Re: [dpdk-dev] [PATCH v3] examples/vhost: fix ioat dependency issue

["Jiang, Cheng1" <cheng1.jiang@intel.com>]

Submitted v4 patch as per Bruce's suggestion.

> -----Original Message-----
> From: Bruce Richardson <bruce.richardson@intel.com>
> Sent: Thursday, November 12, 2020 8:02 PM
> To: Jiang, Cheng1 <cheng1.jiang@intel.com>
> Cc: David Marchand <david.marchand@redhat.com>; Maxime Coquelin
> <maxime.coquelin@redhat.com>; Xia, Chenbo <chenbo.xia@intel.com>;
> dev <dev@dpdk.org>; Fu, Patrick <patrick.fu@intel.com>; Yang, YvonneX
> <yvonnex.yang@intel.com>; Hu, Jiayu <jiayu.hu@intel.com>
> Subject: Re: [dpdk-dev] [PATCH v3] examples/vhost: fix ioat dependency
> issue
> 
> On Thu, Nov 12, 2020 at 11:29:56AM +0000, Jiang, Cheng1 wrote:
> > Hi Bruce,
> >
> > > -----Original Message-----
> > > From: Bruce Richardson <bruce.richardson@intel.com>
> > > Sent: Thursday, November 12, 2020 6:28 PM
> > > To: David Marchand <david.marchand@redhat.com>
> > > Cc: Jiang, Cheng1 <cheng1.jiang@intel.com>; Maxime Coquelin
> > > <maxime.coquelin@redhat.com>; Xia, Chenbo <chenbo.xia@intel.com>;
> > > dev <dev@dpdk.org>; Fu, Patrick <patrick.fu@intel.com>; Yang, YvonneX
> > > <yvonnex.yang@intel.com>; Hu, Jiayu <jiayu.hu@intel.com>
> > > Subject: Re: [dpdk-dev] [PATCH v3] examples/vhost: fix ioat dependency
> > > issue
> > >
> > > On Thu, Nov 12, 2020 at 10:36:50AM +0100, David Marchand wrote:
> > > > On Thu, Nov 12, 2020 at 8:30 AM Cheng Jiang <Cheng1.jiang@intel.com>
> > > wrote:
> > > > >
> > > > > Fix vhost-switch compiling issue when ioat dependency is missing.
> > > > > Change 'RTE_x86' check into 'RTE_RAW_IOAT' check in meson build
> file
> > > > > and update Makefile. Clean some codes.
> > > > >
> > > > > Fixes: abec60e7115d ("examples/vhost: support vhost async data
> > > > > path")
> > > > > Fixes: 3a04ecb21420 ("examples/vhost: add async vhost args parsing")
> > > > >
> > > > > Signed-off-by: Cheng Jiang <Cheng1.jiang@intel.com>
> > > > > ---
> > > > > v3:
> > > > >  * Added fixes lines in commit log.
> > > > >
> > > > > v2:
> > > > >  * Cleaned some codes
> > > > >  * Changed RTE_RAW_IOAT check method in Makefile
> > > > >  * Added ioat function definition when RTE_RAW_IOAT is missing
> > > > >
> > > > >  examples/vhost/Makefile    |  5 +++++
> > > > >  examples/vhost/ioat.h      | 32 +++++++++++++++++++++++++-------
> > > > >  examples/vhost/main.c      | 22 +++++++++++-----------
> > > > >  examples/vhost/meson.build |  2 +-
> > > > >  4 files changed, 42 insertions(+), 19 deletions(-)
> > > > >
> > > > > diff --git a/examples/vhost/Makefile b/examples/vhost/Makefile
> index
> > > > > cec59d0e0..cbe56f742 100644
> > > > > --- a/examples/vhost/Makefile
> > > > > +++ b/examples/vhost/Makefile
> > > > > @@ -28,6 +28,11 @@ CFLAGS += -O3 $(shell $(PKGCONF) --cflags
> > > > > libdpdk)  LDFLAGS_SHARED = $(shell $(PKGCONF) --libs libdpdk)
> > > > > LDFLAGS_STATIC = $(shell $(PKGCONF) --static --libs libdpdk)
> > > > >
> > > > > +HAS_RAW_IOAT=$(shell echo RTE_RAW_IOAT | $(CPP) $(CFLAGS) -
