[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Abraham V.]

Hello everyone,

I've got a copy of mainline uboot. Here's the last log message in it;

> git log --name-status HEAD^..HEAD
commit d19ad726bcd5d9106f7ba9c750462fcc369f1020
Author: Tom Rini <trini@ti.com>
Date:   Mon Nov 25 16:49:32 2013 -0500

    Prepare v2014.01-rc1

    Signed-off-by: Tom Rini <trini@ti.com>

M       Makefile
>


And well, I'm having trouble building images for the sabreauto board. Here
are the steps I follow,

> make distclean
> make mx6qsabreauto_config
> make

And this is the error I get;

make[1]: Leaving directory `/home/abraham/SPACE/temp_uboot/arch/arm/cpu'
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-gcc
-E -g  -Os   -ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float  -D__KERNEL__ -DCONFIG_SYS_TEXT_BASE=0x17800000
 -I/home/abraham/SPACE/temp_uboot/include
 -I/home/abraham/SPACE/temp_uboot/arch/arm/include -fno-builtin
-ffreestanding -nostdinc -isystem
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/include
-pipe  -DCONFIG_ARM -D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux
-mword-relocations -march=armv7-a -include
/home/abraham/SPACE/temp_uboot/include/u-boot/u-boot.lds.h
-DCPUDIR=arch/arm/cpu/armv7  -ansi -D__ASSEMBLY__ -P -
</home/abraham/SPACE/temp_uboot/arch/arm/cpu/u-boot.lds >u-boot.lds
cd /home/abraham/SPACE/temp_uboot &&
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd
 -pie -T u-boot.lds --gc-sections -Bstatic -Ttext 0x17800000
arch/arm/cpu/armv7/start.o --start-group arch/arm/cpu/armv7/built-in.o
arch/arm/cpu/armv7/mx6/built-in.o arch/arm/cpu/built-in.o
arch/arm/imx-common/built-in.o arch/arm/lib/built-in.o
board/freescale/common/built-in.o board/freescale/mx6qsabreauto/built-in.o
common/built-in.o disk/built-in.o drivers/built-in.o drivers/dma/built-in.o
drivers/gpio/built-in.o drivers/i2c/built-in.o drivers/input/built-in.o
drivers/mmc/built-in.o drivers/mtd/built-in.o drivers/mtd/nand/built-in.o
drivers/mtd/onenand/built-in.o drivers/mtd/spi/built-in.o
drivers/mtd/ubi/built-in.o drivers/net/built-in.o
drivers/net/phy/built-in.o drivers/pci/built-in.o
drivers/power/battery/built-in.o drivers/power/built-in.o
drivers/power/fuel_gauge/built-in.o drivers/power/mfd/built-in.o
drivers/power/pmic/built-in.o drivers/serial/built-in.o
drivers/spi/built-in.o drivers/usb/eth/built-in.o
drivers/usb/gadget/built-in.o drivers/usb/host/built-in.o
drivers/usb/musb-new/built-in.o drivers/usb/musb/built-in.o
drivers/usb/phy/built-in.o drivers/usb/ulpi/built-in.o fs/built-in.o
lib/built-in.o lib/libfdt/built-in.o net/built-in.o post/built-in.o
test/built-in.o --end-group
/home/abraham/SPACE/temp_uboot/arch/arm/lib/eabi_compat.o  -L
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2
-lgcc -Map u-boot.map -o u-boot
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
error:
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(bpabi.o)
uses VFP register arguments, u-boot does not
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
failed to merge target specific data of file
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(bpabi.o)
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
error:
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(_divdi3.o)
uses VFP register arguments, u-boot does not
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
failed to merge target specific data of file
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(_divdi3.o)
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
error:
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(_udivdi3.o)
uses VFP register arguments, u-boot does not
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
failed to merge target specific data of file
/home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(_udivdi3.o)
make: *** [u-boot] Error 1
>

Now, what bothers me is NOT the error, but how I resolved it. In the above
run, I've been using linaro's cross-compiler. If I switch over to using the
toolchain from Mentor Graphics (code-sorcery free version), the build
compiles without issues (I can even boot the resulting image on a board!).

Here are details of the cross-compilers I've been using,

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Stefano Babic]

Hi Abraham,

On 28/11/2013 05:22, Abraham V. wrote:

> test/built-in.o --end-group
> /home/abraham/SPACE/temp_uboot/arch/arm/lib/eabi_compat.o  -L
> /home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2
> -lgcc -Map u-boot.map -o u-boot
> /home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
> error:
> /home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(bpabi.o)
> uses VFP register arguments, u-boot does not

The most evident cause here is that you use a -hf toolchain with support
for hard float point. U-Boot does not support it: there is no need of it.

This is the reason why it works using Mentor's toolchain. Even if you
use the Linaro's supplied with Ubuntu, it works.

Best regards,
Stefano Babic

-- 
=====================================================================
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-53 Fax: +49-8142-66989-80 Email: sbabic at denx.de
=====================================================================

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Abraham V.]

Thanks Stefano,

I didn't notice that difference until you pointed it out. Till now, I was
under the assumption that both toolchains were (feature-wise) identical.
And I *almost* sent a complaint letter to Linaro about this too ... Guess,
I've got more to learn than I anticipated.

Sincerely,
Abraham V.



On Thu, Nov 28, 2013 at 12:52 PM, Stefano Babic <sbabic@denx.de> wrote:

> Hi Abraham,
>
> On 28/11/2013 05:22, Abraham V. wrote:
>
> > test/built-in.o --end-group
> > /home/abraham/SPACE/temp_uboot/arch/arm/lib/eabi_compat.o  -L
> >
> /home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2
> > -lgcc -Map u-boot.map -o u-boot
> >
> /home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/arm-linux-gnueabihf-ld.bfd:
> > error:
> >
> /home/abraham/SPACE/BISQUARE/source/uboot_update/gcc-linaro-arm/bin/../lib/gcc/arm-linux-gnueabihf/4.8.2/libgcc.a(bpabi.o)
> > uses VFP register arguments, u-boot does not
>
> The most evident cause here is that you use a -hf toolchain with support
> for hard float point. U-Boot does not support it: there is no need of it.
>
> This is the reason why it works using Mentor's toolchain. Even if you
> use the Linaro's supplied with Ubuntu, it works.
>
> Best regards,
> Stefano Babic
>
> --
> =====================================================================
> DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
> HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
> Phone: +49-8142-66989-53 Fax: +49-8142-66989-80 Email: sbabic at denx.de
> =====================================================================
>

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Wolfgang Denk]

Dear "Abraham V.",

In message <CANiE1qo9fDOkeVjq89W3cEoxRyic9d_DO71vv9BRtg7bnJ-jnQ@mail.gmail.com> you wrote:
>
> I didn't notice that difference until you pointed it out. Till now, I was
> under the assumption that both toolchains were (feature-wise) identical.
> And I *almost* sent a complaint letter to Linaro about this too ... Guess,
> I've got more to learn than I anticipated.

I think feature-wise both configurations are probably identical; but
for more general usability the -hf configuration should also include a
non-hf version of the libgcc library, so that building with
"-msoft-float" actually works.

Alternatively, you can always use USE_PRIVATE_LIBGCC=yes on the make
command line to use U-Boot's builtin libgcc code.

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
The only person who always got his work done by Friday
                                                 was Robinson Crusoe.

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Christian Gmeiner]

2013/11/29 Wolfgang Denk <wd@denx.de>:
> Dear "Abraham V.",
>
> In message <CANiE1qo9fDOkeVjq89W3cEoxRyic9d_DO71vv9BRtg7bnJ-jnQ@mail.gmail.com> you wrote:
>>
>> I didn't notice that difference until you pointed it out. Till now, I was
>> under the assumption that both toolchains were (feature-wise) identical.
>> And I *almost* sent a complaint letter to Linaro about this too ... Guess,
>> I've got more to learn than I anticipated.
>

I am running into the same problem. Me wonders which patch/change
triggers this problem. I am using
the same toolchain since I started to work with u-boot and it fails
with the latest rc.


> I think feature-wise both configurations are probably identical; but
> for more general usability the -hf configuration should also include a
> non-hf version of the libgcc library, so that building with
> "-msoft-float" actually works.
>

Will try this one.

greets
--
Christian Gmeiner, MSc

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Wolfgang Denk]

Dear Christian Gmeiner,

In message <CAH9NwWeb6s+3s4O25D7Cifob9Y3_7jmHoSjO6o78FijQuWZMwg@mail.gmail.com> you wrote:
>
> I am running into the same problem. Me wonders which patch/change
> triggers this problem. I am using
> the same toolchain since I started to work with u-boot and it fails
> with the latest rc.

Maybe Linaro started using a hard-float configuration only recently?

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
It is common sense to take a method and try it. If it fails, admit it
frankly and try another. But above all, try something.
                                              - Franklin D. Roosevelt

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Robert Nelson]

On Fri, Nov 29, 2013 at 4:28 PM, Wolfgang Denk <wd@denx.de> wrote:
> Dear Christian Gmeiner,
>
> In message <CAH9NwWeb6s+3s4O25D7Cifob9Y3_7jmHoSjO6o78FijQuWZMwg@mail.gmail.com> you wrote:
>>
>> I am running into the same problem. Me wonders which patch/change
>> triggers this problem. I am using
>> the same toolchain since I started to work with u-boot and it fails
>> with the latest rc.
>
> Maybe Linaro started using a hard-float configuration only recently?
>

For what it's worth, i just ran a git bisect on this issue today, as
this compiler worked fine with v2013.10

toolchain:
arm-linux-gnueabihf-gcc (crosstool-NG linaro-1.13.1-4.8-2013.10 -
Linaro GCC 2013.10) 4.8.2 20131014 (prerelease)

762a88ccf8540948fbf8c31b40a29d1e0684a25b is the first bad commit
commit 762a88ccf8540948fbf8c31b40a29d1e0684a25b
Author: Pierre Aubert <p.aubert@staubli.com>
Date:   Thu Sep 19 17:48:59 2013 +0200

    mx6: compute PLL PFD frequencies rather than using defines

    Signed-off-by: Pierre Aubert <p.aubert@staubli.com>
    CC: Stefano Babic <sbabic@denx.de>

:040000 040000 c3dba1387015304c14486872c9776d008d9a3b03
1f16de9cee07af71e12f1a746dcd962c9eb84bb0 M      arch

git bisect log
git bisect start
# good: [183acb700378a8cfc5d50a01a65de93fb2c24586] Prepare v2013.10
git bisect good 183acb700378a8cfc5d50a01a65de93fb2c24586
# bad: [d19ad726bcd5d9106f7ba9c750462fcc369f1020] Prepare v2014.01-rc1
git bisect bad d19ad726bcd5d9106f7ba9c750462fcc369f1020
# good: [65ba7add0d609bbd035b8d42fafdaf428ac24751] time: add weak
annotation to timer_read_counter declaration
git bisect good 65ba7add0d609bbd035b8d42fafdaf428ac24751
# bad: [c3ebb8c38a5da5e40da2786c5d850d1f6555ff95] Merge branch
'master' of git://git.denx.de/u-boot-mpc85xx
git bisect bad c3ebb8c38a5da5e40da2786c5d850d1f6555ff95
# bad: [3285d4ca197928a048d3dda86751b5d26e6e0e86] Merge branch
'u-boot-imx/master' into 'u-boot-arm/master'
git bisect bad 3285d4ca197928a048d3dda86751b5d26e6e0e86
# good: [552998e5f7655ba8fd3acff400012bd651edff06] TI:armv7: Change
CONFIG_SYS_SPL_ARGS_ADDR to a higher address
git bisect good 552998e5f7655ba8fd3acff400012bd651edff06
# bad: [195d130da1c2705f96571b7265312bdfa8a83fee] Revert "configs:
imx: Make CONFIG_SYS_PROMPT uniform across FSL boards"
git bisect bad 195d130da1c2705f96571b7265312bdfa8a83fee
# bad: [4867b634b7c0e5ede258b4998fa4b2710e7daacf] ARM: mx5: Enable L2 cache
git bisect bad 4867b634b7c0e5ede258b4998fa4b2710e7daacf
# bad: [0c5e26678b18e136c1514bf769a16060ae1b5ff8] udoo: Add initial
support for mx6q udoo board
git bisect bad 0c5e26678b18e136c1514bf769a16060ae1b5ff8
# bad: [6654f33c9b520bd4073c7f82a13044e79bc14898] ARM: mxs: tools: Use
mkimage for BootStream generation
git bisect bad 6654f33c9b520bd4073c7f82a13044e79bc14898
# bad: [762a88ccf8540948fbf8c31b40a29d1e0684a25b] mx6: compute PLL PFD
frequencies rather than using defines
git bisect bad 762a88ccf8540948fbf8c31b40a29d1e0684a25b
# first bad commit: [762a88ccf8540948fbf8c31b40a29d1e0684a25b] mx6:
compute PLL PFD frequencies rather than using defines

using USE_PRIVATE_LIBGCC=yes causes another error..

arch/arm/cpu/armv7/mx6/built-in.o: In function `mxc_get_pll_pfd':
/opt/github/u-boot/arch/arm/cpu/armv7/mx6/clock.c:126: undefined
reference to `__aeabi_uldivmod'
/opt/gcc/gcc-linaro-arm-linux-gnueabihf-4.8-2013.10_linux/bin/arm-linux-gnueabihf-ld.bfd:
BFD (crosstool-NG linaro-1.13.1-4.8-2013.10 - Linaro GCC 2013.10)
2.23.2.20130610 Linaro 2013.10-4 assertion fail
/cbuild/slaves/oorts/crosstool-ng/builds/arm-linux-gnueabihf-linux/.build/src/binutils-linaro-2.23.2-2013.10-4/bfd/elf32-arm.c:7687
/opt/gcc/gcc-linaro-arm-linux-gnueabihf-4.8-2013.10_linux/bin/arm-linux-gnueabihf-ld.bfd:
error: required section '.rel.plt' not found in the linker script
/opt/gcc/gcc-linaro-arm-linux-gnueabihf-4.8-2013.10_linux/bin/arm-linux-gnueabihf-ld.bfd:
final link failed: Invalid operation
make: *** [u-boot] Error 1

Regards,

-- 
Robert Nelson
http://www.rcn-ee.com/

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Måns Rullgård]

Robert Nelson <robertcnelson@gmail.com> writes:

> On Fri, Nov 29, 2013 at 4:28 PM, Wolfgang Denk <wd@denx.de> wrote:
>> Dear Christian Gmeiner,
>>
>> In message <CAH9NwWeb6s+3s4O25D7Cifob9Y3_7jmHoSjO6o78FijQuWZMwg@mail.gmail.com> you wrote:
>>>
>>> I am running into the same problem. Me wonders which patch/change
>>> triggers this problem. I am using
>>> the same toolchain since I started to work with u-boot and it fails
>>> with the latest rc.
>>
>> Maybe Linaro started using a hard-float configuration only recently?
>>
>
> For what it's worth, i just ran a git bisect on this issue today, as
> this compiler worked fine with v2013.10
>
> toolchain:
> arm-linux-gnueabihf-gcc (crosstool-NG linaro-1.13.1-4.8-2013.10 -
> Linaro GCC 2013.10) 4.8.2 20131014 (prerelease)
>
> 762a88ccf8540948fbf8c31b40a29d1e0684a25b is the first bad commit
> commit 762a88ccf8540948fbf8c31b40a29d1e0684a25b
> Author: Pierre Aubert <p.aubert@staubli.com>
> Date:   Thu Sep 19 17:48:59 2013 +0200
>
>     mx6: compute PLL PFD frequencies rather than using defines

That commit introduces a 64-bit division without using the lldiv()
function, which pulls in previously unused libgcc stuff.

Try this patch:

diff --git a/arch/arm/cpu/armv7/mx6/clock.c b/arch/arm/cpu/armv7/mx6/clock.c
index 873d9d0..752b3c8 100644
--- a/arch/arm/cpu/armv7/mx6/clock.c
+++ b/arch/arm/cpu/armv7/mx6/clock.c
@@ -5,6 +5,7 @@
  */
 
 #include <common.h>
+#include <div64.h>
 #include <asm/io.h>
 #include <asm/errno.h>
 #include <asm/arch/imx-regs.h>
@@ -123,8 +124,8 @@ static u32 mxc_get_pll_pfd(enum pll_clocks pll, int pfd_num)
                return 0;
        }
 
-       return (freq * 18) / ((div & ANATOP_PFD_FRAC_MASK(pfd_num)) >>
-                             ANATOP_PFD_FRAC_SHIFT(pfd_num));
+       return lldiv(freq * 18, (div & ANATOP_PFD_FRAC_MASK(pfd_num)) >>
+                               ANATOP_PFD_FRAC_SHIFT(pfd_num));
 }
 
 static u32 get_mcu_main_clk(void)

-- 
M?ns Rullg?rd
mans at mansr.com

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Robert Nelson]

On Wed, Dec 4, 2013 at 9:35 AM, M?ns Rullg?rd <mans@mansr.com> wrote:
> Robert Nelson <robertcnelson@gmail.com> writes:
>
>> On Fri, Nov 29, 2013 at 4:28 PM, Wolfgang Denk <wd@denx.de> wrote:
>>> Dear Christian Gmeiner,
>>>
>>> In message <CAH9NwWeb6s+3s4O25D7Cifob9Y3_7jmHoSjO6o78FijQuWZMwg@mail.gmail.com> you wrote:
>>>>
>>>> I am running into the same problem. Me wonders which patch/change
>>>> triggers this problem. I am using
>>>> the same toolchain since I started to work with u-boot and it fails
>>>> with the latest rc.
>>>
>>> Maybe Linaro started using a hard-float configuration only recently?
>>>
>>
>> For what it's worth, i just ran a git bisect on this issue today, as
>> this compiler worked fine with v2013.10
>>
>> toolchain:
>> arm-linux-gnueabihf-gcc (crosstool-NG linaro-1.13.1-4.8-2013.10 -
>> Linaro GCC 2013.10) 4.8.2 20131014 (prerelease)
>>
>> 762a88ccf8540948fbf8c31b40a29d1e0684a25b is the first bad commit
>> commit 762a88ccf8540948fbf8c31b40a29d1e0684a25b
>> Author: Pierre Aubert <p.aubert@staubli.com>
>> Date:   Thu Sep 19 17:48:59 2013 +0200
>>
>>     mx6: compute PLL PFD frequencies rather than using defines
>
> That commit introduces a 64-bit division without using the lldiv()
> function, which pulls in previously unused libgcc stuff.
>
> Try this patch:
>
> diff --git a/arch/arm/cpu/armv7/mx6/clock.c b/arch/arm/cpu/armv7/mx6/clock.c
> index 873d9d0..752b3c8 100644
> --- a/arch/arm/cpu/armv7/mx6/clock.c
> +++ b/arch/arm/cpu/armv7/mx6/clock.c
> @@ -5,6 +5,7 @@
>   */
>
>  #include <common.h>
> +#include <div64.h>
>  #include <asm/io.h>
>  #include <asm/errno.h>
>  #include <asm/arch/imx-regs.h>
> @@ -123,8 +124,8 @@ static u32 mxc_get_pll_pfd(enum pll_clocks pll, int pfd_num)
>                 return 0;
>         }
>
> -       return (freq * 18) / ((div & ANATOP_PFD_FRAC_MASK(pfd_num)) >>
> -                             ANATOP_PFD_FRAC_SHIFT(pfd_num));
> +       return lldiv(freq * 18, (div & ANATOP_PFD_FRAC_MASK(pfd_num)) >>
> +                               ANATOP_PFD_FRAC_SHIFT(pfd_num));
>  }
>
>  static u32 get_mcu_main_clk(void)

Thanks M?ns!

That fixed it...

Regards,

-- 
Robert Nelson
http://www.rcn-ee.com/

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Pierre Aubert]

Dear M?ns Rullg?rd ,

On Wed, Dec 4, 2013 at 9:35 AM, M?ns Rullg?rd <[hidden email]> wrote: 

>>     mx6: compute PLL PFD frequencies rather than using defines
>
> That commit introduces a 64-bit division without using the lldiv()
> function, which pulls in previously unused libgcc stuff.

Thank you for the fix. I missed the 64 bits division.

Best regards




--
View this message in context: http://u-boot.10912.n7.nabble.com/toolchain-problems-when-building-iMX6-mx6qsabreauto-and-another-bootloader-tool-tp168514p169107.html
Sent from the U-Boot mailing list archive at Nabble.com.

