commit: e4dbf2424f80030489d714272de6c7317bedf5c3
Author: Sam James <sam <AT> gentoo <DOT> org>
AuthorDate: Sat May 31 18:47:34 2025 +0000
Commit: Sam James <sam <AT> gentoo <DOT> org>
CommitDate: Sat May 31 18:47:34 2025 +0000
URL: https://gitweb.gentoo.org/proj/gcc-patches.git/commit/?id=e4dbf242
16.0.0: don't run afdo tests
Noticed by zen_deus after an abnormally high load on a machine I was
building GCC on. Seems to have started with
7b76965df10c47616c8f65e1c1134356d635f9ed
which is curious as it doesn't seem to have shown up for me before on
Intel machines, but I'm not so familiar with what it relies on yet.
Signed-off-by: Sam James <sam <AT> gentoo.org>
16.0.0/gentoo/36_all_no-afdo-testing.patch | 13 +++++++++++++
16.0.0/gentoo/README.history | 4 ++++
2 files changed, 17 insertions(+)
diff --git a/16.0.0/gentoo/36_all_no-afdo-testing.patch
b/16.0.0/gentoo/36_all_no-afdo-testing.patch
new file mode 100644
index 0000000..9282804
--- /dev/null
+++ b/16.0.0/gentoo/36_all_no-afdo-testing.patch
@@ -0,0 +1,13 @@
+`perf` doesn't seem to work well in sandbox and this isn't so useful
+anyway for packaging testing purposes. These tests seemed to hang.
+--- a/gcc/testsuite/lib/target-supports.exp
++++ b/gcc/testsuite/lib/target-supports.exp
+@@ -799,7 +799,7 @@ proc check_profiling_available { test_what } {
+ }
+ puts $wrapper
+ global srcdir
+- set status [remote_exec host "$wrapper true -v >/dev/null"]
++ set status [remote_exec host "false >/dev/null"]
+ if { [lindex $status 0] != 0 } {
+ verbose "autofdo not supported because perf does not work"
+ return 0
diff --git a/16.0.0/gentoo/README.history b/16.0.0/gentoo/README.history
index a75f6c8..189da71 100644
--- a/16.0.0/gentoo/README.history
+++ b/16.0.0/gentoo/README.history
@@ -1,3 +1,7 @@
+3 ????
+
+ + 36_all_no-afdo-testing.patch
+
2 11 May 2025
-
85_all_PR120080_Revert-gimple-Switch-bit-test-lowering-testcases-for.patch
