[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Chen-Yu Tsai]

This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
bisect, but v2018.07-rc1 is a working version. From the EHCI error
messages, I'm thinking it might be related to the USB changes lately.
Not sure if any other SoCs (ex. A64) are broken or not at the moment.

See log:

U-Boot SPL 2018.07 (Jul 10 2018 - 10:21:03 +0800)
DRAM: 1024 MiB
Trying to boot from MMC1


U-Boot 2018.07 (Jul 10 2018 - 10:21:03 +0800) Allwinner Technology

CPU:   Allwinner H3 (SUN8I 1680)
Model: Libre Computer Board ALL-H3-CC H3
DRAM:  1 GiB
MMC:   SUNXI SD/MMC: 0
Loading Environment from FAT... Unable to use mmc 0:1... Failed (-5)
In:    serial
Out:   serial
Err:   serial
Net:   phy interface0
eth0: ethernet at 1c30000
starting USB...
USB0:   USB EHCI 1.00
USB1:   USB OHCI 1.0
USB2:   USB EHCI 1.00
USB3:   USB OHCI 1.0
USB4:   USB EHCI 1.00
USB5:   USB OHCI 1.0
scanning bus 0 for devices... 1 USB Device(s) found
scanning bus 2 for devices... 1 USB Device(s) found
scanning bus 4 for devices... 1 USB Device(s) found
       scanning usb for storage devices... 0 Storage Device(s) found
Hit any key to stop autoboot:  0
switch to partitions #0, OK
mmc0 is current device
Scanning mmc 0:1...
Found U-Boot script /boot/boot.scr
385 bytes read in 8 ms (46.9 KiB/s)
## Executing script at 43100000
6343768 bytes read in 295 ms (20.5 MiB/s)
18710 bytes read in 10 ms (1.8 MiB/s)
## Flattened Device Tree blob at 43000000
   Booting using the fdt blob at 0x43000000
EHCI failed to shut down host controller.
EHCI failed to shut down host controller.
EHCI failed to shut down host controller.
   Loading Device Tree to 49ff8000, end 49fff915 ... OK

Starting kernel ...

data abort
pc : [<7dfadc6c>]          lr : [<7df82ed7>]
reloc pc : [<4a02bc6c>]    lr : [<4a000ed7>]
sp : 79f5ce90  ip : 79f62b8a     fp : 00000003
r10: 7dfc3d14  r9 : 79f61ed8     r8 : 00000400
r7 : 00000000  r6 : 42000000     r5 : 49ff8000  r4 : 00000000
r3 : 49ff8000  r2 : 49ff8000     r1 : 00000000  r0 : 00000000
Flags: nZCv  IRQs off  FIQs off  Mode UK6_32
Code: e12fff1e e52de008 fa000001 e3a00000 (e49df008)
Resetting CPU ...

resetting ...

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Jagan Teki]

On Tue, Jul 10, 2018 at 8:34 AM, Chen-Yu Tsai <wens@csie.org> wrote:
> This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
> bisect, but v2018.07-rc1 is a working version. From the EHCI error
> messages, I'm thinking it might be related to the USB changes lately.
> Not sure if any other SoCs (ex. A64) are broken or not at the moment.

I'm able to boot with BPI-M2+, where I changed the configurations
based on Libre Computer board.
- remove/disable ehci0, ohci0
- enable ehci1, 2, 3 and ohci1, 2, 3
- remove/disable usb_otg node
- disable MUSB_GADGET

Here is the log:
U-Boot 2018.07-00008-g9b89c61c72-dirty (Jul 10 2018 - 11:37:46 +0530)
Allwinner Technology

