framebuffer not working

[zasdf22]

the framebuffer node in start;s fdt is not tied to the video output HDMI.
I can't do anything with it from u-boot. Guessing I need some vc4 magic to make it actually work ? The rainbow screen works, but I can't write my own data for a splash screen.
What can I do ?

[6by9]

You'll have to provide a bit more information with regard your system and what you're attempting to do.

If there is no DT alias for display0, and /soc/fb doesn't exist or isn't enabled (happens if vc4-kms-v3d of vc4-fkms-v3d are enabled), then the firmware creates a simple-framebuffer node as /chosen/framebuffer@XXXX with the appropriate geometry, format, and address parameters for the buffer that is on the primary HDMI display (whichever monitor if only one is connected, or HDMI-0 if both are connected on a Pi4).

If those conditions are not met, then /soc/fb is a "brcm,bcm2708_fb" node, and that driver is expected to set up the framebuffer(s) as desired.

[zasdf22]

Yes I have a fdt /chosen/framebuffer@XXXXXX but when I write anything to its memory, it is not updated on the HDMI output as I (novice) would expect. All it does is show the rainbow square until later on linux starts the vc/v3d modules in an init script or udev I forget which.
I would like to know if there is anyway to draw to it as if it were a simple-video device ?

edit: to provid more information

CM4S buildroot. rpi kernel 6.1 but that all works great, no issues.
u-boot is rpi_4_32b defconfig with the following as fragment

CONFIG_BMP_24BPP=y
CONFIG_BMP_32BPP=y
CONFIG_BMP=y
CONFIG_BOOTDELAY=3
CONFIG_CMD_BMP=y
CONFIG_CMD_BSP=y
CONFIG_CMD_CACHE=y
CONFIG_CMD_ECHO=y
CONFIG_CMD_HASH=y
CONFIG_CMD_VIDEO=y
CONFIG_CONSOLE_NORMAL=y
CONFIG_DISPLAY=y
CONFIG_DM_VIDEO=y
CONFIG_DT_VIDEO=y
CONFIG_ENV_IS_IN_FAT=n
CONFIG_ENV_IS_NOWHERE=y
CONFIG_I2C_EDID=y
CONFIG_PREBOOT_DEFINED=y
CONFIG_PREBOOT="fdt addr 2600000"
CONFIG_SHA256_LEGACY=y
CONFIG_SHA256=y
CONFIG_SIMPLE_FRAMEBUFFER=y
CONFIG_SPLASH_SCREEN_ALIGN=y
CONFIG_SPLASH_SCREEN=y
CONFIG_SPLASH_SOURCE=y
CONFIG_SYS_STDIO_DEREGISTER=y
CONFIG_VIDCONSOLE=y
CONFIG_VIDEO_ANSI=y
CONFIG_VIDEO_BMP_GZIP=y
CONFIG_VIDEO_BMP_RLE8=y
CONFIG_VIDEO_BPP32=y
CONFIG_VIDEO_BRIDGE=y
CONFIG_VIDEO_DT_SIMPLEFB=y
CONFIG_VIDEO_FONT_8X16=y
CONFIG_VIDEO_MIPI_DSI=y
CONFIG_VIDEO_SIMPLE=y
CONFIG_VIDEO=y

standard fatload and bmp display in boot script. It all works as expected, except drawing to the fb provided by start elf

config.txt relevant part

start_file=start4.elf
fixup_file=fixup4.dat

device_tree=bcm2711-rpi-cm4s.dtb
kernel=u-boot.bin

uart_2ndstage=1

#disable_overscan=1

gpu_mem=128
dtoverlay=vc4-kms-v3d-pi4
max_framebuffers=2

#disable_splash=1
#framebuffer_priority=1

# force 800x600 60Hz
hdmi_force_hotplug=1
hdmi_group=2
hdmi_mode=9
framebuffer_width=800
framebuffer_height=600
enable_dpi_lcd=1
display_lcd_rotate=0

cmdline.txt (i dont know if start elf looks at this to determine how to init all the blocks..)
root=/dev/mmcblk0p2 rootwait console=tty1 fbcon=map:0 console=ttyAMA0,115200 cma=256M splash loglevel=3

also I made a thread on the forum re this
https://forums.raspberrypi.com/viewtopic.php?p=2320704

[6by9]

I know nothing about how u-boot would be handling this. At one point I believe it used the bcm2708_fb mailbox calls, and those would potentially trash what had been set up for the simple-framebuffer.

