linux-kernel - Re: [PATCH v6 02/20] drm/gpuva_mgr: Helper to get range of unmap from a remap op.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <877cp2gm42.fsf@intel.com>
Date:   Thu, 07 Sep 2023 16:15:57 +0300
From:   Jani Nikula <jani.nikula@...ux.intel.com>
To:     Donald Robson <Donald.Robson@...tec.com>,
        "dri-devel@...ts.freedesktop.org" <dri-devel@...ts.freedesktop.org>,
        Sarah Walker <Sarah.Walker@...tec.com>
Cc:     "corbet@....net" <corbet@....net>,
        "luben.tuikov@....com" <luben.tuikov@....com>,
        "christian.koenig@....com" <christian.koenig@....com>,
        "krzysztof.kozlowski+dt@...aro.org" 
        <krzysztof.kozlowski+dt@...aro.org>,
        "tzimmermann@...e.de" <tzimmermann@...e.de>,
        "mripard@...nel.org" <mripard@...nel.org>,
        "matthew.brost@...el.com" <matthew.brost@...el.com>,
        "afd@...com" <afd@...com>, "hns@...delico.com" <hns@...delico.com>,
        Matt Coster <Matt.Coster@...tec.com>,
        "boris.brezillon@...labora.com" <boris.brezillon@...labora.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "conor+dt@...nel.org" <conor+dt@...nel.org>,
        "devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
        "linux-doc@...r.kernel.org" <linux-doc@...r.kernel.org>,
        "robh+dt@...nel.org" <robh+dt@...nel.org>,
        "dakr@...hat.com" <dakr@...hat.com>,
        "faith.ekstrand@...labora.com" <faith.ekstrand@...labora.com>
Subject: Re: [PATCH v6 02/20] drm/gpuva_mgr: Helper to get range of unmap
 from a remap op.

On Thu, 07 Sep 2023, Donald Robson <Donald.Robson@...tec.com> wrote:
> On Thu, 2023-09-07 at 15:14 +0300, Jani Nikula wrote:
>> On Wed, 06 Sep 2023, Sarah Walker <sarah.walker@...tec.com> wrote:
>> > From: Donald Robson <donald.robson@...tec.com>
>> > 
>> > Signed-off-by: Donald Robson <donald.robson@...tec.com>
>> > ---
>> >  include/drm/drm_gpuva_mgr.h | 27 +++++++++++++++++++++++++++
>> >  1 file changed, 27 insertions(+)
>> > 
>> > diff --git a/include/drm/drm_gpuva_mgr.h b/include/drm/drm_gpuva_mgr.h
>> > index ed8d50200cc3..be7b3a6d7e67 100644
>> > --- a/include/drm/drm_gpuva_mgr.h
>> > +++ b/include/drm/drm_gpuva_mgr.h
>> > @@ -703,4 +703,31 @@ void drm_gpuva_remap(struct drm_gpuva *prev,
>> >  
>> >  void drm_gpuva_unmap(struct drm_gpuva_op_unmap *op);
>> >  
>> > +/**
>> > + * drm_gpuva_op_remap_get_unmap_range() - Helper to get the start and range of
>> > + * the unmap stage of a remap op.
>> > + * @op: Remap op.
>> > + * @start_addr: Output pointer for the start of the required unmap.
>> > + * @range: Output pointer for the length of the required unmap.
>> > + *
>> > + * These parameters can then be used by the caller to unmap memory pages that
>> > + * are no longer required.
>> > + */
>> > +static __always_inline void
>> 
>> IMO __always_inline *always* requires a justification in the commit
>> message.
>> 
>> BR,
>> Jani.
>
> Hi Jani,
> I went with __always_inline because I can't see this being used more than once per driver.
> I can add that to the commit message, but is that suitable justification? I could move
> it to the source file or make it a macro if you prefer.

My personal opinion is that static inlines in general should always have
a performance justification. If there isn't one, it should be a regular
function. Static inlines leak the abstractions and often make the header
dependencies worse.

Not everyone agrees, of course.

More than anything I was looking for justification on __always_inline
rather than just inline, though.


BR,
Jani.



> Thanks,
> Donald
>> 
>> 
>> > +drm_gpuva_op_remap_get_unmap_range(const struct drm_gpuva_op_remap *op,
>> > +				   u64 *start_addr, u64 *range)
>> > +{
>> > +	const u64 va_start = op->prev ?
>> > +			     op->prev->va.addr + op->prev->va.range :
>> > +			     op->unmap->va->va.addr;
>> > +	const u64 va_end = op->next ?
>> > +			   op->next->va.addr :
>> > +			   op->unmap->va->va.addr + op->unmap->va->va.range;
>> > +
>> > +	if (start_addr)
>> > +		*start_addr = va_start;
>> > +	if (range)
>> > +		*range = va_end - va_start;
>> > +}
>> > +
>> >  #endif /* __DRM_GPUVA_MGR_H__ */

-- 
Jani Nikula, Intel Open Source Graphics Center