset_allhashboard = 0x0000ffff
set_reset_allhashboard = 0x0000ffff
chain[1] PIC jump to app
dsPIC33EP16GS202_get_pic_sw_version failed on Chain[1]!
dsPIC33EP16GS202_get_pic_sw_version failed on Chain[1]!
dsPIC33EP16GS202_get_pic_sw_version failed on Chain[1]!
Check chain[1] PIC fw version=0xff
chain[1] PIC need restore ...

So I have reverted back to the old bmminer and its hashing at 3.x TH/S.. any help?

[taserz]

If you flashed that dolz firmware or whatever they are still trying to fix it. If my version of bmminer is not working properly for you just revert back to stock. This is not a firmware flash like the others here but just a modified version of bmminer. Hence the request to update the miners official firmware prior to changing bmminer.

[russianjohn]

Sorry to hear that you board is probably in bad shape a reflow might fix it.

You are the best! reflow fixed it!

[taserz]

You are the best! reflow fixed it!

Did you use a rework station or did you just hit that bastard up.

[cryptacks]

If you flashed that dolz firmware or whatever they are still trying to fix it. If my version of bmminer is not working properly for you just revert back to stock. This is not a firmware flash like the others here but just a modified version of bmminer. Hence the request to update the miners official firmware prior to changing bmminer.

nope.. i didnt flash any firmware before.. i revert back to stocks and its running on 3-4 TH/s which i have update the official firmware prior to changing bmminer..

any other way that i can make it run like normal?

but why is my t9+ not working? it's brand new.. i updated the latest firmware from bitmain.. but still got issues after changing your bmminer.. appreciate your help.. thanks in advance..

[taserz]

nope.. i didnt flash any firmware before.. i revert back to stocks and its running on 3-4 TH/s which i have update the official firmware prior to changing bmminer..

any other way that i can make it run like normal?

but why is my t9+ not working? it's brand new.. i updated the latest firmware from bitmain.. but still got issues after changing your bmminer.. appreciate your help.. thanks in advance..

That is really weird... When you changed the offical bmminer back it didn't change.

You did not flash the S9 firmware right? Because that seems to be causing all sort of hell for people on T9+

[cryptacks]

That is really weird... When you changed the offical bmminer back it didn't change.

You did not flash the S9 firmware right? Because that seems to be causing all sort of hell for people on T9+

nope.. i flash the T9+ firmware from bitmain.. https://service.bitmain.com/support/download?product=Antminer%20T9%2B

after flashed, it work like normal.. but after changing bmminer.. it start to have issues..

[taserz]

nope.. i flash the T9+ firmware from bitmain.. https://service.bitmain.com/support/download?product=Antminer%20T9%2B

after flashed, it work like normal.. but after changing bmminer.. it start to have issues..

Converting bmminer back and reflash didn't change it back? The errors you have are for the sw on the pic aka on the hasboards themselves.

I just pulled the bmminer file for that one just to be safe. It was originally made to fix the firmware issue from flashing that dolzminer which we were trying to help a fell out with and figured it was compiled so might as well release it since it has all the same tweaks.

[cryptacks]

Converting bmminer back and reflash didn't change it back? The errors you have are for the sw on the pic aka on the hasboards themselves.

I just pulled the bmminer file for that one just to be safe. It was originally made to fix the firmware issue from flashing that dolzminer which we were trying to help a fell out with and figured it was compiled so might as well release it since it has all the same tweaks.

reflashed to which version? i dont have the previous version..

yeah better pull it off at the moment..

so what can i do now? lol

[taserz]

Trying to figure that out as this is pretty odd. As we did not flash a different firmware than what is on your machine. We replaced the factory bmminer with one that removed the voltage limitation and fanspeed checks. Was it mining properly on the official T9+ firmware?