> P - |
> > > > > +tail -1) ifeq ($(HAS_RAW_IOAT), 1) SRCS-y += ioat.c endif
> > > > > +
> > > > >  CFLAGS += -DALLOW_EXPERIMENTAL_API
> > > > >
> > > > >  build/$(APP)-shared: $(SRCS-y) Makefile $(PC_FILE) | build
> > >
> > > <snip>
> > >
> > > > I'll let Bruce and Maxime have the last word on this patch.
> > > > But at least it works for me,
> > > > Tested-by: David Marchand <david.marchand@redhat.com>
> > > >
> > > I don't have a really strong objection to this, but I still would rather see
> the
> > > ioat detection done in the C code only rather than in the Makefile.
> > > I'm concerned about us adding too much complexity into our makefiles,
> so
> > > would like to keep them simple as much as possible.
> > >
> > > Therefore, I'd like to see ioat.c always included in SRCS-y, and ioat.c just
> have
> > > #ifdef RTE_RAW_IOAT to block out any ioat-dependent code.
> > >
> > > /Bruce
> >
> > Tomorrow is the deadline for RC4, it's a little bit rush to prepare a new
> framework to cover the conditional compiling. Considering we have test
> efforts and CI check already on-going based on current patch, I would prefer
> to keep current code structure. Plus, pure C code change could also
> introduce more readability problem when we have more than one async
> callback implementations.
> >
> 
> I don't see any readability problems at all, in fact the result is cleaner
> and shorter IMHO - especially the Makefile. I just tested with the
> following changes applied on top of your patch, and all seemed to build ok.
> 
> Do you see any implementation problems with the below solution?
> 
> /Bruce
> 
> diff --git a/examples/vhost/Makefile b/examples/vhost/Makefile
> index cbe56f742..8c969caaa 100644
> --- a/examples/vhost/Makefile
> +++ b/examples/vhost/Makefile
> @@ -5,7 +5,7 @@
>  APP = vhost-switch
> 
>  # all source are stored in SRCS-y
> -SRCS-y := main.c virtio_net.c
> +SRCS-y := main.c virtio_net.c ioat.c
> 
>  # Build using pkg-config variables if possible
>  ifneq ($(shell pkg-config --exists libdpdk && echo 0),0)
> @@ -28,11 +28,6 @@ CFLAGS += -O3 $(shell $(PKGCONF) --cflags libdpdk)
>  LDFLAGS_SHARED = $(shell $(PKGCONF) --libs libdpdk)
>  LDFLAGS_STATIC = $(shell $(PKGCONF) --static --libs libdpdk)
> 
> -HAS_RAW_IOAT=$(shell echo RTE_RAW_IOAT | $(CPP) $(CFLAGS) -P - | tail
> -1)
> -ifeq ($(HAS_RAW_IOAT), 1)
> -SRCS-y += ioat.c
> -endif
> -
>  CFLAGS += -DALLOW_EXPERIMENTAL_API
> 
>  build/$(APP)-shared: $(SRCS-y) Makefile $(PC_FILE) | build
> diff --git a/examples/vhost/ioat.c b/examples/vhost/ioat.c
> index b2c74f653..7c1a4f40d 100644
> --- a/examples/vhost/ioat.c
> +++ b/examples/vhost/ioat.c
> @@ -8,6 +8,8 @@
>  #include "ioat.h"
>  #include "main.h"
> 
> +#ifdef RTE_RAW_IOAT
> +
>  struct dma_for_vhost dma_bind[MAX_VHOST_DEVICE];
> 
>  struct packet_tracker {
> @@ -199,3 +201,5 @@ ioat_check_completed_copies_cb(int vid, uint16_t
> queue_id,
>         /* Opaque data is not supported */
>         return -1;
>  }
> +
> +#endif /* RTE_RAW_IOAT */