a73049b2a1
As shown in the previous commit, qemu_strtosz_MiB sometimes leaves the
result value untouched (we have to audit further to learn that in that
case, the QAPI generator says that visit_type_NumaOptions() will have
zero-initialized it), and sometimes leaves it with the value of a
partial parse before -EINVAL occurs because of trailing garbage.
Rather than blindly treating any string the user may throw at us as
valid, we should check for parse failures.
Fixes:
|
||
---|---|---|
.. | ||
bus.c | ||
clock-vmstate.c | ||
clock.c | ||
cpu-common.c | ||
cpu-sysemu.c | ||
fw-path-provider.c | ||
generic-loader.c | ||
gpio.c | ||
guest-loader.c | ||
guest-loader.h | ||
hotplug-stubs.c | ||
hotplug.c | ||
irq.c | ||
Kconfig | ||
loader-fit.c | ||
loader.c | ||
machine-hmp-cmds.c | ||
machine-qmp-cmds.c | ||
machine-smp.c | ||
machine.c | ||
meson.build | ||
nmi.c | ||
null-machine.c | ||
numa.c | ||
or-irq.c | ||
platform-bus.c | ||
ptimer.c | ||
qdev-clock.c | ||
qdev-fw.c | ||
qdev-hotplug.c | ||
qdev-prop-internal.h | ||
qdev-properties-system.c | ||
qdev-properties.c | ||
qdev.c | ||
register.c | ||
reset.c | ||
resettable.c | ||
split-irq.c | ||
stream.c | ||
sysbus-fdt.c | ||
sysbus.c | ||
trace-events | ||
trace.h | ||
uboot_image.h | ||
vm-change-state-handler.c | ||
vmstate-if.c |