GPT magic number problem #61

wojteklewandowski-pl · 2020-08-27T10:18:40Z

wojteklewandowski-pl
Aug 27, 2020

Hello, I am trying to test the files prepared using the polarfire-soc-buildroot-sdk under the Renode simulation framework. It requires the hss, fit image, boot, kernel and emmc image.There is a problem with emmc image. The hss checks its magic number and returns an error:

[0.959500] GPT_ValidatePartitionEntries(): Validated GPT Partition Entries ...
[0.959800] verifyMagic_(): magic is 00000000 vs expected b007c0de or c08b8355
[0.960000] getBootImageFromMMC_(): verifyMagic_() failed

Is it a problem with image creation or what? What would you advise?
Regards,
Wojtek

Answered by ConchuOD

Aug 28, 2020

Hi Wojtek,
I just built a fresh emmc.img using 6514a11 & running:

make clean
make DEVKIT=icicle-kit-es

I then created an emmc.img by running: (/dev/sdb is an sd card on my system)

sudo make DEVKIT=icicle-kit-es DISK=/dev/sdb format-icicle-emmc-image

Examining the hexdump I could see c0de b007 present at offset 0100000.

What process did you follow to create the emmc.img?
What is present at offset 0100000 in the emmc.img you created?

View full answer

ConchuOD · 2020-08-27T10:28:01Z

ConchuOD
Aug 27, 2020
Collaborator

Hi Wojtek,
Could you share the contents of the resc script that you are using to boot via renode? From that error it looks like the file containing the image has not been correctly mapped to the mmc.

0 replies

wojteklewandowski-pl · 2020-08-27T11:05:48Z

wojteklewandowski-pl
Aug 27, 2020
Author

Hi Conor,
Sure, but the resc sript is the same that is in renode for icycle board, but with my files:

:name: Icicle Kit
:description: This is a sample script running HSS, U-Boot and Linux on Icicle Kit with PolarFire SoC

$bin?=@/home/wojlew/renode/files2/hss.elf
$uboot?=@/home/wojlew/renode/files2/u-boot
$vmlinux?=@/home/wojlew/renode/files2/vmlinux
$fit?=@/home/wojlew/renode/files2/fitImage.fit
$emmc?=@/home/wojlew/renode/files2/emmc.img

i @scripts/single-node/polarfire-soc.resc

showAnalyzer mmuart1
machine SdhcCardFromFile $emmc mmc

macro reset
"""
    sysbus LoadBinary $fit 0x88300000
    sysbus LoadSymbolsFrom $vmlinux
    sysbus LoadSymbolsFrom $uboot
"""
runMacro $reset

To me, it looks like it is something with an image, so maybe I've made a mistake when creating it. When I Hexdump the emmc image provided by renode and mine, is smaller and in the line, wherein renode's image is a magic number, on mine is something else:
Renode emmc image:

...
00004a0 1000 0000 0000 0000 cbee 0000 0000 0000
00004b0 0000 0000 0000 0000 0072 006f 006f 0074
00004c0 0000 0000 0000 0000 0000 0000 0000 0000
*
0100000 c0de b007 0000 0000 0858 0000 0000 0000 <-- this line
0100010 b6ac a270 0000 0000 0610 0000 0000 0000
...

Mine emmc image:

...
00004a0 1000 0000 0000 0000 cbee 0000 0000 0000
00004b0 0000 0000 0000 0000 0072 006f 006f 0074
00004c0 0000 0000 0000 0000 0000 0000 0000 0000
*
197de00 6148 2168 6449 6e6f 4e74 6565 4564 4946 <-- this line
197de10 19c3 3190 f224 4fc3 e794 c850 33a6 e8aa
...

