commit 6ca5f1804d5f7a70bcee036fd5499652837b5f75 tree bc6637c7d7fdf0c347b2c49f6c5748400766d552 parent c25a2875e89bf6cbf205938ed64383ec5e817f0b author David Spickett 2025-08-04 10:03:48 +0100 committer David Spickett 2025-08-04 10:11:07 +0100 tcwg-llvmbot: Restrict amp-01 flang GCC to 16 cores This appears to use around 9GB of RAM per process at peak. Worse than clang builds of flang. 16 cores takes us beyond the 160 that jade has so I've added a hostname check. With amp-01's generated but jade's hand written since it's going away soon anyway. Tempted to restrict every flang bot but want to see if GCC flang really is the main issue first. Change-Id: I66d7da1e043ef5d563e45b906b082f17b4be9579 :100644 100644 ee8e69d44053c5958d0d8da476e8676e06ea6114 29d64b6c57b8b9d5a04f8158a36fa88bbcad3815 M tcwg-base/tcwg-llvmbot/generate_core_assignments.py :100755 100755 c00dcccc45628cf19c9193c786a8ee6479169f66 acf6effdcfb984ce2184bab716696da4c806a479 M tcwg-base/tcwg-llvmbot/start.sh