Re: [PULL 05/22] target/i386/hvf: Include missing 'exec/target_page.h' header

[Wei Liu]

On Wed, May 07, 2025 at 01:20:49PM +0200, Philippe Mathieu-Daudé wrote:
> Hi Wei,
> 
> On 7/5/25 08:35, Wei Liu wrote:
> > On Tue, May 06, 2025 at 04:34:54PM +0200, Philippe Mathieu-Daudé wrote:
> > > Include "exec/target_page.h" to be able to compile HVF on x86_64:
> > > 
> > >    ../target/i386/hvf/hvf.c:139:49: error: use of undeclared identifier 
> > > 'TARGET_PAGE_SIZE'
> > >                uint64_t dirty_page_start = gpa & ~(TARGET_PAGE_SIZE - 1u);
> > >                                                  ^
> > >    ../target/i386/hvf/hvf.c:141:45: error: use of undeclared identifier 
> > > 'TARGET_PAGE_SIZE'
> > >                hv_vm_protect(dirty_page_start, TARGET_PAGE_SIZE,
> > > 
> > > Fixes: 9c2ff9cdc9b ("exec/cpu-all: remove exec/target_page include")
> > > Reported-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
> > > Reported-by: Wei Liu <wei.liu@kernel.org>
> > > Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> > > Message-Id: <20250425174310.70890-1-philmd@linaro.org>
> > > Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
> > 
> > FYI Paolo merged a patch from me that does the same thing. My patch is
> > already in the master branch.
> 
> Paolo said he'd replace your patch by this older one which mentions
> the problematic commit, which is why I didn't notice your patch,
> I was expecting this one to disappear on rebase. I don't plan to
> repost this PR except if there is a serious problem with it. In
> this case, we can still revert if we are picky. Duplicated includes
> aren't a problem, missing ones are.
> 

Sure.

> BTW since you are testing / using HVF on x86, do you know if your
> employer could help extending our CI infrastructure coverage for
> your use cases?
> 

No, we don't have a use case for HVF. I only tested it because I wanted
to refactor the instruction emulator out of the HVF code base.

Thanks,
Wei.

> Regards,
> 
> Phil.