0 replies

ConchuOD · 2020-08-28T09:45:01Z

ConchuOD
Aug 28, 2020
Collaborator

Hi Wojtek,
I just built a fresh emmc.img using 6514a11 & running:

make clean
make DEVKIT=icicle-kit-es

I then created an emmc.img by running: (/dev/sdb is an sd card on my system)

sudo make DEVKIT=icicle-kit-es DISK=/dev/sdb format-icicle-emmc-image

Examining the hexdump I could see c0de b007 present at offset 0100000.

What process did you follow to create the emmc.img?
What is present at offset 0100000 in the emmc.img you created?

0 replies

wojteklewandowski-pl · 2020-08-28T10:53:11Z

wojteklewandowski-pl
Aug 28, 2020
Author

Hi Conor,
You are right, it was my mistake. I am building it within a Docker and did not notice that there was an issue with loop0p2 device. When I fixed it the hexdump is correct. Thanks, we can close this topic.
Should it run under Renode? Do you test it with it or QEMU?

0 replies

ConchuOD · 2020-08-28T11:11:46Z

ConchuOD
Aug 28, 2020
Collaborator

Hi Wojtek,
It should run in renode, however you may encounter this problem which prevents booting as far as a Linux login prompt.

You may also have to change the hash algorithm in conf/osbi-fit-image.its to sha1, if you see the following error in u-boot:

Running Linux...
## Loading kernel from FIT Image at 88300000 ...
   Using 'conf@microchip_icicle-kit-es-a000-microchip.dtb' configuration
   Trying 'kernel@1' kernel subimage
     Description:  Linux kernel
     Type:         Kernel Image
     Compression:  uncompressed
     Data Start:   0x883000fc
     Data Size:    6498192 Bytes = 6.2 MiB
     Architecture: RISC-V
     OS:           Linux
     Load Address: 0x80200000
     Entry Point:  0x80200000
     Hash algo:    sha256
     Hash value:   5b8a39aaa68519e37625d7cf48eb456c2678b006c1dade272fd13949c1387fe5
   Verifying Hash Integrity ... sha256 error!
Unsupported hash algorithm for 'hash@1' hash node in 'kernel@1' image node
Bad Data Hash
ERROR: can't get kernel image!
RISC-V #

0 replies

wojteklewandowski-pl · 2020-09-01T11:44:55Z

wojteklewandowski-pl
Sep 1, 2020
Author

Hi Conor,
One additional question, maybe you can help me or maybe it will be a question to renode. As I mentioned, hex dump is correct, and my emmc image looks ok. It even has exactly the same size as the renode's one. But when I try to boot with my emmc image there are sdhci timeouts and it is not booting. Also, there is a Warning that Input data exceeds 1048576 bytes. Maybe you have a suggestion where the problem can be?

mmuart1

RAM:  1 GiB
MMC:   sdhc@20008000: 0
In:    serial@20100000
Out:   serial@20100000
Err:   serial@20100000
Net:   
Warning: ethernet@20112000 using MAC address from ROM
eth0: ethernet@20112000
Hit any key to stop autoboot:  0 
sdhci_transfer_data: Transfer data timeout
sdhci_send_command: Timeout for status update!
sdhci_send_command: Timeout for status update!
sdhci_send_command: Timeout for status update!
## Warning: Input data exceeds 1048576 bytes - truncated
## Info: input data size = 1048578 = 0x100002
Running boot cmd...
## Error: "boot3" not defined
RISC-V #

