I've spent the last few days recompiling several versions of GCC and
installing them in /usr/local/gccxx so I can try different versions of GCC
against the different versions of Sage I have.
Yes, I'm building Sage as root (uid = 0).
I will report back as soon as my tests are complete.
Thanks for the suggestions.
- Dale
On Sunday, July 28, 2024 at 9:16:01 AM UTC-5 Volker Braun wrote:
> Pretty sure he means the "uid == 0" kind of root ;)
>
> If you have enough memory: try upgrading to gcc 14.1.1 or use our bundled
> gcc 13 (./configure --with-system-gcc=no)
>
> On Sunday, July 28, 2024 at 4:10:32 PM UTC+2 Dima Pasechnik wrote:
>
>>
>>
>> On 24 July 2024 17:15:36 BST, Dale Marshall wrote:
>> >
>> >I'm on a closed network so I cannot provide the log files.
>> >
>> >I'm building Sage 10.4 on a RHEL8.10 box using root and GCC 14.1.0. The
>> >configure line was:
>> >
>> >./configure --prefix= --enable-build-as-root --enable-wheels
>> >--with-python=
>> >
>> >The configure showed lots of packages to be compiled (even though I've
>> >installed as many as I could). The make eventually fails on
>> >giac-1.9.0.15p0.
>> >
>> >The last error messages are:
>> >init2.c:52:MPFR assertion failed: p >= 2 && p <= ((mpfr_prec_t)
>> >((mprf_uprec_t)(~(mpfr_uprec_t)0)>>1))
>> >during GIMPLE pass: fre
>> >modpoly.cc: In function 'std::complex giac::horner_newton(const
>> >vecteur&, conf std::complex&, const context*)';
>> >modpoly.cc:16083:1: internal compile error: Aborted
>> >16083 | } // namespace giac
>> > | ^
>> >0xf0873f crash signal
>> >../../gcc-14.1.0/gcc/toplev.cc:319
>> >0x7f624acd43c2 mpc_div_zero
>>
>> this is usually a sign that the compiler ran out of memory.
>>
>> C++ complilers are very memory-hungry.
>> Can you make sure there is 4Gb or more of RAM
>> available?
>>
>> We should test Sage with root - presume you mean
>> <https://root.cern/> ?
>>
>> Dima
>> >
>> >
>> >Oh... I'm getting the same error on my CentOS7.9 box using the same
>> version
>> >of GCC.
>> >
>> >Any suggestions on what I can do to resolve this issue would be welcome.
>> >
>> >- Dale
>> >
>>
>
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/sage-devel/bf490650-fd1f-4b31-8460-cf8ad76c455an%40googlegroups.com.