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Christian Gmeiner]

2013/12/4 M?ns Rullg?rd <mans@mansr.com>:
> Robert Nelson <robertcnelson@gmail.com> writes:
>
>> On Fri, Nov 29, 2013 at 4:28 PM, Wolfgang Denk <wd@denx.de> wrote:
>>> Dear Christian Gmeiner,
>>>
>>> In message <CAH9NwWeb6s+3s4O25D7Cifob9Y3_7jmHoSjO6o78FijQuWZMwg@mail.gmail.com> you wrote:
>>>>
>>>> I am running into the same problem. Me wonders which patch/change
>>>> triggers this problem. I am using
>>>> the same toolchain since I started to work with u-boot and it fails
>>>> with the latest rc.
>>>
>>> Maybe Linaro started using a hard-float configuration only recently?
>>>
>>
>> For what it's worth, i just ran a git bisect on this issue today, as
>> this compiler worked fine with v2013.10
>>
>> toolchain:
>> arm-linux-gnueabihf-gcc (crosstool-NG linaro-1.13.1-4.8-2013.10 -
>> Linaro GCC 2013.10) 4.8.2 20131014 (prerelease)
>>
>> 762a88ccf8540948fbf8c31b40a29d1e0684a25b is the first bad commit
>> commit 762a88ccf8540948fbf8c31b40a29d1e0684a25b
>> Author: Pierre Aubert <p.aubert@staubli.com>
>> Date:   Thu Sep 19 17:48:59 2013 +0200
>>
>>     mx6: compute PLL PFD frequencies rather than using defines
>
> That commit introduces a 64-bit division without using the lldiv()
> function, which pulls in previously unused libgcc stuff.
>
> Try this patch:
>
> diff --git a/arch/arm/cpu/armv7/mx6/clock.c b/arch/arm/cpu/armv7/mx6/clock.c
> index 873d9d0..752b3c8 100644
> --- a/arch/arm/cpu/armv7/mx6/clock.c
> +++ b/arch/arm/cpu/armv7/mx6/clock.c
> @@ -5,6 +5,7 @@
>   */
>
>  #include <common.h>
> +#include <div64.h>
>  #include <asm/io.h>
>  #include <asm/errno.h>
>  #include <asm/arch/imx-regs.h>
> @@ -123,8 +124,8 @@ static u32 mxc_get_pll_pfd(enum pll_clocks pll, int pfd_num)
>                 return 0;
>         }
>
> -       return (freq * 18) / ((div & ANATOP_PFD_FRAC_MASK(pfd_num)) >>
> -                             ANATOP_PFD_FRAC_SHIFT(pfd_num));
> +       return lldiv(freq * 18, (div & ANATOP_PFD_FRAC_MASK(pfd_num)) >>
> +                               ANATOP_PFD_FRAC_SHIFT(pfd_num));
>  }
>
>  static u32 get_mcu_main_clk(void)
>
> --

works here too... can we get this patch into git asap?

thanks
--
Christian Gmeiner, MSc

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Stefano Babic]

Hi M?ns,

On 04/12/2013 16:35, M?ns Rullg?rd wrote:
>>
>>     mx6: compute PLL PFD frequencies rather than using defines
> 
> That commit introduces a 64-bit division without using the lldiv()
> function, which pulls in previously unused libgcc stuff.
> 
> Try this patch:

As it was proofed that your patch fixes the issue, can you send aagain
the patch using git send-email and adding your Signed-off-by for merging
into the release ?

Thanks,
Stefano Babic

-- 
=====================================================================
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-53 Fax: +49-8142-66989-80 Email: sbabic at denx.de
=====================================================================

[U-Boot] toolchain problems when building iMX6 mx6qsabreauto (and another bootloader tool)

[Tom Rini]

On Fri, Nov 29, 2013 at 07:27:29AM +0100, Wolfgang Denk wrote:
> Dear "Abraham V.",
> 
> In message <CANiE1qo9fDOkeVjq89W3cEoxRyic9d_DO71vv9BRtg7bnJ-jnQ@mail.gmail.com> you wrote:
> >
> > I didn't notice that difference until you pointed it out. Till now, I was
> > under the assumption that both toolchains were (feature-wise) identical.
> > And I *almost* sent a complaint letter to Linaro about this too ... Guess,
> > I've got more to learn than I anticipated.
> 
> I think feature-wise both configurations are probably identical; but
> for more general usability the -hf configuration should also include a
> non-hf version of the libgcc library, so that building with
> "-msoft-float" actually works.
> 
> Alternatively, you can always use USE_PRIVATE_LIBGCC=yes on the make
> command line to use U-Boot's builtin libgcc code.

Well, herp-a-derp.  This is the jffs2 problem as well, which boils down
to a now 64bit division that needs to be lldiv().  Single board
build-tested now for that, testing the rest before posting.

-- 
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20131204/fd16e58e/attachment.pgp>