renode output during this timeouts

1:24:23.7119 [WARNING] mmc: Unhandled write to offset 0x22C. Unhandled bits: [17-19] when writing value 0xE0002.
11:24:23.7121 [WARNING] mmc: Unhandled write to offset 0x204. Unhandled bits: [12-14] when writing value 0x17008.
11:24:23.7122 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [1, 4] when writing value 0x37020012. Tags: Block Count Enable (BCE) (0x1), Data Transfer Direction Select (DTDS) (0x1).
11:24:23.7124 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [19-21] when writing value 0x333A0000. Tags: Command CRC Check Enable (CRCCE) (0x1), Command Index Check Enable (CICE) (0x1), Data Present Select (DPS) (0x1).
11:26:12.9430 [WARNING] mmc: Unhandled write to offset 0x230. Unhandled bits: [9-13, 26, 28-31] when writing value 0xFFFFFFFF.
11:26:12.9430 [WARNING] mmc: Unhandled write to offset 0x22C. Unhandled bits: [25] when writing value 0x2000002. Tags: Software Reset For CMD Line (SRCMD) (0x1).
11:26:12.9431 [WARNING] mmc: Unhandled write to offset 0x22C. Unhandled bits: [26] when writing value 0x4000002. Tags: Software Reset For DAT Line (SRDAT) (0x1).
11:26:12.9435 [WARNING] mmc: Unhandled write to offset 0x230. Unhandled bits: [9-13, 26, 28-31] when writing value 0xFFFFFFFF.
11:26:12.9435 [WARNING] mmc: Unhandled write to offset 0x22C. Unhandled bits: [17-19] when writing value 0xE0002.
11:26:12.9435 [WARNING] mmc: Unhandled write to offset 0x204. Unhandled bits: [12-14] when writing value 0x17008.
11:26:12.9437 [WARNING] mmc.sdCard: Unsupported command: 51. Ignoring it
11:26:12.9438 [WARNING] mmc: Expected a response of length 32 bits (excluding control bits and CRC), but 0 received
11:26:12.9439 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [1, 4] when writing value 0x33020012. Tags: Block Count Enable (BCE) (0x1), Data Transfer Direction Select (DTDS) (0x1).
11:26:12.9439 [WARNING] mmc.sdCard: Unsupported command: 51. Ignoring it
11:26:12.9439 [WARNING] mmc: Expected a response of length 32 bits (excluding control bits and CRC), but 0 received
11:26:12.9439 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [19-21] when writing value 0x333A0000. Tags: Command CRC Check Enable (CRCCE) (0x1), Command Index Check Enable (CICE) (0x1), Data Present Select (DPS) (0x1).
11:26:22.1236 [WARNING] mmc: Unhandled write to offset 0x230. Unhandled bits: [9-13, 26, 28-31] when writing value 0xFFFFFFFF.
11:26:22.1237 [WARNING] mmc: Unhandled write to offset 0x22C. Unhandled bits: [17-19] when writing value 0xE0002.
11:26:22.1237 [WARNING] mmc: Unhandled write to offset 0x204. Unhandled bits: [12-14] when writing value 0x17008.
11:26:22.1238 [WARNING] mmc.sdCard: Unsupported command: 51. Ignoring it
11:26:22.1238 [WARNING] mmc: Expected a response of length 32 bits (excluding control bits and CRC), but 0 received
11:26:22.1238 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [1, 4] when writing value 0x33020012. Tags: Block Count Enable (BCE) (0x1), Data Transfer Direction Select (DTDS) (0x1).
11:26:22.1239 [WARNING] mmc.sdCard: Unsupported command: 51. Ignoring it
11:26:22.1239 [WARNING] mmc: Expected a response of length 32 bits (excluding control bits and CRC), but 0 received
11:26:22.1239 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [19-21] when writing value 0x333A0000. Tags: Command CRC Check Enable (CRCCE) (0x1), Command Index Check Enable (CICE) (0x1), Data Present Select (DPS) (0x1).
11:26:31.3120 [WARNING] mmc: Unhandled write to offset 0x230. Unhandled bits: [9-13, 26, 28-31] when writing value 0xFFFFFFFF.
11:26:31.3120 [WARNING] mmc: Unhandled write to offset 0x22C. Unhandled bits: [17-19] when writing value 0xE0002.
11:26:31.3120 [WARNING] mmc: Unhandled write to offset 0x204. Unhandled bits: [12-14] when writing value 0x17008.
11:26:31.3121 [WARNING] mmc.sdCard: Unsupported command: 51. Ignoring it
11:26:31.3121 [WARNING] mmc: Expected a response of length 32 bits (excluding control bits and CRC), but 0 received
11:26:31.3121 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [1, 4] when writing value 0x33020012. Tags: Block Count Enable (BCE) (0x1), Data Transfer Direction Select (DTDS) (0x1).
11:26:31.3122 [WARNING] mmc.sdCard: Unsupported command: 51. Ignoring it
11:26:31.3123 [WARNING] mmc: Expected a response of length 32 bits (excluding control bits and CRC), but 0 received
11:26:31.3123 [WARNING] mmc: Unhandled write to offset 0x20C. Unhandled bits: [19-21] when writing value 0x333A0000. Tags: Command CRC Check Enable (CRCCE) (0x1), Command Index Check Enable (CICE) (0x1), Data Present Select (DPS) (0x1).

