[dpdk-dev] [PATCH v2] compress/qat: enable dynamic huffman encoding
Trahe, Fiona
fiona.trahe at intel.com
Tue Oct 9 16:09:43 CEST 2018
Hi Akhil,
> -----Original Message-----
> From: Akhil Goyal [mailto:akhil.goyal at nxp.com]
> Sent: Tuesday, October 9, 2018 12:02 PM
> To: Trahe, Fiona <fiona.trahe at intel.com>; dev at dpdk.org
> Cc: Jozwiak, TomaszX <tomaszx.jozwiak at intel.com>
> Subject: Re: [dpdk-dev] [PATCH v2] compress/qat: enable dynamic huffman encoding
>
>
>
> On 10/5/2018 9:48 PM, Fiona Trahe wrote:
> > Enable dynamic huffman encoding in the QAT comp PMD.
> >
> > Signed-off-by: Tomasz Jozwiak <tomaszx.jozwiak at intel.com>
> > Signed-off-by: Fiona Trahe <fiona.trahe at intel.com>
> > ---
> > v2 changes:
> > - allocate 2 buffers per intermediate buffer sgl
> > - Compile out trace for debugging intermediate buffers
> >
> > Depends on:
> > https://patches.dpdk.org/patch/45794/
> >
> //snip
> > +#if 1
> > + QAT_LOG(DEBUG, " : phys addr of sgl[%i] in array_of_pointers"
> > + "= %lx", i, array_of_pointers->pointer[i]);
> > + QAT_LOG(DEBUG, " : virt address of sgl[%i] = %p", i, sgl);
> > + QAT_LOG(DEBUG, " : sgl->buffers[0].addr = %lx, len=%d",
> > + sgl->buffers[0].addr, sgl->buffers[0].len);
> > + QAT_LOG(DEBUG, " : sgl->buffers[1].addr = %lx, len=%d",
> > + sgl->buffers[1].addr, sgl->buffers[1].len);
> > +#endif
> > + }
> > +#if 0
> > + QAT_DP_HEXDUMP_LOG(DEBUG, "IM buffer memzone:", mz_start, 1504);
> > +#endif
>
> I believe #if 0 and #if 1 should be replaced with some debug macros, or
> else dynamic logging shall take care of that.
[Fiona] These are not on the data path, so I didn't think it appropriate to wrap with the DP log level compile option.
And they generate a lot of trace, so make it impossible to see the other debug trace enabled with
dynamic logging using level 8/DEBUG. Really a level 9 would be useful here, but I think this is overkill.
I could have removed this code, but it's handy for debugging.
How about if I add a local QAT define for it?
#define QAT_IM_BUFFER_DEBUG 0
#if QAT_IM_BUFFER_DEBUG
...
#endif
More information about the dev
mailing list