[Note: Adding sunp...@freebsd.org and isotni...@vdura.com for possible
additional example packages that could use such changes.]

On Feb 22, 2025, at 06:53, Mark Millard <mark...@yahoo.com> wrote:

> [Note: I've added rnagy@ for the new www/chromium note.]
> 
> On Feb 22, 2025, at 06:08, Hiroki Tagato <tagat...@freebsd.org> wrote:
> 
>> On 2025/02/22 16:41, Mark Millard wrote:
>>> On Feb 21, 2025, at 21:57, Mark Millard <mark...@yahoo.com> wrote:
>>>> On Feb 21, 2025, at 20:51, Mark Millard <mark...@yahoo.com> wrote:
>>>> 
>>>>> On Feb 21, 2025, at 19:34, Hiroki Tagato <tagat...@freebsd.org> wrote:
>>>>> 
>>>>>> On 2025/02/22 11:29, Hiroki Tagato wrote:
>>>>>>> On 2025/02/20 1:26, Mark Millard wrote:
>>>>>>>> On Feb 19, 2025, at 06:42, Hiroki Tagato <tagat...@freebsd.org> wrote:
>>>>>>>> 
>>>>>>>>> On 2025/02/19 18:05, Jason E. Hale wrote:
>>>>>>>>>> On Wed, Feb 19, 2025 at 3:20 AM Hiroki Tagato <tagat...@freebsd.org> 
>>>>>>>>>> wrote:
>>>>>>>>>>> 
>>>>>>>>>>> On 2025/02/18 19:03, Mark Millard wrote:
>>>>>>>>>>>> 
>>>>>>>>>>>> 
>>>>>>>>>>>> On Feb 17, 2025, at 22:27, Mark Millard <mark...@yahoo.com> wrote:
>>>>>>>>>>>> 
>>>>>>>>>>>>> On Feb 17, 2025, at 19:38, Mark Millard <mark...@yahoo.com> wrote:
>>>>>>>>>>> . . .
>>>>>>>>>>> 
>>>>>>>>>> Hi all,
>>>>>>>>>> I maintain archivers/libunrar and after seeing this, I just committed
>>>>>>>>>> a patch [1] to my port to fix this problem. The libunrar patches have
>>>>>>>>>> some additional guff to add openssl support that should not be
>>>>>>>>>> included for chromium-based projects, but the concept is basically 
>>>>>>>>>> the
>>>>>>>>>> same. Should be pretty easy to adapt.
>>>>>>>>>> [1] https://cgit.freebsd.org/ports/commit/? 
>>>>>>>>>> id=96e53c0c65faa691e18c0a5558d8f445e6010a70
>>>>>>>>> 
>>>>>>>>> Hi Jason,
>>>>>>>>> 
>>>>>>>>> Thanks a lot for letting me know the commit. I will shortly prepare a 
>>>>>>>>> patch for electron ports based on that.
>>>>>>>> . . .
>>>>>>> Hi Mark,
>>>>> 
>>>>> Hello.
>>>>> 
>>>>>>> There was a report from mikael@ that electron32 built fine on aarch64 
>>>>>>> hardware. So I suspect you have some custom settings in make.conf 
>>>>>>> and/or poudriere.conf.
>>>>>>> Do you have custom settings?
>>>>> 
>>>>> My context fits with the reference to 96e53c0c6:
>>>>> 
>>>>> QUOTE
>>>>> archivers/libunrar: Support crc and crypto flags on aarch64
>>>>> If CPU flags crc or crytpo are set, e.g. -march=armv8-a+crc+crypto, build 
>>>>> would fail with missing headers because this was only implemented for 
>>>>> glibc and Apple.
>>>>> END QUOTE
>>>>> 
>>>>> I generally have -mcpu=cortex-a76 involved for my aarch64 builds for
>>>>> the primary machines. That would mean that crc and crypto would be
>>>>> impicitly enabled.
>>>>> 
>>>>>> Anyway, I have prepared a patch for electron 32 and 33 based on the 
>>>>>> commit:
>>>>>> https://cgit.freebsd.org/ports/commit/?id=96e53c0c65faa691e18c0a5558d8f445e6010a70
>>>>> 
>>>>> I've applied the patch file and started an electron32 build
>>>>> with no completing build activity or other significant activity,
>>>>> including not having started building electron33. That should
>>>>> get to a preliminary result as soon as possible. Then, if it
>>>>> looks worthwhile, I could try to build electron33 as well.
>>>>> 
>>>>> The prerequisites had already been built during the prior
>>>>> bulk -a experiment. So only electron is building.
>>>>> 
>>>>>> Can you try again with the attached patch if/when you have available 
>>>>>> resources?
>>>>>> # I don't have an aarch64 hardware so I can't test it by myself.
>>>>> 
>>>>> I'll let you know what happens with electron32.
>>>> 
>>>> electron32 built:
>>>> 
>>>> [01:31:28] [01] [01:30:30] Finished   devel/electron32 | 
>>>> electron32-32.3.0: Success
>>>> 
>>>> I have started an electron33 build.
>>> electron33 also built:
>>> [01:39:26] [08] [01:38:34] Finished   devel/electron33 | electron33-33.4.0: 
>>> Success
>> 
>> Thanks for confirming the builds. The changes have been committed:
>> https://cgit.freebsd.org/ports/commit/?id=e6330e8ee3c6c70b6b8e203dc6ddd581be421638
>> 
> 
> Thanks.
> 
> With archivers/libunrar and devel/electron3[23] fixed for
> build with crc and/or crypto flags for aarch64, that just
> leaves www/chromium as failing to build for such contexts.
> 
> www/chromium for aarch64 with -mcpu=cortex-a76 also failed
> with:
> 
> In file included from ../../third_party/unrar/src/crypt.cpp:1:
> In file included from ../../third_party/unrar/src/rar.hpp:6:
> ../../third_party/unrar/src/os.hpp:167:10: fatal error: 'asm/hwcap.h' file 
> not found
>  167 | #include <asm/hwcap.h>
>      |          ^~~~~~~~~~~~~
> 1 error generated.
> 
> which is what the above fixed in libunrar and electron3[23] .

Hmm. Looking at the logs from the bulk -a attempt:

# grep -rl "asm/hwcap\.h' file not found" 
/usr/local/poudriere/data/logs/bulk/main-CA76-bulk_a-default/2025-02-16_21h21m26s/
 | more
/usr/local/poudriere/data/logs/bulk/main-CA76-bulk_a-default/2025-02-16_21h21m26s/logs/unrar-7.01,6.log
/usr/local/poudriere/data/logs/bulk/main-CA76-bulk_a-default/2025-02-16_21h21m26s/logs/libunrar-7.0.9,1.log
/usr/local/poudriere/data/logs/bulk/main-CA76-bulk_a-default/2025-02-16_21h21m26s/logs/isa-l_crypto-2.25.0.log
/usr/local/poudriere/data/logs/bulk/main-CA76-bulk_a-default/2025-02-16_21h21m26s/logs/electron33-33.4.0.log
/usr/local/poudriere/data/logs/bulk/main-CA76-bulk_a-default/2025-02-16_21h21m26s/logs/electron32-32.3.0.log
/usr/local/poudriere/data/logs/bulk/main-CA76-bulk_a-default/2025-02-16_21h21m26s/logs/chromium-132.0.6834.159.log

unrar-7.01,6.log ( archivers/unrar ) and isa-l_crypto-2.25.0.log
( devel/isa-l_crypto ) may be other examples that could use such
changes for aarch64 to allow use of crc and crypto for compiles.



===
Mark Millard
marklmi at yahoo.com


Reply via email to