CPU:   Allwinner H3 (SUN8I 1680)
Model: Banana Pi BPI-M2-Plus
DRAM:  1 GiB
MMC:   SUNXI SD/MMC: 0, SUNXI SD/MMC: 1
Loading Environment from FAT... Unable to use mmc 1:3... Failed (-5)
In:    serial
Out:   serial
Err:   serial
Net:   No ethernet found.
starting USB...
USB0:   USB EHCI 1.00
USB1:   USB OHCI 1.0
USB2:   USB EHCI 1.00
USB3:   USB OHCI 1.0
USB4:   USB EHCI 1.00
USB5:   USB OHCI 1.0
scanning bus 0 for devices... 1 USB Device(s) found
scanning bus 2 for devices... 1 USB Device(s) found
scanning bus 4 for devices... 1 USB Device(s) found
       scanning usb for storage devices... 0 Storage Device(s) found
Hit any key to stop autoboot:  0
switch to partitions #0, OK
mmc0 is current device
Scanning mmc 0:1...
Found /extlinux/extlinux.conf
Retrieving file: /extlinux/extlinux.conf
155 bytes read in 3 ms (49.8 KiB/s)
1:      linux-4.17.0-rc3
Retrieving file: /zImage
3817968 bytes read in 175 ms (20.8 MiB/s)
append: console=ttyS0,115200 earlyprintk root=/dev/mmcblk0p2 rootwait
Retrieving file: /sun8i-h3-bananapi-m2-plus.dtb
15691 bytes read in 4 ms (3.7 MiB/s)
## Flattened Device Tree blob at 43000000
   Booting using the fdt blob at 0x43000000
EHCI failed to shut down host controller.
EHCI failed to shut down host controller.
EHCI failed to shut down host controller.
   Loading Device Tree to 49ff9000, end 49fffd4a ... OK

Starting kernel ...

[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 4.15.0-rc6-00415-g2f2b41a952de
(jagan at jagan-XPS-13-9350) (gcc version 6.3.1 20170109 (Linaro GCC
6.3-2017.02)) #1 S8
[    0.000000] CPU: ARMv7 Processor [410fc075] revision 5 (ARMv7), cr=10c5387d

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Chen-Yu Tsai]

On Tue, Jul 10, 2018 at 2:13 PM, Jagan Teki <jagannadh.teki@gmail.com> wrote:
> On Tue, Jul 10, 2018 at 8:34 AM, Chen-Yu Tsai <wens@csie.org> wrote:
>> This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
>> bisect, but v2018.07-rc1 is a working version. From the EHCI error
>> messages, I'm thinking it might be related to the USB changes lately.
>> Not sure if any other SoCs (ex. A64) are broken or not at the moment.
>
> I'm able to boot with BPI-M2+, where I changed the configurations
> based on Libre Computer board.
> - remove/disable ehci0, ohci0
> - enable ehci1, 2, 3 and ohci1, 2, 3
> - remove/disable usb_otg node
> - disable MUSB_GADGET

On BPI-M2+, with Sinovoip_BPI_M2_Plus_defconfig,
cross-compiled using gcc version 7.3.0 (Debian 7.3.0-20)

U-Boot SPL 2018.07 (Jul 11 2018 - 10:32:26 +0800)
DRAM: 1024 MiB
Trying to boot from MMC2


U-Boot 2018.07 (Jul 11 2018 - 10:32:26 +0800) Allwinner Technology

CPU:   Allwinner H3 (SUN8I 1680)
Model: Banana Pi BPI-M2-Plus
DRAM:  1 GiB

MMC:   SUNXI SD/MMC: 0, SUNXI SD/MMC: 1
Loading Environment from FAT... Unable to use mmc 1:1... Failed (-5)
In:    serial
Out:   serial
Err:   serial
Allwinner mUSB OTG (Peripheral)
Net:   phy interface7
eth0: ethernet at 1c30000
Warning: usb_ether using MAC address from ROM
, eth1: usb_ether
starting USB...
USB0:   USB EHCI 1.00
USB1:   USB OHCI 1.0
USB2:   USB EHCI 1.00
USB3:   USB OHCI 1.0
USB4:   USB EHCI 1.00
USB5:   USB OHCI 1.0
scanning bus 0 for devices... 1 USB Device(s) found
scanning bus 2 for devices... 1 USB Device(s) found
scanning bus 4 for devices... 1 USB Device(s) found
       scanning usb for storage devices... 0 Storage Device(s) found
Hit any key to stop autoboot:  0
switch to partitions #0, OK
mmc1(part 0) is current device
Scanning mmc 1:1...
Found U-Boot script /boot/boot.scr
494 bytes read in 4 ms (120.1 KiB/s)
## Executing script at 43100000
5847064 bytes read in 268 ms (20.8 MiB/s)
16209 bytes read in 5 ms (3.1 MiB/s)
## Flattened Device Tree blob at 43000000
   Booting using the fdt blob at 0x43000000
EHCI failed to shut down host controller.
EHCI failed to shut down host controller.
   Loading Device Tree to 49ff9000, end 49ffff50 ... OK

