Re: [PATCH] staging: qlge: Remove macro FILL_SEG

[Sumitra Sharma]

On Thu, Apr 06, 2023 at 04:57:44PM +0200, Greg Kroah-Hartman wrote:
> On Thu, Apr 06, 2023 at 07:46:44AM -0700, Sumitra Sharma wrote:
> > On Wed, Apr 05, 2023 at 06:21:57PM +0200, Simon Horman wrote:
> > > On Wed, Apr 05, 2023 at 08:06:27AM -0700, Sumitra Sharma wrote:
> > > > Remove macro FILL_SEG to fix the checkpatch warning:
> > > > 
> > > > WARNING: Macros with flow control statements should be avoided
> > > > 
> > > > Macros with flow control statements must be avoided as they
> > > > break the flow of the calling function and make it harder to
> > > > test the code.
> > > > 
> > > > Replace all FILL_SEG() macro calls with:
> > > > 
> > > > err = err || qlge_fill_seg_(...);
> > > 
> > > Perhaps I'm missing the point here.
> > > But won't this lead to err always either being true or false (1 or 0).
> > > Rather than the current arrangement where err can be
> > > either 0 or a negative error value, such as -EINVAL.
> > >
> > 
> > Hi Simon
> > 
> > 
> > Thank you for the point you mentioned which I missed while working on this
> > patch. 
> > 
> > However, after thinking on it, I am still not able to get any fix to this
> > except that we can possibly implement the Ira's solution here which is:
> > 
> > https://lore.kernel.org/outreachy/64154d438f0c8_28ae5229421@iweiny-mobl.notmuch/
> > 
> > Although we have to then deal with 40 lines of ifs.
> 
> Which implies that the current solution is the best one, so I would
> recommend just leaving it as-is.
>

Hi greg

Before leaving it I would like to know your opinion on the solution Dan is offering.


Regards
Sumitra

> Remember, checkpatch.pl is a tool to provide hints, it does not have
> much context, if any, to determine if it's hints actually make sense.
> 
> thanks,
> 
> greg k-h