Pi4 Raspberry Pi OS Lite.
Created /etc/modprobe.d/blacklist-vc4.conf containing blacklist vc4 to stop the module loading and hence simple-framebuffer to persist.
ls /proc/device-tree/chosen/ gives me framebuffer@3e194000.
sudo memtool mw 0x3e194000 0xffffffff gives me a white dot in the top left. Increase the offset and the white dot is further across the image.

Do remember that these are physical addresses.

[6by9]

Forum thread locked with a link to here to avoid duplicated discussions.

[zasdf22]

U-Boot> fdt list /chosen
chosen {
	user-warnings = [64 74 65 72 72 6f 72 3a 20 20 20 6f 76 65 72 72 69 64 65 20 69 32 63 5f 62 75 73 3a 20 64 61 74 61 20 69 73 20 74 72 75 6e 63 61 74 65 64 20 6f 72 20 6d 61 6e 67 6c 65 64 0a 64 74 65 72 72 6f 72 3a 20 63 61 6e 27 74 20 66 69 6e 64 20 73 79 6d 62 6f 6c 20 27 62 74 27 0a 46 61 69 6c 65 64 20 74 6f 20 72 65 73 6f 6c 76 65 20 6f 76 65 72 6c 61 79 20 27 6d 69 6e 69 75 61 72 74 2d 62 74 27 0a];
	ranges;
	#size-cells = <0x00000001>;
	#address-cells = <0x00000001>;
	compatible = "simple_bus";
	rng-seed = <0x547d27a1 0x3ef588b9 0x3e6e5025 0xccf8d252 0xae7c2a84 0x046927e1 0x371962d0 0x7d018f3b 0x91ba0b2c 0xfe76ca13 0xfb3426af 0xd3630c66 0xb61b8efd 0xfd22a6c6 0x25d204a3 0xe838ad22>;
	kaslr-seed = <0xaf27e83b 0x11469e62>;
	os_prefix = [00];
	overlay_prefix = "overlays/";
	rpi-boardrev-ext = <0x40000000>;
	log = <0x3ff80000 0x0007ffe0>;
	stdout-path = "serial0:115200n8";
	bootargs = "coherent_pool=1M snd_bcm2835.enable_headphones=0 snd_bcm2835.enable_hdmi=0 8250.nr_uarts=1 video=HDMI-A-1:800x600M@60D smsc95xx.macaddr=B8:27:EB:DC:74:21 vc_mem.mem_base=0x3ec00000 vc_mem.mem_size=0x40000000  root=/dev/mmcblk0p2 rootwait console=tty1 fbcon=map:0 console=ttyAMA0,115200 cma=256M splash loglevel=3";
	phandle = <0x00000046>;
	framebuffer@3e9ea000 {
	};
	bootloader {
	};
};
U-Boot> fdt list /chosen/framebuffer@3e9ea000
framebuffer@3e9ea000 {
	status = "okay";
	compatible = "simple-framebuffer";
	format = "a8r8g8b8";
	stride = <0x00000c80>;
	height = <0x00000258>;
	width = <0x00000320>;
	reg = <0x3e9ea000 0x001db000>;
};

U-Boot> mw.l 0x3e9ea000 0xffffffff 0x1000
U-Boot> mw.l 0x3e9ea000 0xffffffff 0x10000
U-Boot> mw.l 0x3e9ea000 0xffffffff 0x100000
U-Boot> 

no visual change, except the last command made the rainbow turn into a gradient of only black and blue, and then uboot crashed. Must have overwritten something, which proves the mw.l command is working. Am I doing something wrong, am I being thick ? Please tell me if I am

[6by9]

Sorry, I have no idea how u-boot has set things up.

You shouldn't have a rainbow screen on the primary monitor at that point though, as the simple framebuffer should be on display.

Whilst unlikely to be significant here, please note that the 6.1 and 6.6 kernels are no longer being updated. 6.12 is the actively developed branch.