Starting kernel ...

data abort
pc : [<3f600018>]          lr : [<7df79ed7>]
reloc pc : [<0b687018>]    lr : [<4a000ed7>]
sp : 79f53658  ip : 79f59c82     fp : 00000400
r10: 7dfc2d70  r9 : 79f58ed8     r8 : 79fb2ca8
r7 : 00000000  r6 : 42000000     r5 : 49ff9000  r4 : 00000000
r3 : 00000006  r2 : 49ff9000     r1 : 00000006  r0 : 00000006
Flags: nzcv  IRQs off  FIQs off  Mode UK6_32
Code: 00012003 00013003 00014003 00015003 (00016003)
Resetting CPU ...

resetting ...

That PC looks really weird.

ChenYu

> Here is the log:
> U-Boot 2018.07-00008-g9b89c61c72-dirty (Jul 10 2018 - 11:37:46 +0530)
> Allwinner Technology
>
> CPU:   Allwinner H3 (SUN8I 1680)
> Model: Banana Pi BPI-M2-Plus
> DRAM:  1 GiB
> MMC:   SUNXI SD/MMC: 0, SUNXI SD/MMC: 1
> Loading Environment from FAT... Unable to use mmc 1:3... Failed (-5)
> In:    serial
> Out:   serial
> Err:   serial
> Net:   No ethernet found.
> starting USB...
> USB0:   USB EHCI 1.00
> USB1:   USB OHCI 1.0
> USB2:   USB EHCI 1.00
> USB3:   USB OHCI 1.0
> USB4:   USB EHCI 1.00
> USB5:   USB OHCI 1.0
> scanning bus 0 for devices... 1 USB Device(s) found
> scanning bus 2 for devices... 1 USB Device(s) found
> scanning bus 4 for devices... 1 USB Device(s) found
>        scanning usb for storage devices... 0 Storage Device(s) found
> Hit any key to stop autoboot:  0
> switch to partitions #0, OK
> mmc0 is current device
> Scanning mmc 0:1...
> Found /extlinux/extlinux.conf
> Retrieving file: /extlinux/extlinux.conf
> 155 bytes read in 3 ms (49.8 KiB/s)
> 1:      linux-4.17.0-rc3
> Retrieving file: /zImage
> 3817968 bytes read in 175 ms (20.8 MiB/s)
> append: console=ttyS0,115200 earlyprintk root=/dev/mmcblk0p2 rootwait
> Retrieving file: /sun8i-h3-bananapi-m2-plus.dtb
> 15691 bytes read in 4 ms (3.7 MiB/s)
> ## Flattened Device Tree blob at 43000000
>    Booting using the fdt blob at 0x43000000
> EHCI failed to shut down host controller.
> EHCI failed to shut down host controller.
> EHCI failed to shut down host controller.
>    Loading Device Tree to 49ff9000, end 49fffd4a ... OK
>
> Starting kernel ...
>
> [    0.000000] Booting Linux on physical CPU 0x0
> [    0.000000] Linux version 4.15.0-rc6-00415-g2f2b41a952de
> (jagan at jagan-XPS-13-9350) (gcc version 6.3.1 20170109 (Linaro GCC
> 6.3-2017.02)) #1 S8
> [    0.000000] CPU: ARMv7 Processor [410fc075] revision 5 (ARMv7), cr=10c5387d

