[PATCH i-g-t] configure.ac: correctly manage DRM_INTEL_{CFLAGS, LIBS}

[PATCH i-g-t] configure.ac: correctly manage DRM_INTEL_{CFLAGS, LIBS}

[Emil Velikov]

From: Emil Velikov <emil.velikov@collabora.com>

Currently the latter is only set when using --enable-intel.

Whereas for the CFLAGS with "enable", it's set by PKG_CHECK_MODULES
and it's set locally for "disable". Yet, in either case it's not
propagated through, this one can get a range of build issues regardless of
the actual state of the toggle.

Cc: Brian Starkey <brian.starkey@arm.com>
Cc: Robert Foss <robert.foss@collabora.com>
Reported-by: Brian Starkey <brian.starkey@arm.com>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
---
If interested, one can do a fine grained addition in the respective
files, but this is the quickest fix ;-)

On a related note: seems like the nouveau and vc4 CFLAGS are _not_
propagated either. Any volunteers ?
---
 configure.ac | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/configure.ac b/configure.ac
index 735cfd5..1c747b3 100644
--- a/configure.ac
+++ b/configure.ac
@@ -178,12 +178,15 @@ fi
 if test "x$INTEL" = xyes; then
 	PKG_CHECK_MODULES(DRM_INTEL, [libdrm_intel >= 2.4.64])
 	AC_DEFINE(HAVE_LIBDRM_INTEL, 1, [Have intel support])
-	DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
-	AC_SUBST([DRM_LIBS])
 else
 	DRM_INTEL_CFLAGS=$(top_srcdir)/lib/stubs/drm/
-	AC_SUBST([DRM_INTEL_CFLAGS])
+	DRM_INTEL_LIBS=
 fi
+DRM_CFLAGS="$DRM_CFLAGS $DRM_INTEL_CFLAGS"
+DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
+AC_SUBST([DRM_CFLAGS])
+AC_SUBST([DRM_LIBS])
+
 AM_CONDITIONAL(HAVE_LIBDRM_INTEL, [test "x$INTEL" = xyes])
 
 # for dma-buf tests
-- 
2.9.3

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH i-g-t v2] configure.ac: correctly manage DRM_INTEL_{CFLAGS, LIBS}

[Emil Velikov]

From: Emil Velikov <emil.velikov@collabora.com>

Currently the latter is only set when using --enable-intel.

Whereas for the CFLAGS: if we "enable" PKG_CHECK_MODULES sets the
variable, while for "disable" we do it locally. In either case the
CFLAGS is not propagated through, this one can get build issues
regardless of the actual state of the toggle.

v2: Add -I for the include directive and correctly propagate
$(top_srcdir).

Cc: Brian Starkey <brian.starkey@arm.com>
Cc: Robert Foss <robert.foss@collabora.com>
Reported-by: Brian Starkey <brian.starkey@arm.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
---
This time fully tested:
 - enable/disable/auto with and w/o libdrm_intel :-)

Did not see much point in splitting the v2 only changes to 1/2 and keep
this as 2/2. I don't mind either way though :-)
---
 configure.ac | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/configure.ac b/configure.ac
index 735cfd5..e181c83 100644
--- a/configure.ac
+++ b/configure.ac
@@ -178,12 +178,15 @@ fi
 if test "x$INTEL" = xyes; then
 	PKG_CHECK_MODULES(DRM_INTEL, [libdrm_intel >= 2.4.64])
 	AC_DEFINE(HAVE_LIBDRM_INTEL, 1, [Have intel support])
-	DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
-	AC_SUBST([DRM_LIBS])
 else
-	DRM_INTEL_CFLAGS=$(top_srcdir)/lib/stubs/drm/
-	AC_SUBST([DRM_INTEL_CFLAGS])
+	DRM_INTEL_CFLAGS=-I$\(top_srcdir\)/lib/stubs/drm/
+	DRM_INTEL_LIBS=
 fi
+DRM_CFLAGS="$DRM_CFLAGS $DRM_INTEL_CFLAGS"
+DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
+AC_SUBST([DRM_CFLAGS])
+AC_SUBST([DRM_LIBS])
+
 AM_CONDITIONAL(HAVE_LIBDRM_INTEL, [test "x$INTEL" = xyes])
 
 # for dma-buf tests
-- 
2.9.3

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH i-g-t v2] configure.ac: correctly manage DRM_INTEL_{CFLAGS, LIBS}

[Robert Foss]

Thanks Emil!
This looks good to me.

