[Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Karicheri, Murali via buildroot]

Hi All,

We were using buildroot version 2023.08 version and encountered a year 2038 crash when using time/date utilities. So, my research found some work done in the recent version of buildroot. Where BR2_TIME_BITS_64=y was added to build with 64bit time field. My board is based on a STM32MP133c SoC and using an external project to build buildroot. So, I have migrated our project to using 2024.02.x version of buildroot and enabled BR2_TIME_BITS_64=y to get the fix for the 2038 issue. But my build fails with the logs below.

Could you help me answer the following questions?
1. Is the Y2038 time issue resolved in buildroot? If not, is there some work, though officially not supported,  I can investigate for a potential solution
2. If it is in fact resolved and expected to work in 2024.02.x, what I might be missing? Is there specific version of the toolchain packages I need to select in the Toolchain configuration and other packages?

libtool: compile:  /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc/xgcc -shared-libgcc -B/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc -nostdinc++ -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src/.libs -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/libsupc++/.libs -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/bin/ -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/lib/ -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/include -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/sys-
 include -D_GNU_SOURCE -D_DEBUG -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS -DHAVE_RPC_XDR_H=0 -DHAVE_TIRPC_RPC_XDR_H=0 -I. -I../../../../libsanitizer/sanitizer_common -I.. -I ../../../../libsanitizer/include -I ../../../../libsanitizer -isystem ../../../../libsanitizer/include/system -Wall -W -Wno-unused-parameter -Wwrite-strings -pedantic -Wno-long-long -fPIC -fno-builtin -fno-exceptions -fno-rtti -fomit-frame-pointer -funwind-tables -fvisibility=hidden -Wno-variadic-macros -I../../libstdc++-v3/include -I../../libstdc++-v3/include/arm-buildroot-linux-gnueabihf -I../../../../libsanitizer/../libstdc++-v3/libsupc++ -std=gnu++14 -DSANITIZER_LIBBACKTRACE -DSANITIZER_CP_DEMANGLE -I ../../../../libsanitizer/../libbacktrace -I ../libbacktrace -I ../../../../libsanitizer/../include -include ../../../../libsanitizer/libbacktrace/backtrace-rename.h -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_TIME_BITS=64 -Og -g2 -D_GNU_SOURCE -MT sanitizer_pro
 cmaps_common.lo -MD -MP -MF .deps/sanitizer_procmaps_common.Tpo -c ../../../../libsanitizer/sanitizer_common/sanitizer_procmaps_common.cpp -o sanitizer_procmaps_common.o >/dev/null 2>&1
mv -f .deps/sanitizer_linux_libcdep.Tpo .deps/sanitizer_linux_libcdep.Plo
mv -f .deps/sanitizer_procmaps_linux.Tpo .deps/sanitizer_procmaps_linux.Plo
mv -f .deps/sanitizer_linux.Tpo .deps/sanitizer_linux.Plo
mv -f .deps/sanitizer_printf.Tpo .deps/sanitizer_printf.Plo
mv -f .deps/sanitizer_posix.Tpo .deps/sanitizer_posix.Plo
mv -f .deps/sanitizer_procmaps_common.Tpo .deps/sanitizer_procmaps_common.Plo
mv -f .deps/sanitizer_posix_libcdep.Tpo .deps/sanitizer_posix_libcdep.Plo
make[6]: *** [Makefile:533: all-recursive] Error 1
make[5]: *** [Makefile:420: all] Error 2
make[4]: *** [Makefile:13138: all-target-libsanitizer] Error 2
make[3]: *** [Makefile:1033: all] Error 2
make[2]: *** [package/pkg-generic.mk:283: /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/.stamp_built] Error 2
make[1]: *** [Makefile:82: _all] Error 2
make: *** [Makefile:106: /src/linux-bsp-stm32mp133c/output/development/host/bin/arm-none-linux-gnueabihf-gcc] Error 2


Murali Karicheri
Software Architect
S&C Electric Company
________________________________
NOTICE OF CONFIDENTIALITY:
This message may contain information that is considered confidential and which may be prohibited from disclosure under applicable law or by contractual agreement. The information is intended solely for the use of the individual or entity named above. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution or use of the information contained in or attached to this message is strictly prohibited. If you have received this email transmission in error, please notify the sender by replying to this email and then delete it from your system.
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Thomas Petazzoni via buildroot]

Hello Murali,

On Wed, 17 Jul 2024 18:18:35 +0000
"Karicheri, Murali via buildroot" <buildroot@buildroot.org> wrote:

> We were using buildroot version 2023.08 version and encountered a
> year 2038 crash when using time/date utilities. So, my research found
> some work done in the recent version of buildroot. Where
> BR2_TIME_BITS_64=y was added to build with 64bit time field. My board
> is based on a STM32MP133c SoC and using an external project to build
> buildroot. So, I have migrated our project to using 2024.02.x version
> of buildroot and enabled BR2_TIME_BITS_64=y to get the fix for the
> 2038 issue. But my build fails with the logs below.