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Chen-Yu Tsai]

On Wed, Jul 11, 2018 at 11:53 AM, Chen-Yu Tsai <wens@csie.org> wrote:
> On Tue, Jul 10, 2018 at 2:13 PM, Jagan Teki <jagannadh.teki@gmail.com> wrote:
>> On Tue, Jul 10, 2018 at 8:34 AM, Chen-Yu Tsai <wens@csie.org> wrote:
>>> This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
>>> bisect, but v2018.07-rc1 is a working version. From the EHCI error
>>> messages, I'm thinking it might be related to the USB changes lately.
>>> Not sure if any other SoCs (ex. A64) are broken or not at the moment.
>>
>> I'm able to boot with BPI-M2+, where I changed the configurations
>> based on Libre Computer board.
>> - remove/disable ehci0, ohci0
>> - enable ehci1, 2, 3 and ohci1, 2, 3
>> - remove/disable usb_otg node
>> - disable MUSB_GADGET
>
> On BPI-M2+, with Sinovoip_BPI_M2_Plus_defconfig,
> cross-compiled using gcc version 7.3.0 (Debian 7.3.0-20)

OK. I dropped back to GCC 6.4.1 from Linaro's 2017.11 release.
It now works. Will try Linaro's other toolchains later.

ChenYu

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Jagan Teki]

On Wed, Jul 11, 2018 at 12:28 PM, Chen-Yu Tsai <wens@csie.org> wrote:
> On Wed, Jul 11, 2018 at 11:53 AM, Chen-Yu Tsai <wens@csie.org> wrote:
>> On Tue, Jul 10, 2018 at 2:13 PM, Jagan Teki <jagannadh.teki@gmail.com> wrote:
>>> On Tue, Jul 10, 2018 at 8:34 AM, Chen-Yu Tsai <wens@csie.org> wrote:
>>>> This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
>>>> bisect, but v2018.07-rc1 is a working version. From the EHCI error
>>>> messages, I'm thinking it might be related to the USB changes lately.
>>>> Not sure if any other SoCs (ex. A64) are broken or not at the moment.
>>>
>>> I'm able to boot with BPI-M2+, where I changed the configurations
>>> based on Libre Computer board.
>>> - remove/disable ehci0, ohci0
>>> - enable ehci1, 2, 3 and ohci1, 2, 3
>>> - remove/disable usb_otg node
>>> - disable MUSB_GADGET
>>
>> On BPI-M2+, with Sinovoip_BPI_M2_Plus_defconfig,
>> cross-compiled using gcc version 7.3.0 (Debian 7.3.0-20)
>
> OK. I dropped back to GCC 6.4.1 from Linaro's 2017.11 release.
> It now works. Will try Linaro's other toolchains later.

OK, I used Linaro GCC 6.3-2017.02

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Peter Robinson]

On Tue, Jul 10, 2018 at 4:04 AM, Chen-Yu Tsai <wens@csie.org> wrote:
> This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
> bisect, but v2018.07-rc1 is a working version. From the EHCI error
> messages, I'm thinking it might be related to the USB changes lately.
> Not sure if any other SoCs (ex. A64) are broken or not at the moment.

I'm still seeing this issue in 2018.09 RC1, what's the status of a fix for this?

