Re: [PATCH v3 1/7] fpga-mgr: wrap the write_init() op

[Tom Rix]

On 6/24/21 12:54 AM, Xu Yilun wrote:
> On Wed, Jun 23, 2021 at 11:24:04AM -0700, trix@xxxxxxxxxx wrote:
> > From: Tom Rix <trix@xxxxxxxxxx>
> >
> > An FPGA manager should not be required to provide a
> > write_init() op if there is nothing for it do.
> > So add a wrapper and move the op checking.
> > Default to success.
> >
> > Signed-off-by: Tom Rix <trix@xxxxxxxxxx>
> > ---
> >   drivers/fpga/fpga-mgr.c | 14 +++++++++++---
> >   1 file changed, 11 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/fpga/fpga-mgr.c b/drivers/fpga/fpga-mgr.c
> > index ecb4c3c795fa5..87bbb940c9504 100644
> > --- a/drivers/fpga/fpga-mgr.c
> > +++ b/drivers/fpga/fpga-mgr.c
> > @@ -69,6 +69,14 @@ void fpga_image_info_free(struct fpga_image_info *info)
> >   }
> >   EXPORT_SYMBOL_GPL(fpga_image_info_free);
> >   
> > +static int fpga_mgr_write_init(struct fpga_manager *mgr,
> > +			       struct fpga_image_info *info,
> > +			       const char *buf, size_t count)
> > +{
> > +	if (mgr->mops && mgr->mops->write_init)
> Maybe we don't have to check mgr->mops, it is already checked on
> creation.

The check was on purpose because my earlier patchset responding to 
problems with sec-mgr.

Focusing on Greg's comment that why can't sec-mgr be done with existing 
code,

I think the sec-mgr can be folded into the exiting fpga-mgr via a new 
set of ops.

the 'generalize fpga_mgr_load' patchset set has two sets of ops,

one for existing partial reconfiguration

and one for reimaging the whole board, what the sec-mgr is doing.

Since dfl has the only instance of need, it would have the only reimage ops.

The check at creation has been deferred to at use.

other targets could have null ops.

Having maybe null ops means the wrappers need to check.

Here is a ref to the earlier patchset

https://lore.kernel.org/linux-fpga/20210524162721.2220782-1-trix@xxxxxxxxxx/

I'll respin 'generalize fpga_mgr_load' within the context this patchset 
to give you some more context.

It will test is the check is needed and give folks a chance to comment 
if this a way sec-mgr should go.

Tom


> The same concern to all the following patches.
>
> Thanks,
> Yilun
>
> > +		return  mgr->mops->write_init(mgr, info, buf, count);
> > +	return 0;
> > +}
> >   /*
> >    * Call the low level driver's write_init function.  This will do the
> >    * device-specific things to get the FPGA into the state where it is ready to
> > @@ -83,9 +91,9 @@ static int fpga_mgr_write_init_buf(struct fpga_manager *mgr,
> >   
> >   	mgr->state = FPGA_MGR_STATE_WRITE_INIT;
> >   	if (!mgr->mops->initial_header_size)
> > -		ret = mgr->mops->write_init(mgr, info, NULL, 0);
> > +		ret = fpga_mgr_write_init(mgr, info, NULL, 0);
> >   	else
> > -		ret = mgr->mops->write_init(
> > +		ret = fpga_mgr_write_init(
> >   		    mgr, info, buf, min(mgr->mops->initial_header_size, count));
> >   
> >   	if (ret) {
> > @@ -569,7 +577,7 @@ struct fpga_manager *fpga_mgr_create(struct device *parent, const char *name,
> >   	int id, ret;
> >   
> >   	if (!mops || !mops->write_complete || !mops->state ||
> > -	    !mops->write_init || (!mops->write && !mops->write_sg) ||
> > +	    (!mops->write && !mops->write_sg) ||
> >   	    (mops->write && mops->write_sg)) {
> >   		dev_err(parent, "Attempt to register without fpga_manager_ops\n");
> >   		return NULL;
> > --
> > 2.26.3