[dutchlincoln]

@taserz: today i want to try one of my S9's to flash to your modded fw and try to squeeze ot 16-17Th with the stock psu.
Can you tell me what temperatures do with this mod, as there is not that much info about it. Will the miner work more efficient and temps stay roughly equal, or will they significantly rise?
I assume the latter, as the extra power should go somewhere...?  

Thanks.


Edit:

So, i installed (actually my son did, but hey...) the firmware, and it seems to work.
I put it onto 2 of my miners: one with one defective hashboard, and one that is on my top drawer and one of the coolest miners at 77 celcius as we speak.
The 2-boarder indeed increased quite some at 10.7Th stable@800 and 825Mhz. When set to 850, it drops to 10.3TH. Temp rises, but not able to compare.
The other has been set to 800 also, not tried higher yet, and gets a decent 16.3Th@800Mhz setting, with a temp. increase of 11 deg. reaching 88 celcius.

One thing which i find a bit odd, is that my Awesome miner states 16.3Th, while my pool within 10-20 seconds went to 13Th readout, and further increases by .1Th each 5-10 minutes or so. As we speak it has crawled to 13.8, still not doing better as some of my S9i's (exept according to awesome miner and the actual webbased miner screen.)

All i did so far was setting the frequency.

What are the advised next steps?

Also a few tips/requests if possible:

- Can you make the advanced screen appear just in the manu, instead of url change?
- is there apossibility to set the fan speeds?
- please add in your description how to change the rights of the firmware, not everyone knows Linux that well...
- it would be highly advisable to make a small index of which firmware does what to which miner.
- i find it a little confusing how you set the 2 firmwares in your start post and state both are optimised for S9i (?)

No harm was meant by my comment, just good intentions. I know from expirience that as im not native, sometimes i miss the "subtle way" of writing.

[cryptacks]

Trying to figure that out as this is pretty odd. As we did not flash a different firmware than what is on your machine. We replaced the factory bmminer with one that removed the voltage limitation and fanspeed checks. Was it mining properly on the official T9+ firmware?

yes.. it was.. mining was fine with the default and upgraded firmware. My machine is new..

anyone have default version of the t9+?

[taserz]

[...]

It is a bit of a mix between the 2 versions based off what psu you have. I much prefer using the 8.8v one for the s9i as it runs very stable and not too hot at 800mhz-825mhz. Anything beyond that you will get a drop in hashrate because you would need more power to run it better hence the 9volt firmware or any of the others on the github. You can tweak the frequency to the max it can for each voltage but as voltage goes up and mhz goes up efficiency is slowly lost so you need to find a happy medium. For me that is 8.8@800mhz 9.2@900mhz.

The pool doesn't know the hasrate of the miner. It just does some math of how many accepted shares over time. Most pools take 1 hour to figure out the hashrate such as slush.

-I could but that would need to be a firmware flash and I have no built firmware packages yet. Also making this a little bit more difficult than just flashing a firmware makes sure the users will read the thread before flashing a thirdparty software. I want people to know what the changes are and not just flash and hope for the best granted that will be coming.

-Fan speed can be set I posted somewhere above in the threads a way to hard set the fan speeds while connected over winscp by changing the config file.

-For the proper linux permissons it should be 0777

-All the firmwares do the same thing they just have different fixed voltages.

-The orignal firmware was to run at 9v. Then I figured I should make a bunch of different ones for different psu configurations and the more efficent s9i,s9j. You have to see which works best for you. For me I like the s9i at 8.8v@800mhz with apw3+ and 9.2v@900mhz with apw5+

[alikas74]

hi.
i have old s9 13.5th. psu - apw3++1600w
what max ths i can get using 650fixed firmware + your 8.8v bmminer?
what is your devfee?

[taserz]

hi.
i have old s9 13.5th. psu - apw3++1600w
what max ths i can get using 650fixed firmware + your 8.8v bmminer?
what is your devfee?

no dev fee. Try 700-750 mhz. Let me know how it goes on the 1600w psu.