> See log:
>
> U-Boot SPL 2018.07 (Jul 10 2018 - 10:21:03 +0800)
> DRAM: 1024 MiB
> Trying to boot from MMC1
>
>
> U-Boot 2018.07 (Jul 10 2018 - 10:21:03 +0800) Allwinner Technology
>
> CPU:   Allwinner H3 (SUN8I 1680)
> Model: Libre Computer Board ALL-H3-CC H3
> DRAM:  1 GiB
> MMC:   SUNXI SD/MMC: 0
> Loading Environment from FAT... Unable to use mmc 0:1... Failed (-5)
> In:    serial
> Out:   serial
> Err:   serial
> Net:   phy interface0
> eth0: ethernet at 1c30000
> starting USB...
> USB0:   USB EHCI 1.00
> USB1:   USB OHCI 1.0
> USB2:   USB EHCI 1.00
> USB3:   USB OHCI 1.0
> USB4:   USB EHCI 1.00
> USB5:   USB OHCI 1.0
> scanning bus 0 for devices... 1 USB Device(s) found
> scanning bus 2 for devices... 1 USB Device(s) found
> scanning bus 4 for devices... 1 USB Device(s) found
>        scanning usb for storage devices... 0 Storage Device(s) found
> Hit any key to stop autoboot:  0
> switch to partitions #0, OK
> mmc0 is current device
> Scanning mmc 0:1...
> Found U-Boot script /boot/boot.scr
> 385 bytes read in 8 ms (46.9 KiB/s)
> ## Executing script at 43100000
> 6343768 bytes read in 295 ms (20.5 MiB/s)
> 18710 bytes read in 10 ms (1.8 MiB/s)
> ## Flattened Device Tree blob at 43000000
>    Booting using the fdt blob at 0x43000000
> EHCI failed to shut down host controller.
> EHCI failed to shut down host controller.
> EHCI failed to shut down host controller.
>    Loading Device Tree to 49ff8000, end 49fff915 ... OK
>
> Starting kernel ...
>
> data abort
> pc : [<7dfadc6c>]          lr : [<7df82ed7>]
> reloc pc : [<4a02bc6c>]    lr : [<4a000ed7>]
> sp : 79f5ce90  ip : 79f62b8a     fp : 00000003
> r10: 7dfc3d14  r9 : 79f61ed8     r8 : 00000400
> r7 : 00000000  r6 : 42000000     r5 : 49ff8000  r4 : 00000000
> r3 : 49ff8000  r2 : 49ff8000     r1 : 00000000  r0 : 00000000
> Flags: nZCv  IRQs off  FIQs off  Mode UK6_32
> Code: e12fff1e e52de008 fa000001 e3a00000 (e49df008)
> Resetting CPU ...
>
> resetting ...
> _______________________________________________
> U-Boot mailing list
> U-Boot at lists.denx.de
> https://lists.denx.de/listinfo/u-boot

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Chen-Yu Tsai]

Hi,

On Wed, Aug 1, 2018 at 9:22 PM Peter Robinson <pbrobinson@gmail.com> wrote:
>
> On Tue, Jul 10, 2018 at 4:04 AM, Chen-Yu Tsai <wens@csie.org> wrote:
> > This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
> > bisect, but v2018.07-rc1 is a working version. From the EHCI error
> > messages, I'm thinking it might be related to the USB changes lately.
> > Not sure if any other SoCs (ex. A64) are broken or not at the moment.
>
> I'm still seeing this issue in 2018.09 RC1, what's the status of a fix for this?

This is a bug in binutils: https://sourceware.org/bugzilla/show_bug.cgi?id=23571
It has been fixed upstream.

