Bug#1121356: opm-simulators FTBFS with scotch 7.0.10

[Adrian Bunk]

On Wed, Nov 26, 2025 at 06:26:26PM +0100, Markus Blatt wrote:
> 
> On 11/26/25 17:38, Adrian Bunk wrote:
>...
> > And the opm-grid autopkgtest failure blocking testing migration of
> > scotch is also caused by that:
> > https://tracker.debian.org/pkg/scotch
> 
> I see. so migration would happen if scotch opm-grid and opm-simulators
> migrates at the same time.
> 
> The latter is blocked by a failing test on ppc64el

I assume you mean riscv64?

> (thought it was due to lto but it isn't.

LTO is not enabled by default in Debian (it is in Ubuntu).

> Could not really replicate it on a porter machine, though).

Not strictly required, but reducing debug info can get the build time 
on riscv64 from 4 days to < 1 days with the workaround below (see [1,2]).

> I'll try to work around it.

Thanks!

> Best,
> 
> Markus

cu
Adrian

[1] https://lists.debian.org/debian-riscv/2025/09/msg00017.html
[2] https://lists.debian.org/debian-riscv/2025/09/msg00019.html

--- opm-simulators-2025.10+ds/debian/rules	2025-11-18 09:47:25.000000000 +0000
+++ opm-simulators-2025.10+ds/debian/rules	2025-11-18 15:30:46.000000000 +0000
@@ -24,6 +24,13 @@
 
 # require 7GB of RAM per make process. Yes, peak seems to be above 6 GB currently
 need_gb_ram_per_process=7
+
+# build time reduction                  
+ifneq (,$(filter $(DEB_HOST_ARCH), riscv64))
+    export DEB_CXXFLAGS_MAINT_APPEND += -g1
+    need_gb_ram_per_process=3
+endif
+
 free_ram=$(shell free -g | sed -n 2p| sed "s/ \+/ /g"| cut -d " " -f 2)
 max_procs=$(shell echo "$(free_ram)/$(need_gb_ram_per_process)" | bc)
 parallel_procs=$(shell if test "$(max_procs)" -lt "1"; then echo 1; else echo "$(max_procs)"; fi)