File gcc10-maybe-uninitialized.patch of Package qemu.20153

From: Bruce Rogers <brogers@suse.com>
Date: Wed, 22 Apr 2020 08:50:55 -0600
Subject: gcc10: maybe-uninitialized
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

References: boo#1169728

gcc 10 needs some help to understand that indeed cpu_irqs[0] does get
initialized in all cases. In this case an assert is sufficient.

Reported-by: Martin Liška <mliska@suse.cz>
Signed-off-by: Bruce Rogers <brogers@suse.com>
---
 hw/openrisc/openrisc_sim.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/hw/openrisc/openrisc_sim.c b/hw/openrisc/openrisc_sim.c
index 79e70493fc77e50556a4a92a4231..bc4b71059ff6d922e3cdc83bfc79 100644
--- a/hw/openrisc/openrisc_sim.c
+++ b/hw/openrisc/openrisc_sim.c
@@ -134,6 +134,7 @@ static void openrisc_sim_init(MachineState *machine)
     int n;
     unsigned int smp_cpus = machine->smp.cpus;
 
+    assert(smp_cpus >= 1 && smp_cpus <= 2);
     for (n = 0; n < smp_cpus; n++) {
         cpu = OPENRISC_CPU(cpu_create(machine->cpu_type));
         if (cpu == NULL) {