crypto/scheduler: fix header install with meson
Checks
Commit Message
The headers variable is only used in lib/meson.build context.
For drivers, an explicit install_headers() is necessary.
Reproduced while compiling the l2fwd-crypto example out of meson
(which can be done by adding it in devtools/test-meson-builds.sh
examples build test).
Fixes: cd2b6458a1cb ("crypto/scheduler: add in meson build")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
---
drivers/crypto/scheduler/meson.build | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Comments
On Thu, Sep 17, 2020 at 05:09:17PM +0200, David Marchand wrote:
> The headers variable is only used in lib/meson.build context.
> For drivers, an explicit install_headers() is necessary.
>
> Reproduced while compiling the l2fwd-crypto example out of meson
> (which can be done by adding it in devtools/test-meson-builds.sh
> examples build test).
>
> Fixes: cd2b6458a1cb ("crypto/scheduler: add in meson build")
> Cc: stable@dpdk.org
>
> Signed-off-by: David Marchand <david.marchand@redhat.com>
> ---
> drivers/crypto/scheduler/meson.build | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
Originally I only put the "headers" variable in the "lib" folder
meson.build because it was not expected that many drivers should have
private headers. However, perhaps it's worthwhile adding it now:
a) because private headers, while not common, are not unusual
b) for consistency and ease of use.
Thoughts?
/Bruce
On Thu, Sep 17, 2020 at 05:09:17PM +0200, David Marchand wrote:
> The headers variable is only used in lib/meson.build context.
> For drivers, an explicit install_headers() is necessary.
>
> Reproduced while compiling the l2fwd-crypto example out of meson
> (which can be done by adding it in devtools/test-meson-builds.sh
> examples build test).
>
> Fixes: cd2b6458a1cb ("crypto/scheduler: add in meson build")
> Cc: stable@dpdk.org
>
> Signed-off-by: David Marchand <david.marchand@redhat.com>
> ---
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
On Thu, Sep 17, 2020 at 5:29 PM Bruce Richardson
<bruce.richardson@intel.com> wrote:
>
> On Thu, Sep 17, 2020 at 05:09:17PM +0200, David Marchand wrote:
> > The headers variable is only used in lib/meson.build context.
> > For drivers, an explicit install_headers() is necessary.
> >
> > Reproduced while compiling the l2fwd-crypto example out of meson
> > (which can be done by adding it in devtools/test-meson-builds.sh
> > examples build test).
> >
> > Fixes: cd2b6458a1cb ("crypto/scheduler: add in meson build")
> > Cc: stable@dpdk.org
> >
> > Signed-off-by: David Marchand <david.marchand@redhat.com>
> > ---
> > drivers/crypto/scheduler/meson.build | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> Originally I only put the "headers" variable in the "lib" folder
> meson.build because it was not expected that many drivers should have
> private headers. However, perhaps it's worthwhile adding it now:
> a) because private headers, while not common, are not unusual
> b) for consistency and ease of use.
>
> Thoughts?
Yep, I can do it in a followup patch now, still looking at your RFC
series though.. :-)
On Thu, Sep 17, 2020 at 05:33:23PM +0200, David Marchand wrote:
> On Thu, Sep 17, 2020 at 5:29 PM Bruce Richardson
> <bruce.richardson@intel.com> wrote:
> >
> > On Thu, Sep 17, 2020 at 05:09:17PM +0200, David Marchand wrote:
> > > The headers variable is only used in lib/meson.build context.
> > > For drivers, an explicit install_headers() is necessary.
> > >
> > > Reproduced while compiling the l2fwd-crypto example out of meson
> > > (which can be done by adding it in devtools/test-meson-builds.sh
> > > examples build test).
> > >
> > > Fixes: cd2b6458a1cb ("crypto/scheduler: add in meson build")
> > > Cc: stable@dpdk.org
> > >
> > > Signed-off-by: David Marchand <david.marchand@redhat.com>
> > > ---
> > > drivers/crypto/scheduler/meson.build | 2 +-
> > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > Originally I only put the "headers" variable in the "lib" folder
> > meson.build because it was not expected that many drivers should have
> > private headers. However, perhaps it's worthwhile adding it now:
> > a) because private headers, while not common, are not unusual
> > b) for consistency and ease of use.
> >
> > Thoughts?
>
> Yep, I can do it in a followup patch now, still looking at your RFC
> series though.. :-)
>
No rush, what we have now has worked reasonably well for the last few
years! :-)
If you don't get to it, I can always look at it post-19.11.
@@ -13,7 +13,7 @@ sources = files(
'scheduler_roundrobin.c',
)
-headers = files(
+install_headers(
'rte_cryptodev_scheduler.h',
'rte_cryptodev_scheduler_operations.h',
)