> > See log:
> >
> > U-Boot SPL 2018.07 (Jul 10 2018 - 10:21:03 +0800)
> > DRAM: 1024 MiB
> > Trying to boot from MMC1
> >
> >
> > U-Boot 2018.07 (Jul 10 2018 - 10:21:03 +0800) Allwinner Technology
> >
> > CPU:   Allwinner H3 (SUN8I 1680)
> > Model: Libre Computer Board ALL-H3-CC H3
> > DRAM:  1 GiB
> > MMC:   SUNXI SD/MMC: 0
> > Loading Environment from FAT... Unable to use mmc 0:1... Failed (-5)
> > In:    serial
> > Out:   serial
> > Err:   serial
> > Net:   phy interface0
> > eth0: ethernet at 1c30000
> > starting USB...
> > USB0:   USB EHCI 1.00
> > USB1:   USB OHCI 1.0
> > USB2:   USB EHCI 1.00
> > USB3:   USB OHCI 1.0
> > USB4:   USB EHCI 1.00
> > USB5:   USB OHCI 1.0
> > scanning bus 0 for devices... 1 USB Device(s) found
> > scanning bus 2 for devices... 1 USB Device(s) found
> > scanning bus 4 for devices... 1 USB Device(s) found
> >        scanning usb for storage devices... 0 Storage Device(s) found
> > Hit any key to stop autoboot:  0
> > switch to partitions #0, OK
> > mmc0 is current device
> > Scanning mmc 0:1...
> > Found U-Boot script /boot/boot.scr
> > 385 bytes read in 8 ms (46.9 KiB/s)
> > ## Executing script at 43100000
> > 6343768 bytes read in 295 ms (20.5 MiB/s)
> > 18710 bytes read in 10 ms (1.8 MiB/s)
> > ## Flattened Device Tree blob at 43000000
> >    Booting using the fdt blob at 0x43000000
> > EHCI failed to shut down host controller.
> > EHCI failed to shut down host controller.
> > EHCI failed to shut down host controller.
> >    Loading Device Tree to 49ff8000, end 49fff915 ... OK
> >
> > Starting kernel ...
> >
> > data abort
> > pc : [<7dfadc6c>]          lr : [<7df82ed7>]
> > reloc pc : [<4a02bc6c>]    lr : [<4a000ed7>]
> > sp : 79f5ce90  ip : 79f62b8a     fp : 00000003
> > r10: 7dfc3d14  r9 : 79f61ed8     r8 : 00000400
> > r7 : 00000000  r6 : 42000000     r5 : 49ff8000  r4 : 00000000
> > r3 : 49ff8000  r2 : 49ff8000     r1 : 00000000  r0 : 00000000
> > Flags: nZCv  IRQs off  FIQs off  Mode UK6_32
> > Code: e12fff1e e52de008 fa000001 e3a00000 (e49df008)
> > Resetting CPU ...
> >
> > resetting ...
> > _______________________________________________
> > U-Boot mailing list
> > U-Boot at lists.denx.de
> > https://lists.denx.de/listinfo/u-boot

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Peter Robinson]

> > On Tue, Jul 10, 2018 at 4:04 AM, Chen-Yu Tsai <wens@csie.org> wrote:
> > > This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
> > > bisect, but v2018.07-rc1 is a working version. From the EHCI error
> > > messages, I'm thinking it might be related to the USB changes lately.
> > > Not sure if any other SoCs (ex. A64) are broken or not at the moment.
> >
> > I'm still seeing this issue in 2018.09 RC1, what's the status of a fix for this?
>
> This is a bug in binutils: https://sourceware.org/bugzilla/show_bug.cgi?id=23571
> It has been fixed upstream.

For reference the Fedora binutils maintainer (and also one of the
upstream developers made the following note about the fix:

https://bugzilla.redhat.com/show_bug.cgi?id=1624751#c3
---
It should also be noted that this problem can be avoided in the U-Boot linker
script if the align expression is put on the left of the colon.  Like this:

        .__secure_start
  #ifndef CONFIG_ARMV7_SECURE_BASE
                ALIGN(CONSTANT(COMMONPAGESIZE))
  #endif
        : {
                KEEP(*(.__secure_start))
        }

This has the advantage that the solution will work with older versions of the
binutils package.
---

So you may wish to fix it like that so it's not dependent on versions
of binutils.

Peter

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Chen-Yu Tsai]

On Tue, Sep 4, 2018 at 8:22 PM Peter Robinson <pbrobinson@gmail.com> wrote:
>
> > > On Tue, Jul 10, 2018 at 4:04 AM, Chen-Yu Tsai <wens@csie.org> wrote:
> > > > This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
> > > > bisect, but v2018.07-rc1 is a working version. From the EHCI error
> > > > messages, I'm thinking it might be related to the USB changes lately.
> > > > Not sure if any other SoCs (ex. A64) are broken or not at the moment.
> > >
> > > I'm still seeing this issue in 2018.09 RC1, what's the status of a fix for this?
> >
> > This is a bug in binutils: https://sourceware.org/bugzilla/show_bug.cgi?id=23571
> > It has been fixed upstream.
>
> For reference the Fedora binutils maintainer (and also one of the
> upstream developers made the following note about the fix:
>
> https://bugzilla.redhat.com/show_bug.cgi?id=1624751#c3
> ---
> It should also be noted that this problem can be avoided in the U-Boot linker
> script if the align expression is put on the left of the colon.  Like this:
>
>         .__secure_start
>   #ifndef CONFIG_ARMV7_SECURE_BASE
>                 ALIGN(CONSTANT(COMMONPAGESIZE))
>   #endif
>         : {
>                 KEEP(*(.__secure_start))
>         }
>
> This has the advantage that the solution will work with older versions of the
> binutils package.

