Re: [PATCH 01/14] Makefile: Only build virtiofsd if system-mode is enabled

[Richard Henderson]

On 3/15/20 3:16 PM, Philippe Mathieu-Daudé wrote:
> On 3/15/20 8:40 PM, Richard Henderson wrote:
>> On 3/13/20 3:28 PM, Laurent Vivier wrote:
>>>>   ifeq ($(CONFIG_LINUX)$(CONFIG_SECCOMP)$(CONFIG_LIBCAP_NG),yyy)
>>>> -HELPERS-y += virtiofsd$(EXESUF)
>>>> -vhost-user-json-y += tools/virtiofsd/50-qemu-virtiofsd.json
>>>> +HELPERS-$(CONFIG_SOFTMMU) += virtiofsd$(EXESUF)
>>>> +vhost-user-json-$(CONFIG_SOFTMMU) += 
>>>> tools/virtiofsd/50-qemu-virtiofsd.json
>> ...
>>>
>>> Why don't you do that instead?
>>>
>>>    ifeq
>>> ($(CONFIG_LINUX)$(CONFIG_SECCOMP)$(CONFIG_LIBCAP_NG)$(CONFIG_SOFTMMU),yyyy)
> 
> This was my first option but checkpatch complained because it is 80 chars 
> long...
> 
> I also tried:
> 
> ifeq \
> ($(CONFIG_LINUX)$(CONFIG_SECCOMP)$(CONFIG_LIBCAP_NG)$(CONFIG_SOFTMMU),yyyy)
> 
> which is uglier and still too long.
> 
> Following qemu-bridge-helper syntax I also tried:
> 
> ifeq ($(CONFIG_SECCOMP)$(CONFIG_LIBCAP_NG),yy)
> HELPERS-$(call land,$(CONFIG_SOFTMMU),$(CONFIG_LINUX)) += virtiofsd$(EXESUF)
> 
> which is also too long ¯\_(ツ)_/¯
> 
> If you accept checkpatch complain I can use the form you prefer :)

Hmph.  Of the choices, I prefer the 80+ char line.


r~