[pelwell]

Would you mind posting the rest of config.txt, or at least the dtoverlay/dtparam lines? There are some interesting messages encoded in your fdt:

dterror:   override i2c_bus: data is truncated or mangled
dterror: can't find symbol 'bt'
Failed to resolve overlay 'miniuart-bt'

[zasdf22]

Would you mind posting the rest of config.txt, or at least the dtoverlay/dtparam lines? There are some interesting messages encoded in your fdt:

dterror:   override i2c_bus: data is truncated or mangled
dterror: can't find symbol 'bt'
Failed to resolve overlay 'miniuart-bt'

serial dump of boot

  0.52 RPi: BOOTSYS release VERSION:69471177 DATE: 2025/05/08 TIME: 16:21:35
  0.52 BOOTMODE: 0x06 partition 0 build-ts BUILD_TIMESTAMP=1746717695 serial 4cdc7421 boardrev a03150 stc 528159
  0.53 PM_RSTS 00001000
  0.54 POWER_OFF_ON_HALT: 1 WAIT_FOR_POWER_BUTTON 0 power-on-reset 1
  0.54 part 00000000 reset_info 00000000
  0.54 uSD voltage 3.3V
  0.57 Initialising SDRAM rank 1 total-size: 8 Gbit 3200 (0x07 0x00)
  0.57 DDR 3200 0 0 8 152 BL:3
  1.53 OTP boardrev a03150 bootrom c8b0 c8b0
  1.53 Customer key hash 3c26a85c347c623373c9a66d3cd572f79b3f8daa22208f84562260a5c57c411f
  1.54 VC-JTAG unlocked
  1.57 bootconf.sig
  1.57 hash: b503f8ad7f8aea93a272a5ba5248cc5222d0c55af28a4345d0a496bdba9d16bf
  1.57 rsa2048: 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
  1.62 RSA verify
  1.39 rsa-verify pass (0x0)


  1.88 RPi: BOOTLOADER release VERSION:69471177 DATE: 2025/05/08 TIME: 16:21:35
  1.88 BOOTMODE: 0x06 partition 0 build-ts BUILD_TIMESTAMP=1746717695 serial 4cdc7421 boardrev a03150 stc 1887687
  1.97 Unsupported boot order 4
  1.17 Unsupported boot order 6
  1.20 Unsupported boot order 2
  1.28 Boot mode: SD (01) order f5
  2.07 SD HOST: 200000000 CTL0: 0x00800000 BUS: 400000 Hz actual: 390625 HZ div: 512 (256) status: 0x1fff0000 delay: 276
  2.08 SD HOST: 200000000 CTL0: 0x00800f00 BUS: 400000 Hz actual: 390625 HZ div: 512 (256) status: 0x1fff0000 delay: 276
  2.19 EMMC
  2.29 SD retry 1 oc 0
  2.32 SD HOST: 200000000 CTL0: 0x00800000 BUS: 400000 Hz actual: 390625 HZ div: 512 (256) status: 0x1fff0000 delay: 276
  2.33 OCR c0ff8080 [0]
CID: 00150100384754463452061493ba5d6a
  2.35 SD HOST: 200000000 CTL0: 0x00800f04 BUS: 25000000 Hz actual: 25000000 HZ div: 8 (4) status: 0x1fff0000 delay: 4
  2.36 SD HOST: 200000000 CTL0: 0x00800f04 BUS: 50000000 Hz actual: 50000000 HZ div: 4 (2) status: 0x1fff0000 delay: 2
  2.37 MBR: 0x00000001,  262144 type: 0x0c
  2.37 MBR: 0x00040001,    2048 type: 0x0c
  2.37 MBR: 0x00040801,    2048 type: 0x0c
  2.38 MBR: 0x00041001,13778949 type: 0x0f
  2.85 Trying partition: 0
  2.99 type: 16 lba: 1 'mkfs.fat' '  V       ^ ' clusters 65399 (4)
  2.40 rsc 4 fat-sectors 256 root dir cluster 1 sectors 32 entries 512
  2.40 FAT16 clusters 65399
  2.41 [sdcard] autoboot.txt not found
  2.41 Select partition rsts 0 C(boot_partition) 0 EEPROM config 0 result 1
  2.21 Trying partition: 1
  2.24 type: 16 lba: 1 'mkfs.fat' '  V       ^ ' clusters 65399 (4)
  2.43 rsc 4 fat-sectors 256 root dir cluster 1 sectors 32 entries 512
  2.43 FAT16 clusters 65399
  2.38 secure-boot
  2.40 Loading boot.img ...
  2.44 boot.sig
  2.44 hash: 6bae94c5315de6158fe2c37e94eb25d655474c39e5c7e62ee560fdd47a7e9e81
  2.45 ts: 1750167487
  2.45 rsa2048: 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
  3.13 Verifying
  3.99 RSA verify
  4.07 rsa-verify pass (0x0)
  4.03 MBR: 0x00000000,       0 type: 0x00
  4.03 MBR: 0x00000000,       0 type: 0x00
  4.04 MBR: 0x00000000,       0 type: 0x00
  4.04 MBR: 0x00000000,       0 type: 0x00
  4.49 Trying partition: 0
  4.51 type: 16 lba: 0 'mkfs.fat' '  V       ^ ' clusters 4589 (4)
  4.05 rsc 4 fat-sectors 20 root dir cluster 1 sectors 32 entries 512
  4.06 FAT16 clusters 4589
  4.67 Read config.txt bytes      816 hnd 0x7b8
  4.07 [ramdisk] recover4.elf not found
  4.07 [ramdisk] recovery.elf not found
  4.96 Read start4.elf bytes  2256800 hnd 0x369
  4.98 Read fixup4.dat bytes     5436 hnd 0x7bc
  4.02 0x00a03150 0x00000000 0x00001fff
  4.05 MEM GPU: 128 ARM: 896 TOTAL: 1024
  4.88 Firmware: 86ccc427f35fdc604edc511881cdf579df945fb4 Apr 17 2024 17:27:09
  4.21 Starting start4.elf @ 0xfec00200 partition 0
  4.24 PCI0 reset
  4.29 +