Thanks. Apparently ALIGN and other commands should come before the colon.
I suppose the linker just ignores commands in incorrect places, or just
tries its best to comply. Anyway, I'm the one that got it wrong the first
time around.

> ---
>
> So you may wish to fix it like that so it's not dependent on versions
> of binutils.

Will do.

ChenYu

[U-Boot] U-Boot 2018.07 Still Broken for Allwinner H3 SoCs

[Chen-Yu Tsai]

On Wed, Sep 5, 2018 at 11:23 AM Chen-Yu Tsai <wens@csie.org> wrote:
>
> On Tue, Sep 4, 2018 at 8:22 PM Peter Robinson <pbrobinson@gmail.com> wrote:
> >
> > > > On Tue, Jul 10, 2018 at 4:04 AM, Chen-Yu Tsai <wens@csie.org> wrote:
> > > > > This is on a Libre Computer ALL-H3-CC H3 variant. Still running a
> > > > > bisect, but v2018.07-rc1 is a working version. From the EHCI error
> > > > > messages, I'm thinking it might be related to the USB changes lately.
> > > > > Not sure if any other SoCs (ex. A64) are broken or not at the moment.
> > > >
> > > > I'm still seeing this issue in 2018.09 RC1, what's the status of a fix for this?
> > >
> > > This is a bug in binutils: https://sourceware.org/bugzilla/show_bug.cgi?id=23571
> > > It has been fixed upstream.
> >
> > For reference the Fedora binutils maintainer (and also one of the
> > upstream developers made the following note about the fix:
> >
> > https://bugzilla.redhat.com/show_bug.cgi?id=1624751#c3
> > ---
> > It should also be noted that this problem can be avoided in the U-Boot linker
> > script if the align expression is put on the left of the colon.  Like this:
> >
> >         .__secure_start
> >   #ifndef CONFIG_ARMV7_SECURE_BASE
> >                 ALIGN(CONSTANT(COMMONPAGESIZE))
> >   #endif
> >         : {
> >                 KEEP(*(.__secure_start))
> >         }
> >
> > This has the advantage that the solution will work with older versions of the
> > binutils package.
>
> Thanks. Apparently ALIGN and other commands should come before the colon.
> I suppose the linker just ignores commands in incorrect places, or just
> tries its best to comply. Anyway, I'm the one that got it wrong the first
> time around.

More on this:

According to the ld documents, ALIGN before the colon is an expression
specifying an address for the section [1], while ALIGN after the colon is
a section attribute for a forced alignment [2].

A third way to accomplish our goal, have the __secure_start symbol within
the .__secure_start section aligned, is to have

    . = ALIGN(CONSTANT(COMMONPAGESIZE));

before

    KEEP(*(.__secure_start))

This is used to align the __secure_stack_end symbol [3].

All these methods accomplish the goal, but have subtle differences. In
particular, the third method aligns the symbols themselves, but not the
start of the section.

ChenYu

[1] https://sourceware.org/binutils/docs-2.31/ld/Output-Section-Address.html
[2] https://sourceware.org/binutils/docs-2.31/ld/Output-Section-Attributes.html
[3] https://elixir.bootlin.com/u-boot/latest/source/arch/arm/cpu/u-boot.lds#L91

> > ---
> >
> > So you may wish to fix it like that so it's not dependent on versions
> > of binutils.
>
> Will do.
>
> ChenYu