qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [RFC PATCH v4 16/44] hw/alpha/Makefile.objs: Create CON


From: Thomas Huth
Subject: Re: [Qemu-devel] [RFC PATCH v4 16/44] hw/alpha/Makefile.objs: Create CONFIG_* for alpha
Date: Thu, 24 Jan 2019 10:09:22 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.9.1

On 2019-01-23 22:11, Richard Henderson wrote:
> On 1/22/19 10:55 PM, Yang Zhong wrote:
>> @@ -19,3 +19,5 @@ CONFIG_I8259=y
>>  CONFIG_MC146818RTC=y
>>  CONFIG_ISA_TESTDEV=y
>>  CONFIG_SMC37C669=y
>> +CONFIG_DP264=y
>> +CONFIG_TYPHOON=y
> 
> There's not really a point in two configs; just use DP264.

Ah, right, typhoon belongs to the clipper/dp264 machine, so that should
be fine to use only one config here.

> But beyond that, is there a point in adding configury that must be set to Y?
> There is exactly one machine type implemented for Alpha.  I don't see anyone
> implementing a different machine type in the future either.

I see two reasons:

1) Someone might want to build QEMU as plain instruction set simulator
with just the "none" machine included. Then you need the CONFIG_DP264
switch to disable the main machine. Ok, that's not very likely, but
still a possibility.

2) Since this series is about adding a Kconfig-style system to QEMU,
there might be other CONFIG switches that depend on this one later, e.g.
for the x86 machines, a later patch adds:

 config ACPI_NVDIMM
     bool
     default y
     depends on PC || Q35

Agreed, we do not have any "depends on DP264" in this series yet, but I
think it's cleaner if we add a proper config switch for each machine
right from the start.

 Thomas



reply via email to

[Prev in Thread] Current Thread [Next in Thread]