Signed-off-by: Robert Foss <robert.foss@collabora.com>

On 2016-10-26 12:18 PM, Emil Velikov wrote:
> From: Emil Velikov <emil.velikov@collabora.com>
>
> Currently the latter is only set when using --enable-intel.
>
> Whereas for the CFLAGS: if we "enable" PKG_CHECK_MODULES sets the
> variable, while for "disable" we do it locally. In either case the
> CFLAGS is not propagated through, this one can get build issues
> regardless of the actual state of the toggle.
>
> v2: Add -I for the include directive and correctly propagate
> $(top_srcdir).
>
> Cc: Brian Starkey <brian.starkey@arm.com>
> Cc: Robert Foss <robert.foss@collabora.com>
> Reported-by: Brian Starkey <brian.starkey@arm.com>
> Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
> ---
> This time fully tested:
>  - enable/disable/auto with and w/o libdrm_intel :-)
>
> Did not see much point in splitting the v2 only changes to 1/2 and keep
> this as 2/2. I don't mind either way though :-)
> ---
>  configure.ac | 11 +++++++----
>  1 file changed, 7 insertions(+), 4 deletions(-)
>
> diff --git a/configure.ac b/configure.ac
> index 735cfd5..e181c83 100644
> --- a/configure.ac
> +++ b/configure.ac
> @@ -178,12 +178,15 @@ fi
>  if test "x$INTEL" = xyes; then
>  	PKG_CHECK_MODULES(DRM_INTEL, [libdrm_intel >= 2.4.64])
>  	AC_DEFINE(HAVE_LIBDRM_INTEL, 1, [Have intel support])
> -	DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
> -	AC_SUBST([DRM_LIBS])
>  else
> -	DRM_INTEL_CFLAGS=$(top_srcdir)/lib/stubs/drm/
> -	AC_SUBST([DRM_INTEL_CFLAGS])
> +	DRM_INTEL_CFLAGS=-I$\(top_srcdir\)/lib/stubs/drm/
> +	DRM_INTEL_LIBS=
>  fi
> +DRM_CFLAGS="$DRM_CFLAGS $DRM_INTEL_CFLAGS"
> +DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
> +AC_SUBST([DRM_CFLAGS])
> +AC_SUBST([DRM_LIBS])
> +
>  AM_CONDITIONAL(HAVE_LIBDRM_INTEL, [test "x$INTEL" = xyes])
>
>  # for dma-buf tests
>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH i-g-t v2] configure.ac: correctly manage DRM_INTEL_{CFLAGS, LIBS}

[Brian Starkey]

Hi Emil,

On Wed, Oct 26, 2016 at 05:18:47PM +0100, Emil Velikov wrote:
>From: Emil Velikov <emil.velikov@collabora.com>
>
>Currently the latter is only set when using --enable-intel.
>
>Whereas for the CFLAGS: if we "enable" PKG_CHECK_MODULES sets the
>variable, while for "disable" we do it locally. In either case the
>CFLAGS is not propagated through, this one can get build issues
>regardless of the actual state of the toggle.
>
>v2: Add -I for the include directive and correctly propagate
>$(top_srcdir).
>
>Cc: Brian Starkey <brian.starkey@arm.com>
>Cc: Robert Foss <robert.foss@collabora.com>
>Reported-by: Brian Starkey <brian.starkey@arm.com>
>Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
>---

This fixes my include path problem, so:

Tested-by: Brian Starkey <brian.starkey@arm.com>

Thanks very much for your help,
-Brian

>This time fully tested:
> - enable/disable/auto with and w/o libdrm_intel :-)
>
>Did not see much point in splitting the v2 only changes to 1/2 and keep
>this as 2/2. I don't mind either way though :-)
>---
> configure.ac | 11 +++++++----
> 1 file changed, 7 insertions(+), 4 deletions(-)
>
>diff --git a/configure.ac b/configure.ac
>index 735cfd5..e181c83 100644
>--- a/configure.ac
>+++ b/configure.ac
>@@ -178,12 +178,15 @@ fi
> if test "x$INTEL" = xyes; then
> 	PKG_CHECK_MODULES(DRM_INTEL, [libdrm_intel >= 2.4.64])
> 	AC_DEFINE(HAVE_LIBDRM_INTEL, 1, [Have intel support])
>-	DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
>-	AC_SUBST([DRM_LIBS])
> else
>-	DRM_INTEL_CFLAGS=$(top_srcdir)/lib/stubs/drm/
>-	AC_SUBST([DRM_INTEL_CFLAGS])
>+	DRM_INTEL_CFLAGS=-I$\(top_srcdir\)/lib/stubs/drm/
>+	DRM_INTEL_LIBS=
> fi
>+DRM_CFLAGS="$DRM_CFLAGS $DRM_INTEL_CFLAGS"
>+DRM_LIBS="$DRM_LIBS $DRM_INTEL_LIBS"
>+AC_SUBST([DRM_CFLAGS])
>+AC_SUBST([DRM_LIBS])
>+
> AM_CONDITIONAL(HAVE_LIBDRM_INTEL, [test "x$INTEL" = xyes])
>
> # for dma-buf tests
>-- 
>2.9.3
>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH i-g-t v2] configure.ac: correctly manage DRM_INTEL_{CFLAGS, LIBS}

[Robert Foss]

Reviewed-by: Robert Foss <robert.foss@collabora.com>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx