All of lore.kernel.org
 help / color / mirror / Atom feed
* intel-gpu-tools: --disable-tests option should make cairo-check obsolete
@ 2013-08-14 20:33 Sedat Dilek
       [not found] ` <20130814210330.GB6077@nuc-i3427.alporthouse.com>
  0 siblings, 1 reply; 3+ messages in thread
From: Sedat Dilek @ 2013-08-14 20:33 UTC (permalink / raw)
  To: intel-gfx

Hi,

I am here on Ubuntu/precise and wanted to avoid to upgrade to a higher
cairo-version.

[ configure.ac ]
...
# for testdisplay
PKG_CHECK_MODULES(CAIRO, [cairo >= 1.12.0])
PKG_CHECK_MODULES(LIBUDEV, [libudev], [udev=yes], [udev=no])
if test x"$udev" = xyes; then
AC_DEFINE(HAVE_UDEV,1,[Enable udev-based monitor hotplug detection])
fi
PKG_CHECK_MODULES(GLIB, glib-2.0)
...

$ find ./ -name 'testdisplay*'
./tests/testdisplay.h
./tests/testdisplay.c
./tests/testdisplay_hotplug.c

So, testdisplay is a member of tests, with --disable-tests I expect
that any checks will be dropped - means no cairo-update for me.

Can someone with better autotools skillz fix that?
Thanks in advance.

Regards,
- Sedat -

P.S.: BTW, I am only interested in building the tools only.

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: intel-gpu-tools: --disable-tests option should make cairo-check obsolete
       [not found] ` <20130814210330.GB6077@nuc-i3427.alporthouse.com>
@ 2013-08-14 21:14   ` Sedat Dilek
       [not found]     ` <20130814212526.GC6077@nuc-i3427.alporthouse.com>
  0 siblings, 1 reply; 3+ messages in thread
From: Sedat Dilek @ 2013-08-14 21:14 UTC (permalink / raw)
  To: Chris Wilson, Sedat Dilek, intel-gfx

On Wed, Aug 14, 2013 at 11:03 PM, Chris Wilson <chris@chris-wilson.co.uk> wrote:
> On Wed, Aug 14, 2013 at 10:33:42PM +0200, Sedat Dilek wrote:
>> Hi,
>>
>> I am here on Ubuntu/precise and wanted to avoid to upgrade to a higher
>> cairo-version.
>
> Daniel's position is that he wants to make it hard for QA to build the
> tests incorrectly and so skip half of them. I think defaulting to
> erroring out at configure if cairo is too old, but allowing an explicit
> --disable-cairo should make all us happy.

Hmm, just curious how this should work?

Beyond testdisplay... kms_flip and kms_render tests require cairo.h
include - not sure which version of cairo is minimum.
So, --disable-cairo configure-option should disable above tests?

[ tests/Makefile.am ]
...
testdisplay_SOURCES = \
        testdisplay.c \
        testdisplay.h \
        testdisplay_hotplug.c \
        $(NULL)

TESTS_progs += testdisplay
LDADD += $(CAIRO_LIBS) $(LIBUDEV_LIBS) $(GLIB_LIBS)
AM_CFLAGS += $(CAIRO_CFLAGS) $(LIBUDEV_CFLAGS) $(GLIB_CFLAGS)
...

$ grep -i cairo.h -nr ./
./lib/drmtest.h:34:#include <cairo.h>
./tests/kms_flip.c:28:#include <cairo.h>
./tests/kms_render.c:27:#include <cairo.h>
./tests/testdisplay.c:52:#include <cairo.h>

Thanks in advance.

- Sedat -

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: intel-gpu-tools: --disable-tests option should make cairo-check obsolete
       [not found]     ` <20130814212526.GC6077@nuc-i3427.alporthouse.com>
@ 2013-08-14 21:41       ` Sedat Dilek
  0 siblings, 0 replies; 3+ messages in thread
From: Sedat Dilek @ 2013-08-14 21:41 UTC (permalink / raw)
  To: Chris Wilson, Sedat Dilek, intel-gfx

On Wed, Aug 14, 2013 at 11:25 PM, Chris Wilson <chris@chris-wilson.co.uk> wrote:
> On Wed, Aug 14, 2013 at 11:14:05PM +0200, Sedat Dilek wrote:
>> On Wed, Aug 14, 2013 at 11:03 PM, Chris Wilson <chris@chris-wilson.co.uk> wrote:
>> > On Wed, Aug 14, 2013 at 10:33:42PM +0200, Sedat Dilek wrote:
>> >> Hi,
>> >>
>> >> I am here on Ubuntu/precise and wanted to avoid to upgrade to a higher
>> >> cairo-version.
>> >
>> > Daniel's position is that he wants to make it hard for QA to build the
>> > tests incorrectly and so skip half of them. I think defaulting to
>> > erroring out at configure if cairo is too old, but allowing an explicit
>> > --disable-cairo should make all us happy.
>>
>> Hmm, just curious how this should work?
>>
>> Beyond testdisplay... kms_flip and kms_render tests require cairo.h
>> include - not sure which version of cairo is minimum.
>> So, --disable-cairo configure-option should disable above tests?
>>
>> [ tests/Makefile.am ]
>> ...
>> testdisplay_SOURCES = \
>>         testdisplay.c \
>>         testdisplay.h \
>>         testdisplay_hotplug.c \
>>         $(NULL)
>>
>> TESTS_progs += testdisplay
>> LDADD += $(CAIRO_LIBS) $(LIBUDEV_LIBS) $(GLIB_LIBS)
>> AM_CFLAGS += $(CAIRO_CFLAGS) $(LIBUDEV_CFLAGS) $(GLIB_CFLAGS)
>> ...
>>
>> $ grep -i cairo.h -nr ./
>> ./lib/drmtest.h:34:#include <cairo.h>
>> ./tests/kms_flip.c:28:#include <cairo.h>
>> ./tests/kms_render.c:27:#include <cairo.h>
>> ./tests/testdisplay.c:52:#include <cairo.h>
>
> First you would need to split the cairo portion of drmtest.[ch] into its
> own files and then conditionally compile those, along with the cairo
> dependent tests.

As said I am not an autotools expert.

But I was interested in the tools not building the tests.
That's why I asked why I am forced to upgrade to a higher cairo release.

Why does tools require $(CAIRO_CFLAGS) and  $(CAIRO_LIBS)?
Hmm, looks like intel_perf_counters and intel_l3_parity do...

tools/Makefile.am:40:AM_CFLAGS = $(DRM_CFLAGS) $(PCIACCESS_CFLAGS)
$(CWARNFLAGS) $(CAIRO_CFLAGS)
tools/Makefile.am:41:LDADD = $(top_builddir)/lib/libintel_tools.la
$(DRM_LIBS) $(PCIACCESS_LIBS) $(CAIRO_LIBS)

tools/intel_perf_counters.c:47:#include "drmtest.h"
tools/intel_l3_parity.c:38:#include "drmtest.h"

--disable-cairo might be a good idea, but I cannot help here.

- Sedat -

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2013-08-14 21:41 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-08-14 20:33 intel-gpu-tools: --disable-tests option should make cairo-check obsolete Sedat Dilek
     [not found] ` <20130814210330.GB6077@nuc-i3427.alporthouse.com>
2013-08-14 21:14   ` Sedat Dilek
     [not found]     ` <20130814212526.GC6077@nuc-i3427.alporthouse.com>
2013-08-14 21:41       ` Sedat Dilek

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.