On 2017-12-13 13:37:54 -0800, Andres Freund wrote:
> Hi Michael,
>
> On 2017-12-13 01:01:19 +0000, Andres Freund wrote:
> > Provide overflow safe integer math inline functions.
>
> https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=dangomushi&dt=2017-12-13%2018%3A00%3A18
>
> which seems half like a compiler bug to me. But either way, we gotta
> work around it. I suspect the reason configure test doesn't
> sufficiently detect this here is because it's testing the function with
> constant arguments.
>
> Could you perhaps test whether replacing PGAC_C_BUILTIN_OP_OVERFLOW's body with something like
> result
> PG_INT64_TYPE a;
> PG_INT64_TYPE b;
> PG_INT64_TYPE result;
> __builtin_mul_overflow(*(volatile PG_INT64_TYPE*) &a, *(volatile PG_INT64_TYPE*) &b, &result);
>
> makes it fail? I'd rather not test this via the buildfarm, given that
> dangomushi isn't the most frequently running / fastest animal.
I've since tried this via the buildfarm, but still:
configure:14480: checking for __builtin_mul_overflow
configure:14500: ccache clang -o conftest -Wall -Wmissing-prototypes -Wpointer-arith -Wdeclaration-after-statement
-Wendif-labels-Wmissing-format-attribute -Wformat-security -fno-strict-aliasing -fwrapv
-Wno-unused-command-line-argument-g -O2 -D_GNU_SOURCE -I/usr/include/et conftest.c -lssl -lcrypto -lz -lreadline
-lrt-lcrypt -ldl -lm >&5
configure:14500: $? = 0
configure:14508: result: yes
I'm not quite following. Could you check if the same happens without
-O2? Not because that'd be a solution, but to narrow down how this
happens?
Greetings,
Andres Freund
