1:26 p.m.
On 6/22/21 11:35 AM, Grumbach, Emmanuel wrote:
>>>> Could this be a random issue with the bot?
>>> Hi Luca,
>>>
>>> The reproduce step is not inaccurate and misleading, I can reproduce
>>> it with out-of-tree builds.
>> Sorry, 'not' is redundant, the reproduce step is wrong.
>>
> Reproduce step is wrong?
> In an case, sap.h is there. So really, I can't understand how this error can
happen.
Hi Grumbach,
It can be reproduced with out-of-tree builds:
make ARCH=i386 -C source O=$(pwd) --keep-going
and I tried to add $(srctree), the problem can be fixed.
--- a/drivers/net/wireless/intel/iwlwifi/mei/Makefile
+++ b/drivers/net/wireless/intel/iwlwifi/mei/Makefile
@@ -3,6 +3,6 @@ obj-$(CONFIG_IWLMEI) += iwlmei.o
iwlmei-y += main.o
iwlmei-y += net.o
iwlmei-$(CONFIG_IWLWIFI_DEVICE_TRACING) += trace.o -CFLAGS_trace.o
:= -I$(src)
+CFLAGS_trace.o := -I$(srctree)/$(src)
-ccflags-y += -I$(src)/../
+ccflags-y += -I$(srctree)/$(src)/../
Hmm... interesting, not sure I understand the difference between srctree and src
Looks like the kernel is more using srctree though..