Sorry to hear that your build fails with Buildroot 2024.02.x. See below
some details.

> 
> Could you help me answer the following questions?

> 1. Is the Y2038 time issue resolved in buildroot? If not, is there
> some work, though officially not supported,  I can investigate for a
> potential solution

Things are not as black and white as "Is the Y2038 time issue
resolved". How 64-bit time_t is handled depends on the C library, and
also different packages might exhibit different problems. So it's not
like we can absolutely guarantee that everything is fixed.

> 2. If it is in fact resolved and expected to work in 2024.02.x, what
> I might be missing? Is there specific version of the toolchain
> packages I need to select in the Toolchain configuration and other
> packages?

We expect things to work, but we have lots of packages.

> libtool: compile:  /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc/xgcc -shared-libgcc -B/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc -nostdinc++ -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src/.libs -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/libsupc++/.libs -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/bin/ -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/lib/ -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/include -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/sy
 s-
>  include -D_GNU_SOURCE -D_DEBUG -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS -DHAVE_RPC_XDR_H=0 -DHAVE_TIRPC_RPC_XDR_H=0 -I. -I../../../../libsanitizer/sanitizer_common -I.. -I ../../../../libsanitizer/include -I ../../../../libsanitizer -isystem ../../../../libsanitizer/include/system -Wall -W -Wno-unused-parameter -Wwrite-strings -pedantic -Wno-long-long -fPIC -fno-builtin -fno-exceptions -fno-rtti -fomit-frame-pointer -funwind-tables -fvisibility=hidden -Wno-variadic-macros -I../../libstdc++-v3/include -I../../libstdc++-v3/include/arm-buildroot-linux-gnueabihf -I../../../../libsanitizer/../libstdc++-v3/libsupc++ -std=gnu++14 -DSANITIZER_LIBBACKTRACE -DSANITIZER_CP_DEMANGLE -I ../../../../libsanitizer/../libbacktrace -I ../libbacktrace -I ../../../../libsanitizer/../include -include ../../../../libsanitizer/libbacktrace/backtrace-rename.h -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_TIME_BITS=64 -Og -g2 -D_GNU_SOURCE -MT sanitizer_p
 ro
>  cmaps_common.lo -MD -MP -MF .deps/sanitizer_procmaps_common.Tpo -c ../../../../libsanitizer/sanitizer_common/sanitizer_procmaps_common.cpp -o sanitizer_procmaps_common.o >/dev/null 2>&1
> mv -f .deps/sanitizer_linux_libcdep.Tpo .deps/sanitizer_linux_libcdep.Plo
> mv -f .deps/sanitizer_procmaps_linux.Tpo .deps/sanitizer_procmaps_linux.Plo
> mv -f .deps/sanitizer_linux.Tpo .deps/sanitizer_linux.Plo
> mv -f .deps/sanitizer_printf.Tpo .deps/sanitizer_printf.Plo
> mv -f .deps/sanitizer_posix.Tpo .deps/sanitizer_posix.Plo
> mv -f .deps/sanitizer_procmaps_common.Tpo .deps/sanitizer_procmaps_common.Plo
> mv -f .deps/sanitizer_posix_libcdep.Tpo .deps/sanitizer_posix_libcdep.Plo
> make[6]: *** [Makefile:533: all-recursive] Error 1
> make[5]: *** [Makefile:420: all] Error 2
> make[4]: *** [Makefile:13138: all-target-libsanitizer] Error 2
> make[3]: *** [Makefile:1033: all] Error 2
> make[2]: *** [package/pkg-generic.mk:283: /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/.stamp_built] Error 2
> make[1]: *** [Makefile:82: _all] Error 2
> make: *** [Makefile:106: /src/linux-bsp-stm32mp133c/output/development/host/bin/arm-none-linux-gnueabihf-gcc] Error 2

This error log is unfortunately useless, as you didn't copy/paste
enough of the build log to have the actual error. All we can see from
this is that the compilation failed during the host-gcc-final step.

Please provide:

 1. A Buildroot .config file that allows to reproduce the issue with
    upstream Buildroot (without any custom changes)

 2. The full build log, preferably posted on some pastebin site, as the
    mailing list will not accept a mail that contains a full build log.

Best regards,

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Florian Fainelli]

On 7/17/24 13:10, Thomas Petazzoni via buildroot wrote:
> Hello Murali,
> 
> On Wed, 17 Jul 2024 18:18:35 +0000
> "Karicheri, Murali via buildroot" <buildroot@buildroot.org> wrote:
> 
>> We were using buildroot version 2023.08 version and encountered a
>> year 2038 crash when using time/date utilities. So, my research found
>> some work done in the recent version of buildroot. Where
>> BR2_TIME_BITS_64=y was added to build with 64bit time field. My board
>> is based on a STM32MP133c SoC and using an external project to build
>> buildroot. So, I have migrated our project to using 2024.02.x version
>> of buildroot and enabled BR2_TIME_BITS_64=y to get the fix for the
>> 2038 issue. But my build fails with the logs below.
> 
> Sorry to hear that your build fails with Buildroot 2024.02.x. See below
> some details.
> 
>>
>> Could you help me answer the following questions?
> 
>> 1. Is the Y2038 time issue resolved in buildroot? If not, is there
>> some work, though officially not supported,  I can investigate for a
>> potential solution
> 
> Things are not as black and white as "Is the Y2038 time issue
> resolved". How 64-bit time_t is handled depends on the C library, and
> also different packages might exhibit different problems. So it's not
> like we can absolutely guarantee that everything is fixed.
> 
>> 2. If it is in fact resolved and expected to work in 2024.02.x, what
>> I might be missing? Is there specific version of the toolchain
>> packages I need to select in the Toolchain configuration and other
>> packages?
> 
> We expect things to work, but we have lots of packages.
> 
>> libtool: compile:  /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc/xgcc -shared-libgcc -B/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc -nostdinc++ -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src/.libs -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/libsupc++/.libs -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/bin/ -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/lib/ -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/include -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/sy
>   s-
>>   include -D_GNU_SOURCE -D_DEBUG -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS -DHAVE_RPC_XDR_H=0 -DHAVE_TIRPC_RPC_XDR_H=0 -I. -I../../../../libsanitizer/sanitizer_common -I.. -I ../../../../libsanitizer/include -I ../../../../libsanitizer -isystem ../../../../libsanitizer/include/system -Wall -W -Wno-unused-parameter -Wwrite-strings -pedantic -Wno-long-long -fPIC -fno-builtin -fno-exceptions -fno-rtti -fomit-frame-pointer -funwind-tables -fvisibility=hidden -Wno-variadic-macros -I../../libstdc++-v3/include -I../../libstdc++-v3/include/arm-buildroot-linux-gnueabihf -I../../../../libsanitizer/../libstdc++-v3/libsupc++ -std=gnu++14 -DSANITIZER_LIBBACKTRACE -DSANITIZER_CP_DEMANGLE -I ../../../../libsanitizer/../libbacktrace -I ../libbacktrace -I ../../../../libsanitizer/../include -include ../../../../libsanitizer/libbacktrace/backtrace-rename.h -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_TIME_BITS=64 -Og -g2 -D_GNU_SOURCE -MT sanitizer_p
>   ro
>>   cmaps_common.lo -MD -MP -MF .deps/sanitizer_procmaps_common.Tpo -c ../../../../libsanitizer/sanitizer_common/sanitizer_procmaps_common.cpp -o sanitizer_procmaps_common.o >/dev/null 2>&1
>> mv -f .deps/sanitizer_linux_libcdep.Tpo .deps/sanitizer_linux_libcdep.Plo
>> mv -f .deps/sanitizer_procmaps_linux.Tpo .deps/sanitizer_procmaps_linux.Plo
>> mv -f .deps/sanitizer_linux.Tpo .deps/sanitizer_linux.Plo
>> mv -f .deps/sanitizer_printf.Tpo .deps/sanitizer_printf.Plo
>> mv -f .deps/sanitizer_posix.Tpo .deps/sanitizer_posix.Plo
>> mv -f .deps/sanitizer_procmaps_common.Tpo .deps/sanitizer_procmaps_common.Plo
>> mv -f .deps/sanitizer_posix_libcdep.Tpo .deps/sanitizer_posix_libcdep.Plo
>> make[6]: *** [Makefile:533: all-recursive] Error 1
>> make[5]: *** [Makefile:420: all] Error 2
>> make[4]: *** [Makefile:13138: all-target-libsanitizer] Error 2
>> make[3]: *** [Makefile:1033: all] Error 2
>> make[2]: *** [package/pkg-generic.mk:283: /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/.stamp_built] Error 2
>> make[1]: *** [Makefile:82: _all] Error 2
>> make: *** [Makefile:106: /src/linux-bsp-stm32mp133c/output/development/host/bin/arm-none-linux-gnueabihf-gcc] Error 2
> 
> This error log is unfortunately useless, as you didn't copy/paste
> enough of the build log to have the actual error. All we can see from
> this is that the compilation failed during the host-gcc-final step.
> 
> Please provide:
> 
>   1. A Buildroot .config file that allows to reproduce the issue with
>      upstream Buildroot (without any custom changes)
> 
>   2. The full build log, preferably posted on some pastebin site, as the
>      mailing list will not accept a mail that contains a full build log.

This looks similar to what I encountered a few days ago, the actual 
issue was that libsanitizer was referencing scsi/scsi.h (sorry about the 
message being in French, he he):

../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:128:10: 
erreur fatale: scsi/scsi.h : No such file or directory
   128 | #include <scsi/scsi.h>
       |          ^~~~~~~~~~~~~
compilation terminée.
make[6]: *** [Makefile:623: sanitizer_platform_limits_posix.lo] Error 1
make[6]: *** Waiting for unfinished jobs....
Dans le fichier inclus depuis 
../../../../libsanitizer/sanitizer_common/sanitizer_linux.h:22,
                  depuis 
../../../../libsanitizer/sanitizer_common/sanitizer_linux.cpp:24:
../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.h:592:10: 
attention: le C++ ISO interdit les structures anonymes [-Wpedantic]
   592 |   struct {
       |          ^
Dans le fichier inclus depuis 
../../../../libsanitizer/sanitizer_common/sanitizer_posix_libcdep.cpp:21:
../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.h:592:10: 
attention: le C++ ISO interdit les structures anonymes [-Wpedantic]
   592 |   struct {
       |          ^

-- 
Florian

_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Thomas Petazzoni via buildroot]

Hello Florian,

On Wed, 17 Jul 2024 14:33:03 -0700
Florian Fainelli <f.fainelli@gmail.com> wrote:

> This looks similar to what I encountered a few days ago, the actual 
> issue was that libsanitizer was referencing scsi/scsi.h (sorry about the 
> message being in French, he he):

I think I should be able to read French :)

> ../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:128:10: 
> erreur fatale: scsi/scsi.h : No such file or directory
>    128 | #include <scsi/scsi.h>
>        |          ^~~~~~~~~~~~~
> compilation terminée.

I am not sure this is the same issue. Indeed, I don't see why this
missing <scsi/scsi.h> would be caused by BR2_TIME_BITS_64=y.

However, I see some host-gcc-final failures in our autobuilders that
seem related to BR2_TIME_BITS_64, such
as http://autobuild.buildroot.net/results/ff2/ff2dbfdabf0bb6a0d82ea8a80122ab97fd75bd3f/build-end.log:

/home/autobuild/autobuild/instance-0/output-1/host/i686-buildroot-linux-gnu/sysroot/usr/include/features-time64.h:26:5: error: #error "_TIME_BITS=64 is allowed only with _FILE_OFFSET_BITS=64"

I don't find your <scsi/scsi.h> issue in our autobuilders though. Do
you have a defconfig that reproduces the issue? <scsi/scsi.h> is
supposed to be installed by glibc.

Best regards,

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Florian Fainelli]

On 7/17/24 14:53, Thomas Petazzoni wrote:
> Hello Florian,
> 
> On Wed, 17 Jul 2024 14:33:03 -0700
> Florian Fainelli <f.fainelli@gmail.com> wrote:
> 
>> This looks similar to what I encountered a few days ago, the actual
>> issue was that libsanitizer was referencing scsi/scsi.h (sorry about the
>> message being in French, he he):
> 
> I think I should be able to read French :)
> 
>> ../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:128:10:
>> erreur fatale: scsi/scsi.h : No such file or directory
>>     128 | #include <scsi/scsi.h>
>>         |          ^~~~~~~~~~~~~
>> compilation terminée.
> 
> I am not sure this is the same issue. Indeed, I don't see why this
> missing <scsi/scsi.h> would be caused by BR2_TIME_BITS_64=y.
> 
> However, I see some host-gcc-final failures in our autobuilders that
> seem related to BR2_TIME_BITS_64, such
> as http://autobuild.buildroot.net/results/ff2/ff2dbfdabf0bb6a0d82ea8a80122ab97fd75bd3f/build-end.log:
> 
> /home/autobuild/autobuild/instance-0/output-1/host/i686-buildroot-linux-gnu/sysroot/usr/include/features-time64.h:26:5: error: #error "_TIME_BITS=64 is allowed only with _FILE_OFFSET_BITS=64"
> 
> I don't find your <scsi/scsi.h> issue in our autobuilders though. Do
> you have a defconfig that reproduces the issue? <scsi/scsi.h> is
> supposed to be installed by glibc.

