Re: [PATCH] drm/xe: Prefer BIT/GENMASK macros over shifts

[Matthew Brost <matthew.brost@intel.com>]

On Tue, Jan 23, 2024 at 09:14:06PM +0100, Michal Wajdeczko wrote:
> 
> 
> On 23.01.2024 20:02, Matthew Brost wrote:
> > Using BIT/GENMASK macros is a better convention than using manual shift
> > and will also fix build errors [1].
> > 
> > [1] http://kisskb.ellerman.id.au/kisskb/buildresult/15112475/log/
> 
> commit message and diff doesn't match (missed patch ?)
> 

Yep, will resend with abi/*.h changes.

> and the actual problem was only due to missed "u" suffix in
> GUC_CTB_MSG_0_FENCE and GUC_HXG_MSG_0_ORIGIN definitions (as using 0x
> notation alone really doesn't make those integer definitions unsigned)
> 
> note that those were already fixed on i915 by commit 962bd34bb457f6353f
> and discussed here in [1]
> 
> [1] https://patchwork.freedesktop.org/patch/573620/?series=128312&rev=1
> 

Thanks for the reference.

I prefer the way I have fixed this. Let me resend and we can discuss more

Matt

> > 
> > Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
> > Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
> > Cc: Lucas De Marchi <lucas.demarchi@intel.com>
> > Signed-off-by: Matthew Brost <matthew.brost@intel.com>
> > ---
> >  drivers/gpu/drm/xe/xe_guc_ct.c    | 2 +-
> >  drivers/gpu/drm/xe/xe_guc_relay.c | 2 +-
> >  2 files changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c b/drivers/gpu/drm/xe/xe_guc_ct.c
> > index ee5d99456aeb..fc5cc2b2bcc5 100644
> > --- a/drivers/gpu/drm/xe/xe_guc_ct.c
> > +++ b/drivers/gpu/drm/xe/xe_guc_ct.c
> > @@ -866,7 +866,7 @@ static int parse_g2h_response(struct xe_guc_ct *ct, u32 *msg, u32 len)
> >  	 */
> >  	if (fence & CT_SEQNO_UNTRACKED) {
> >  		if (type == GUC_HXG_TYPE_RESPONSE_FAILURE)
> > -			xe_gt_err(gt, "FAST_REQ H2G fence 0x%x failed! e=0x%x, h=%u\n",
> > +			xe_gt_err(gt, "FAST_REQ H2G fence 0x%x failed! e=0x%lx, h=%lu\n",
> >  				  fence,
> >  				  FIELD_GET(GUC_HXG_FAILURE_MSG_0_ERROR, hxg[0]),
> >  				  FIELD_GET(GUC_HXG_FAILURE_MSG_0_HINT, hxg[0]));
> > diff --git a/drivers/gpu/drm/xe/xe_guc_relay.c b/drivers/gpu/drm/xe/xe_guc_relay.c
> > index c0a2d8d5d3b3..2d1f19270d9a 100644
> > --- a/drivers/gpu/drm/xe/xe_guc_relay.c
> > +++ b/drivers/gpu/drm/xe/xe_guc_relay.c
> > @@ -300,7 +300,7 @@ static int relay_send_transaction(struct xe_guc_relay *relay, struct relay_trans
> >  		ret = -EPROTO;
> >  	}
> >  	if (unlikely(ret < 0)) {
> > -		relay_notice(relay, "Failed to send %s.%x to GuC (%pe) %*ph ...\n",
> > +		relay_notice(relay, "Failed to send %s.%lx to GuC (%pe) %*ph ...\n",
> >  			     guc_hxg_type_to_string(FIELD_GET(GUC_HXG_MSG_0_TYPE, buf[0])),
> >  			     FIELD_GET(GUC_HXG_REQUEST_MSG_0_ACTION, buf[0]),
> >  			     ERR_PTR(ret), (int)sizeof(u32) * txn->offset, buf);