Regards,
Wojtek

0 replies

ConchuOD · 2020-09-01T11:51:34Z

ConchuOD
Sep 1, 2020
Collaborator

I'm sorry, I can't help you with that at all. Best thing to do is file an issue against renode instead here: https://github.com/renode/renode/issues/

Thanks,
Conor.

0 replies

wojteklewandowski-pl · 2020-09-04T08:14:29Z

wojteklewandowski-pl
Sep 4, 2020
Author

Hi Conor, just for your information: I've found that the problem is with bootcmd, and mmc in u-boot. It looks that even mmcinfo gives timeout in u-boot on renode.
When I change the bootcmd from this SDK - bootcmd=setenv fileaddr 0x90000000; load mmc 0:2 ${fileaddr} uEnv_s-mode.txt; env import -t ${fileaddr} ${filesize}; echo 'Running boot cmd...'; run boot3 to renode's build - bootcmd=echo 'Running Linux...'; bootm ${ramdisk_addr_r} and it boots.
Regards,
Wojtek

1 reply

venkatmakam Jan 11, 2024

Hello Wojtek,
I am also trying to run the renode simulation with my yocto build. I am also facing the above verifyMagic_(): magic is 00000000 vs expected b007c0de or c08b8355 issue. I am using the latest Renode 1.14v

What I tried:

I have built the yocto image with latest sources (emmc) from https://github.com/polarfire-soc/meta-polarfire-soc-yocto-bsp

(used azure vm for the build)

I have created latest hss from https://github.com/polarfire-soc/hart-software-services
Other binaries like uboot, fit, vmlinux are taken from renode sample (from icicle-kit.resc)

Please suggest how to resolve this issue boot verification.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

PolarFire-SoC

GPT magic number problem #61

{{title}}

Replies: 8 comments 1 reply

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{title}}

{{title}}

Select a reply

PolarFire-SoC

GPT magic number problem #61

wojteklewandowski-pl Aug 27, 2020

Replies: 8 comments · 1 reply

ConchuOD Aug 27, 2020 Collaborator

wojteklewandowski-pl Aug 27, 2020 Author

ConchuOD Aug 28, 2020 Collaborator

wojteklewandowski-pl Aug 28, 2020 Author

ConchuOD Aug 28, 2020 Collaborator

wojteklewandowski-pl Sep 1, 2020 Author

ConchuOD Sep 1, 2020 Collaborator

wojteklewandowski-pl Sep 4, 2020 Author

venkatmakam Jan 11, 2024

wojteklewandowski-pl
Aug 27, 2020

Replies: 8 comments 1 reply

ConchuOD
Aug 27, 2020
Collaborator

wojteklewandowski-pl
Aug 27, 2020
Author

ConchuOD
Aug 28, 2020
Collaborator

wojteklewandowski-pl
Aug 28, 2020
Author

ConchuOD
Aug 28, 2020
Collaborator

wojteklewandowski-pl
Sep 1, 2020
Author

ConchuOD
Sep 1, 2020
Collaborator

wojteklewandowski-pl
Sep 4, 2020
Author