Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 26 Nov 2021 09:21:54 +0000
From:      "Bjoern A. Zeeb" <bzeeb-lists@lists.zabbadoz.net>
To:        greg@unrelenting.technology
Cc:        freebsd-arm@freebsd.org
Subject:   Re: world linker error "can't create dynamic relocation R_AARCH64_LDST64_ABS_LO12_NC against symbol: __stderrp in readonly segment"
Message-ID:  <4C688B42-E443-4B44-9048-7B84BCE35982@lists.zabbadoz.net>
In-Reply-To: <5B9D4DF1-32FE-4918-9A8F-DF528E88BD05@unrelenting.technology>
References:  <alpine.BSF.2.00.2111250136570.68830@ai.fobar.qr> <5B9D4DF1-32FE-4918-9A8F-DF528E88BD05@unrelenting.technology>

next in thread | previous in thread | raw e-mail | index | archive | help
On 25 Nov 2021, at 23:53, Greg V via freebsd-arm wrote:

>> On 25 Nov 2021, at 04:38, Bjoern A. Zeeb =

>> <bzeeb-lists@lists.zabbadoz.net> wrote:
>>
>> =EF=BB=BFHi,
>>
>> I am trying to update one machine (also across the LLVM update) and I =

>> get the below error.
>> Anyone any ideas?
>>
>>
>> ld: error: relocation R_AARCH64_ADR_PREL_PG_HI21 cannot be used =

>> against symbol __stderrp; recompile with -fPIC
>>>>> defined in =

>>>>> base14-251054.3ede04c78c7c726ed79a39d22c65a58d0ecc5d00/arm64.aarch6=
4/tmp/lib/libc.so.7
>>>>> referenced by assert.c
>>>>>              assert.o:(libspl_assertf) in archive =

>>>>> base14-251054.3ede04c78c7c726ed79a39d22c65a58d0ecc5d00/arm64.aarch6=
4/tmp/usr/lib/libspl.a
>
> That would be https://reviews.freebsd.org/D32521#739094 =E2=80=93 this =
was =

> fixed in the final version of the patch that actually was committed =

> though.
>
> And I=E2=80=99ve just checked that it all builds on aarch64 too. Works =
for =

> me, and presumably for anyone else building world in the last 10 days =

> or so. So that=E2=80=99s weird. Maybe some leftover /usr/obj state some=
how =

> causing issues or something?

We found that this particular build issue is caused by WITH_CDDL and =

WITHOUT_ZFS as a result of a recent commit to fix library dependencies =

in that area;  markj has given me a patch and I tested it successfully =

for that specific case last night at almost 2AM;  I assume the issue =

will be solved the next days in main.   Sorry for not following up =

anymore last night right away.

/bz



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4C688B42-E443-4B44-9048-7B84BCE35982>