MESS:00:00:04.390835:0: brfs: File read: /mfs/sd/config.txt
MESS:00:00:04.393974:0: brfs: File read: 816 bytes
MESS:00:00:04.418066:0: HDMI0:EDID error reading EDID block 0 attempt 0
MESS:00:00:04.422574:0: HDMI0:EDID giving up on reading EDID block 0
MESS:00:00:04.427928:0: brfs: File read: /mfs/sd/config.txt
MESS:00:00:04.433587:0: gpioman: gpioman_get_pin_num: pin DISPLAY_SDA not defined
MESS:00:00:04.440155:0: gpioman: gpioman_get_pin_num: pin LEDS_PWR_OK not defined
MESS:00:00:04.776784:0: gpioman: gpioman_get_pin_num: pin LEDS_PWR_OK not defined
MESS:00:00:04.781190:0: *** Restart logging
MESS:00:00:04.785072:0: brfs: File read: 816 bytes
MESS:00:00:04.794722:0: hdmi: HDMI0:EDID error reading EDID block 0 attempt 0
MESS:00:00:04.799751:0: hdmi: HDMI0:EDID giving up on reading EDID block 0
MESS:00:00:04.810378:0: hdmi: HDMI0:EDID error reading EDID block 0 attempt 0
MESS:00:00:04.815401:0: hdmi: HDMI0:EDID giving up on reading EDID block 0
MESS:00:00:04.827889:0: HDMI0: hdmi_pixel_encoding: 300000000
MESS:00:00:04.831766:0: brfs: File read: /mfs/sd/bcm2711-rpi-cm4s.dtb
MESS:00:00:04.836692:0: Loaded 'bcm2711-rpi-cm4s.dtb' to 0x2600000 size 0xce28
MESS:00:00:04.856654:0: brfs: File read: 52776 bytes
MESS:00:00:04.860456:0: brfs: File read: /mfs/sd/overlays/overlay_map.dtb
MESS:00:00:04.883415:0: brfs: File read: 5195 bytes
MESS:00:00:04.885301:0: brfs: File read: /mfs/sd/config.txt
MESS:00:00:04.891545:0: brfs: File read: 816 bytes
MESS:00:00:04.898199:0: brfs: File read: /mfs/sd/overlays/vc4-kms-v3d-pi4.dtbo
MESS:00:00:04.964899:0: Loaded overlay 'vc4-kms-v3d-pi4'
MESS:00:00:04.967197:0: dtparam: i2c1=on
MESS:00:00:05.303028:0: brfs: File read: 3913 bytes
MESS:00:00:05.307607:0: brfs: File read: /mfs/sd/overlays/m41t80.dtbo
MESS:00:00:05.325054:0: Loaded overlay 'm41t80'
MESS:00:00:05.326504:0: dtparam: i2c_bus=1
MESS:00:00:05.330439:0: dterror:   override i2c_bus: data is truncated or mangled
MESS:00:00:05.359946:0: brfs: File read: 433 bytes
MESS:00:00:05.365488:0: brfs: File read: /mfs/sd/overlays/pwm.dtbo
MESS:00:00:05.384983:0: Loaded overlay 'pwm'
MESS:00:00:05.425663:0: brfs: File read: 948 bytes
MESS:00:00:05.431250:0: brfs: File read: /mfs/sd/overlays/pwmtest.dtbo
MESS:00:00:05.451206:0: Loaded overlay 'pwmtest'
MESS:00:00:05.452772:0: dtparam: spi=on
MESS:00:00:05.527126:0: brfs: File read: 495 bytes
MESS:00:00:05.531759:0: brfs: File read: /mfs/sd/overlays/mcp2515-can1.dtbo
MESS:00:00:05.559733:0: Loaded overlay 'mcp2515-can1'
MESS:00:00:05.561710:0: dtparam: oscillator=20000000
MESS:00:00:05.567278:0: dtparam: interrupt=25
MESS:00:00:05.571996:0: dtparam: spimaxfrequency=500000
MESS:00:00:05.684406:0: brfs: File read: 1793 bytes
MESS:00:00:05.686589:0: brfs: File read: /mfs/sd/overlays/ads7846.dtbo
MESS:00:00:05.723720:0: Loaded overlay 'ads7846'
MESS:00:00:05.725255:0: dtparam: cs=0
MESS:00:00:05.729871:0: dtparam: penirq=26
MESS:00:00:05.734944:0: dtparam: xohms=600
MESS:00:00:05.737555:0: dtparam: settle=5
MESS:00:00:05.747672:0: Unknown dtparam 'settle' - ignored
MESS:00:00:05.750078:0: dtparam: speed=100000
MESS:00:00:05.869894:0: brfs: File read: 2402 bytes
MESS:00:00:05.874984:0: brfs: File read: /mfs/sd/overlays/miniuart-bt.dtbo
MESS:00:00:05.900921:0: dterror: can't find symbol 'bt'
MESS:00:00:05.903066:0: Failed to resolve overlay 'miniuart-bt'
MESS:00:00:05.910808:0: brfs: File read: 1566 bytes
MESS:00:00:05.918464:0: brfs: File read: /mfs/sd/overlays/uart1.dtbo
MESS:00:00:05.938693:0: Loaded overlay 'uart1'
MESS:00:00:05.940062:0: dtparam: txd1_pin=32
MESS:00:00:05.944600:0: dtparam: rxd1_pin=33
MESS:00:00:06.030992:0: brfs: File read: 988 bytes
MESS:00:00:06.032896:0: brfs: File read: /mfs/sd/cmdline.txt
MESS:00:00:06.038140:0: Read command line from file 'cmdline.txt':
MESS:00:00:06.043996:0: 'root=/dev/mmcblk0p2 rootwait console=tty1 fbcon=map:0 console=ttyAMA0,115200 cma=256M splash loglevel=3'
MESS:00:00:06.115161:0: CMA size reduced to 256MB (total_mem 1024MB, gpu_mem 128MB)
MESS:00:00:06.210941:0: brfs: File read: 104 bytes
MESS:00:00:06.217816:0: brfs: File read: /mfs/sd/u-boot.bin
MESS:00:00:06.220285:0: Loaded 'u-boot.bin' to 0x8000 size 0x9a03c
MESS:00:00:06.226200:0: Device tree loaded to 0x2600000 (size 0xd8f8)
MESS:00:00:06.234053:0: uart: Set PL011 baud rate to 103448.300000 Hz
MESS:00:00:06.241420:0: uart: Baud rate change done...
MESS:00:00:06.243443:0: uart: Baud rate change done...


U-Boot 2025.01 (Jun 17 2025 - 14:37:53 +0100)

DRAM:  896 MiB
RPI: Board rev 0x15 unknown
RPI Unknown model (0xa03150)
Core:  211 devices, 15 uclasses, devicetree: board
MMC:   mmc@7e340000: 0
Loading Environment from nowhere... OK

entire config.txt

start_file=start4.elf
fixup_file=fixup4.dat

device_tree=bcm2711-rpi-cm4s.dtb
kernel=u-boot.bin

uart_2ndstage=1

#disable_overscan=1

gpu_mem=128
dtoverlay=vc4-kms-v3d-pi4
max_framebuffers=2

#disable_splash=1
#framebuffer_priority=1

# force 800x600 60Hz
hdmi_force_hotplug=1
hdmi_group=2
hdmi_mode=9
framebuffer_width=800
framebuffer_height=600
enable_dpi_lcd=1
display_lcd_rotate=0

dtparam=i2c1=on
dtoverlay=m41t80,i2c_bus=1

dtoverlay=pwm
dtoverlay=pwmtest

dtparam=spi=on
dtoverlay=mcp2515-can1,oscillator=20000000,interrupt=25,spimaxfrequency=500000
dtoverlay=ads7846,cs=0,penirq=26,xohms=600,settle=5,speed=100000

# Enable UART0 for serial console on ttyAMA0
dtoverlay=miniuart-bt

dtoverlay=uart1,txd1_pin=32,rxd1_pin=33

Thanks for having a look. I really appreciate it

[pelwell]

Thanks. Let's go through those errors:

dterror:   override i2c_bus: data is truncated or mangled

I can now see that that message is triggered by the m41t80 overlay, which isn't one of ours. The message is saying that there is a kind of syntax error in the overlay, but it might not be enough to stop it working for you. I'm curious about where it came from, but it's not important.

dterror: can't find symbol 'bt'
Failed to resolve overlay 'miniuart-bt'

This is to be expected: the CM4S does not have on-board Bluetooth, hence the error, but because of that UART0 is already available for use on GPIOs 14 & 15 - you just need enable_uart=1. I presume from the use of the uart1 overlay that you have something connected to the UART on GPIOs 32 & 33. If it's a Bluetooth modem, is it working for you? The CM4S DTB file is missing some of the infrastructure required to make it Bluetooth-capable in the modern way.

[zasdf22]

GP32-33 is just a plain uart for modbus. I should clean up this file a little.

$ cat m41t80-overlay.dts
/dts-v1/;
/plugin/;

/ {
    compatible = "brcm,bcm2711";

    fragment@0 {
        target = <&i2c_arm>;
        __overlay__ {
            #address-cells = <1>;
            #size-cells = <0>;

            rtc@68 {
                compatible = "st,m41t80";
                reg = <0x68>;
            };
        };
    };

    __overrides__ {
        i2c_bus = <&i2c_arm>,"";
    };
};

It does work in linux ! Thanks

[pelwell]

I see - that i2c_bus parameter declaration is completely broken, but the default is I2C1 (a.k.a. i2c_arm) so using i2c_bus=1 does nothing but cause an error message. You could just delete the whole __overrides__ node and drop the parameter use.

What I've done is add m41t80 to the standard i2c-rtc overlay, since the driver is already enabled in our builds. See raspberrypi/linux@a5f5dc5.

[zasdf22]

Any idea what's going wrong? does start.elf lock the memory or prevent the buf from being written to/updated or is this entire process so complex that it's hard to describe?

[pelwell]

The VPU (and hence start.elf) has no way to prevent the ARM cores from accessing regions of memory, so that can't be the cause of what you are seeing.

[6by9]

The VPU (and hence start.elf) has no way to prevent the ARM cores from accessing regions of memory, so that can't be the cause of what you are seeing.

Have we still got the VPU cache alias stuff going on with 2711 and need to account for it on the ARM?
Address in DT is 0x3e9ea000, but I did check the HVS dlist which had 0xfe9ea000 as it wants the VPU direct cache alias.

[zasdf22]

The VPU (and hence start.elf) has no way to prevent the ARM cores from accessing regions of memory, so that can't be the cause of what you are seeing.

Have we still got the VPU cache alias stuff going on with 2711 and need to account for it on the ARM? Address in DT is 0x3e9ea000, but I did check the HVS dlist which had 0xfe9ea000 as it wants the VPU direct cache alias.

is the actual fb memory somewhere else and it only exposes an alias region ? how to find this?