This is not happening with a defconfig, however it is currently 
happening after refreshing my existing .config file:

https://gist.github.com/ffainelli/8fe7988c86ee8a3016b893268b9a2f49

Thanks!
--
Florian

_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Thomas Petazzoni via buildroot]

On Wed, 17 Jul 2024 15:06:50 -0700
Florian Fainelli <f.fainelli@gmail.com> wrote:

> This is not happening with a defconfig, however it is currently 
> happening after refreshing my existing .config file:

When I say "provide a defconfig", I don't mean "a defconfig that is in
configs/", but rather "whatever configuration you're using, minified as
a defconfig for brevity".

> https://gist.github.com/ffainelli/8fe7988c86ee8a3016b893268b9a2f49

Thanks! I don't know when I'll have a look, so if others want to have a
look, don't hesitate :-)

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Thomas Petazzoni via buildroot]

Hello Florian,

On Wed, 17 Jul 2024 15:06:50 -0700
Florian Fainelli <f.fainelli@gmail.com> wrote:

> This is not happening with a defconfig, however it is currently 
> happening after refreshing my existing .config file:
> 
> https://gist.github.com/ffainelli/8fe7988c86ee8a3016b893268b9a2f49

For me, this configuration builds fine. Or more precisely:
host-gcc-final builds fine, several target packages are built OK, but
it fails when building the Linux kernel, because
BR2_LINUX_KERNEL_CUSTOM_REPO_VERSION="5.14" doesn't work, it should be
v5.14.

But clearly, I don't hit the <scsi/scsi.h> issue. Are you sure you can
reproduce with this configuration? If so, on which host distro (even
though I don't really see how this could be related to the host system,
but who knows)?

Thanks!

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Florian Fainelli]

On 7/18/2024 1:05 PM, Thomas Petazzoni wrote:
> Hello Florian,
> 
> On Wed, 17 Jul 2024 15:06:50 -0700
> Florian Fainelli <f.fainelli@gmail.com> wrote:
> 
>> This is not happening with a defconfig, however it is currently
>> happening after refreshing my existing .config file:
>>
>> https://gist.github.com/ffainelli/8fe7988c86ee8a3016b893268b9a2f49
> 
> For me, this configuration builds fine. Or more precisely:
> host-gcc-final builds fine, several target packages are built OK, but
> it fails when building the Linux kernel, because
> BR2_LINUX_KERNEL_CUSTOM_REPO_VERSION="5.14" doesn't work, it should be
> v5.14.

Good point, not really an issue here because I have a local.mk 
overriding the Linux sources but I will go fix it anyway.

> 
> But clearly, I don't hit the <scsi/scsi.h> issue. Are you sure you can
> reproduce with this configuration? If so, on which host distro (even
> though I don't really see how this could be related to the host system,
> but who knows)?

Thanks for taking a look! I can reproduce it consistently from an 
incremental build. Distribution is Ubuntu 22.04.4. 
/usr/include/scsi/scsi.h is present and valid, so I am not too sure yet 
why this happening, I will try to investigate over the weekend.
-- 
Florian
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Thomas Petazzoni via buildroot]

On Thu, 18 Jul 2024 19:55:28 -0700
Florian Fainelli <f.fainelli@gmail.com> wrote:

> Thanks for taking a look! I can reproduce it consistently from an 
> incremental build. Distribution is Ubuntu 22.04.4. 
> /usr/include/scsi/scsi.h is present and valid, so I am not too sure yet 
> why this happening, I will try to investigate over the weekend.

I don't think /usr/include/scsi/scsi.h matters here, it's really the
scsi/scsi.h in the toolchain sysroot that matters as libsanitizer is
compiled for the target if I'm not mistaken. You're building with
glibc, and glibc definitely installs <scsi/scsi.h>, so I'm a bit
puzzled by your build issue.

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] [External] Re: Year 2038 time util issue - Build failure in 2024.02.x

[Karicheri, Murali via buildroot]

-----Original Message-----
From: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Sent: Wednesday, July 17, 2024 5:53 PM
To: Florian Fainelli <f.fainelli@gmail.com>
Cc: Karicheri, Murali via buildroot <buildroot@buildroot.org>; Karicheri, Murali <Murali.Karicheri@sandc.com>
Subject: [External] Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x


Hello Florian,

On Wed, 17 Jul 2024 14:33:03 -0700
Florian Fainelli <f.fainelli@gmail.com> wrote:

> This looks similar to what I encountered a few days ago, the actual
> issue was that libsanitizer was referencing scsi/scsi.h (sorry about
> the message being in French, he he):

I think I should be able to read French :)

> ../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:128:10:
> erreur fatale: scsi/scsi.h : No such file or directory
>    128 | #include <scsi/scsi.h>
>        |          ^~~~~~~~~~~~~
> compilation terminée.

I am not sure this is the same issue. Indeed, I don't see why this missing <scsi/scsi.h> would be caused by BR2_TIME_BITS_64=y.

However, I see some host-gcc-final failures in our autobuilders that seem related to BR2_TIME_BITS_64, such as http://autobuild.buildroot.net/results/ff2/ff2dbfdabf0bb6a0d82ea8a80122ab97fd75bd3f/build-end.log:

/home/autobuild/autobuild/instance-0/output-1/host/i686-buildroot-linux-gnu/sysroot/usr/include/features-time64.h:26:5: error: #error "_TIME_BITS=64 is allowed only with _FILE_OFFSET_BITS=64"

Hi Thomas,

I have checked the above log, and it seems to me the same issue.

Murai
I don't find your <scsi/scsi.h> issue in our autobuilders though. Do you have a defconfig that reproduces the issue? <scsi/scsi.h> is supposed to be installed by glibc.

Best regards,

Thomas
--
Thomas Petazzoni, co-owner and CEO, Bootlin Embedded Linux and Kernel engineering and training
https://bootlin.com/
________________________________
NOTICE OF CONFIDENTIALITY:
This message may contain information that is considered confidential and which may be prohibited from disclosure under applicable law or by contractual agreement. The information is intended solely for the use of the individual or entity named above. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution or use of the information contained in or attached to this message is strictly prohibited. If you have received this email transmission in error, please notify the sender by replying to this email and then delete it from your system.
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] [External] Re: Year 2038 time util issue - Build failure in 2024.02.x

[Karicheri, Murali via buildroot]

> -----Original Message-----
> From: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
> Sent: Wednesday, July 17, 2024 5:53 PM
> To: Florian Fainelli <f.fainelli@gmail.com>
> Cc: Karicheri, Murali via buildroot <buildroot@buildroot.org>; Karicheri, Murali
> <Murali.Karicheri@sandc.com>
> Subject: [External] Re: [Buildroot] Year 2038 time util issue - Build failure in
> 2024.02.x
>
> EXTERNAL EMAIL: This email originated from outside of S&C. Please handle
> with care and check for warning signs before opening attachments, clicking
> links, or responding to this email.
>
>
>
> Hello Florian,
>
> On Wed, 17 Jul 2024 14:33:03 -0700
> Florian Fainelli <f.fainelli@gmail.com> wrote:
>
> > This looks similar to what I encountered a few days ago, the actual
> > issue was that libsanitizer was referencing scsi/scsi.h (sorry about
> > the message being in French, he he):
>
> I think I should be able to read French :)
>
> >
> ../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:1
> 28:10:
> > erreur fatale: scsi/scsi.h : No such file or directory
> >    128 | #include <scsi/scsi.h>
> >        |          ^~~~~~~~~~~~~
> > compilation terminée.
>
> I am not sure this is the same issue. Indeed, I don't see why this missing
> <scsi/scsi.h> would be caused by BR2_TIME_BITS_64=y.
>
> However, I see some host-gcc-final failures in our autobuilders that seem
> related to BR2_TIME_BITS_64, such as
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fauto
> build.buildroot.net%2Fresults%2Fff2%2Fff2dbfdabf0bb6a0d82ea8a80122ab
> 97fd75bd3f%2Fbuild-
> end.log&data=05%7C02%7CMurali.Karicheri%40sandc.com%7Cdd0d63a10c
> af48bad13108dca6aae284%7C21c4b724f908455e813c11c8d17f3e9a%7C0
> %7C0%7C638568500087203885%7CUnknown%7CTWFpbGZsb3d8eyJWIjoi
> MC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0
> %7C%7C%7C&sdata=L5M0l3d3AZUCXIZOQMHHg7y0WRV%2BuhbrM2k%2F
> 0eVRy08%3D&reserved=0:
>
> /home/autobuild/autobuild/instance-0/output-1/host/i686-buildroot-linux-
> gnu/sysroot/usr/include/features-time64.h:26:5: error: #error
> "_TIME_BITS=64 is allowed only with _FILE_OFFSET_BITS=64"
>
[Karicheri, Murali
________________________________
NOTICE OF CONFIDENTIALITY:
This message may contain information that is considered confidential and which may be prohibited from disclosure under applicable law or by contractual agreement. The information is intended solely for the use of the individual or entity named above. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution or use of the information contained in or attached to this message is strictly prohibited. If you have received this email transmission in error, please notify the sender by replying to this email and then delete it from your system.
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] [External] Re: Year 2038 time util issue - Build failure in 2024.02.x

[Karicheri, Murali via buildroot]

> -----Original Message-----
> From: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
> Sent: Wednesday, July 17, 2024 5:53 PM
> To: Florian Fainelli <f.fainelli@gmail.com>
> Cc: Karicheri, Murali via buildroot <buildroot@buildroot.org>; Karicheri, Murali
> <Murali.Karicheri@sandc.com>
> Subject: [External] Re: [Buildroot] Year 2038 time util issue - Build failure in
> 2024.02.x
>
>
> Hello Florian,
>
> On Wed, 17 Jul 2024 14:33:03 -0700
> Florian Fainelli <f.fainelli@gmail.com> wrote:
>
> > This looks similar to what I encountered a few days ago, the actual
> > issue was that libsanitizer was referencing scsi/scsi.h (sorry about
> > the message being in French, he he):
>
> I think I should be able to read French :)
>
> >
> ../../../../libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:1
> 28:10:
> > erreur fatale: scsi/scsi.h : No such file or directory
> >    128 | #include <scsi/scsi.h>
> >        |          ^~~~~~~~~~~~~
> > compilation terminée.
>
> I am not sure this is the same issue. Indeed, I don't see why this missing
> <scsi/scsi.h> would be caused by BR2_TIME_BITS_64=y.
>
> However, I see some host-gcc-final failures in our autobuilders that seem
> related to BR2_TIME_BITS_64, such as
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fauto
> build.buildroot.net%2Fresults%2Fff2%2Fff2dbfdabf0bb6a0d82ea8a80122ab
> 97fd75bd3f%2Fbuild-
> end.log&data=05%7C02%7CMurali.Karicheri%40sandc.com%7Cdd0d63a10c
> af48bad13108dca6aae284%7C21c4b724f908455e813c11c8d17f3e9a%7C0
> %7C0%7C638568500087203885%7CUnknown%7CTWFpbGZsb3d8eyJWIjoi
> MC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0
> %7C%7C%7C&sdata=L5M0l3d3AZUCXIZOQMHHg7y0WRV%2BuhbrM2k%2F
> 0eVRy08%3D&reserved=0:
>
> /home/autobuild/autobuild/instance-0/output-1/host/i686-buildroot-linux-
> gnu/sysroot/usr/include/features-time64.h:26:5: error: #error
> "_TIME_BITS=64 is allowed only with _FILE_OFFSET_BITS=64"
>

Sorry to send my response multiple times as it got messed up by our email server.
The issue above appears to be same as what I saw.

Murali
> I don't find your <scsi/scsi.h> issue in our autobuilders though. Do you have a
> defconfig that reproduces the issue? <scsi/scsi.h> is supposed to be installed
> by glibc.
>
> Best regards,
>
> Thomas
> --
> Thomas Petazzoni, co-owner and CEO, Bootlin Embedded Linux and Kernel
> engineering and training
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fboo
> tlin.com%2F&data=05%7C02%7CMurali.Karicheri%40sandc.com%7Cdd0d63
> a10caf48bad13108dca6aae284%7C21c4b724f908455e813c11c8d17f3e9a
> %7C0%7C0%7C638568500087212824%7CUnknown%7CTWFpbGZsb3d8ey
> JWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D
> %7C0%7C%7C%7C&sdata=BFBdZRQW5LF0QSuELFtLbH3TF%2F1p2L0%2B5C
> R0Xj1PaN4%3D&reserved=0
________________________________
NOTICE OF CONFIDENTIALITY:
This message may contain information that is considered confidential and which may be prohibited from disclosure under applicable law or by contractual agreement. The information is intended solely for the use of the individual or entity named above. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution or use of the information contained in or attached to this message is strictly prohibited. If you have received this email transmission in error, please notify the sender by replying to this email and then delete it from your system.
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] Year 2038 time util issue - Build failure in 2024.02.x

[Karicheri, Murali via buildroot]

-----Original Message-----
From: Karicheri, Murali
Sent: Wednesday, July 17, 2024 2:19 PM
To: buildroot@buildroot.org
Subject: Year 2038 time util issue - Build failure in 2024.02.x

Hi All,

We were using buildroot version 2023.08 version and encountered a year 2038 crash when using time/date utilities. So, my research found some work done in the recent version of buildroot. Where BR2_TIME_BITS_64=y was added to build with 64bit time field. My board is based on a STM32MP133c SoC and using an external project to build buildroot. So, I have migrated our project to using 2024.02.x version of buildroot and enabled BR2_TIME_BITS_64=y to get the fix for the 2038 issue. But my build fails with the logs below.

Could you help me answer the following questions?
1. Is the Y2038 time issue resolved in buildroot? If not, is there some work, though officially not supported,  I can investigate for a potential solution 2. If it is in fact resolved and expected to work in 2024.02.x, what I might be missing? Is there specific version of the toolchain packages I need to select in the Toolchain configuration and other packages?

libtool: compile:  /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc/xgcc -shared-libgcc -B/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/./gcc -nostdinc++ -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/src/.libs -L/src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/build/arm-buildroot-linux-gnueabihf/libstdc++-v3/libsupc++/.libs -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/bin/ -B/src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/lib/ -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/include -isystem /src/linux-bsp-stm32mp133c/output/development/host/arm-buildroot-linux-gnueabihf/sys-
 include -D_GNU_SOURCE -D_DEBUG -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS -DHAVE_RPC_XDR_H=0 -DHAVE_TIRPC_RPC_XDR_H=0 -I. -I../../../../libsanitizer/sanitizer_common -I.. -I ../../../../libsanitizer/include -I ../../../../libsanitizer -isystem ../../../../libsanitizer/include/system -Wall -W -Wno-unused-parameter -Wwrite-strings -pedantic -Wno-long-long -fPIC -fno-builtin -fno-exceptions -fno-rtti -fomit-frame-pointer -funwind-tables -fvisibility=hidden -Wno-variadic-macros -I../../libstdc++-v3/include -I../../libstdc++-v3/include/arm-buildroot-linux-gnueabihf -I../../../../libsanitizer/../libstdc++-v3/libsupc++ -std=gnu++14 -DSANITIZER_LIBBACKTRACE -DSANITIZER_CP_DEMANGLE -I ../../../../libsanitizer/../libbacktrace -I ../libbacktrace -I ../../../../libsanitizer/../include -include ../../../../libsanitizer/libbacktrace/backtrace-rename.h -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_TIME_BITS=64 -Og -g2 -D_GNU_SOURCE -MT sanitizer_pro
 cmaps_common.lo -MD -MP -MF .deps/sanitizer_procmaps_common.Tpo -c ../../../../libsanitizer/sanitizer_common/sanitizer_procmaps_common.cpp -o sanitizer_procmaps_common.o >/dev/null 2>&1 mv -f .deps/sanitizer_linux_libcdep.Tpo .deps/sanitizer_linux_libcdep.Plo mv -f .deps/sanitizer_procmaps_linux.Tpo .deps/sanitizer_procmaps_linux.Plo
mv -f .deps/sanitizer_linux.Tpo .deps/sanitizer_linux.Plo mv -f .deps/sanitizer_printf.Tpo .deps/sanitizer_printf.Plo mv -f .deps/sanitizer_posix.Tpo .deps/sanitizer_posix.Plo mv -f .deps/sanitizer_procmaps_common.Tpo .deps/sanitizer_procmaps_common.Plo
mv -f .deps/sanitizer_posix_libcdep.Tpo .deps/sanitizer_posix_libcdep.Plo
make[6]: *** [Makefile:533: all-recursive] Error 1
make[5]: *** [Makefile:420: all] Error 2
make[4]: *** [Makefile:13138: all-target-libsanitizer] Error 2
make[3]: *** [Makefile:1033: all] Error 2
make[2]: *** [package/pkg-generic.mk:283: /src/linux-bsp-stm32mp133c/output/development/build/host-gcc-final-12.3.0/.stamp_built] Error 2
make[1]: *** [Makefile:82: _all] Error 2
make: *** [Makefile:106: /src/linux-bsp-stm32mp133c/output/development/host/bin/arm-none-linux-gnueabihf-gcc] Error 2

Hi Thomas, Florian,

I could see your responses on the webui, but couldn't get them in my inbox. First of al,l thanks for your responses. I have some updates. I will also provide the dotconfg for the failure case. I would provide you a snapshot of the logs several lines before the issue as well. The good news is that I am able to set time beyond 2038 which was crashing earlier. We are not using C++ in our project and could disable it by removing BR2_TOOLCHAIN_BUILDROOT_CXX=y from my defconfig. So the issue happens with CPP part of toolchain.

Murali Karicheri
Software Architect
S&C Electric Company
________________________________
NOTICE OF CONFIDENTIALITY:
This message may contain information that is considered confidential and which may be prohibited from disclosure under applicable law or by contractual agreement. The information is intended solely for the use of the individual or entity named above. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution or use of the information contained in or attached to this message is strictly prohibited. If you have received this email transmission in error, please notify the sender by replying to this email and then delete it from your system.
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot