[PATCH] include mount.h instead of include linux/fs.h

This should fix pgoff_t no defined with some linux headers.

Ok to commit?

Thanks,
Kay

Index: configure.in
===================================================================
RCS file: /cvs/hal/hal/configure.in,v
retrieving revision 1.87
diff -u -r1.87 configure.in
--- configure.in	29 Jun 2005 20:21:04 -0000	1.87
+++ configure.in	1 Jul 2005 10:27:38 -0000
 <at>  <at>  -294,8 +294,7  <at>  <at> 
 AC_MSG_CHECKING(for BLKGETSIZE64)
 AC_TRY_COMPILE(
 [#include <sys/types.h>
-#include <linux/ioctl.h>
-#include <linux/fs.h>],
+#include <sys/mount.h>],
 [int main(int argc, char *argv[])
 {
 	int fd = 0;
Index: hald/linux2/addons/addon-storage.c
===================================================================
RCS file: /cvs/hal/hal/hald/linux2/addons/addon-storage.c,v
retrieving revision 1.6
diff -u -r1.6 addon-storage.c

Re: [PATCH] include mount.h instead of include linux/fs.h

On Fri, 2005-07-01 at 12:54 +0200, Kay Sievers wrote:
> This should fix pgoff_t no defined with some linux headers.
> 
> Ok to commit?

Go for it.

Thanks,
David

Re: [geromanas <at> mailas.com: Bug#315757: libhal-storage0: libhal-storage.c:hal_volume_from_udi():vol->disk_type is not set correctly for DVD+R and DVD+RW disks (typos)]

Hi,

I actually committed a fix for this a few days ago.

 http://cvs.freedesktop.org/hal/hal/libhal-storage/libhal-storage.c?r1=1.15&r2=1.16

Cheers,
David

h-d-m handle empty properties

Want to get rid of the patches I found in the SUSE package... :)
This makes hal-device-manager handle empty properties where a value is
expected.

Ok to commit?

Thanks,
Kay

Index: tools/device-manager/Device.py
===================================================================
RCS file: /cvs/hal/hal/tools/device-manager/Device.py,v
retrieving revision 1.2
diff -u -p -r1.2 Device.py
--- tools/device-manager/Device.py	22 Dec 2003 04:03:10 -0000	1.2
+++ tools/device-manager/Device.py	1 Jul 2005 16:14:11 -0000
 <at>  <at>  -21,7 +21,7  <at>  <at>  class Device:
     def populate_gtk_tree(self, tree_model, dont_show_virtual, representation):
         # see if we should show virtual devices
         if dont_show_virtual:
-            if self.properties.has_key("info.virtual"):
+            try:
                 if self.properties["info.virtual"]:
                     # do show all block devices, ide channels
                     if not self.properties["info.bus"] in ["block", "ide_host"]:
 <at>  <at>  -32,7 +32,8  <at>  <at>  class Device:
                                                 dont_show_virtual,
                                                 representation)
                         return

Re: h-d-m handle empty properties

On Fri, 2005-07-01 at 18:20 +0200, Kay Sievers wrote:
> Want to get rid of the patches I found in the SUSE package... :)
> This makes hal-device-manager handle empty properties where a value is
> expected.

Sounds good.

> 
> Ok to commit?

Yeah, go ahead.

Thanks,
David

make doc-dir configurable

This makes the place where the documentation gets installed
configurable.

Ok to commit?

Thanks,
Kay

Index: configure.in
===================================================================
RCS file: /cvs/hal/hal/configure.in,v
retrieving revision 1.88
diff -u -r1.88 configure.in
--- configure.in	1 Jul 2005 15:50:31 -0000	1.88
+++ configure.in	1 Jul 2005 17:42:42 -0000
 <at>  <at>  -441,6 +441,17  <at>  <at> 
 fi

 
+# documentation target
+AC_ARG_WITH(doc-dir,   [  --with-doc-dir=[dirname] directory to install documentation])
+if ! test -z "$with_doc_dir"; then
+   DOCDIR=$with_doc_dir
+else
+   DOCDIR="$DATADIR/doc/hal-$VERSION"
+fi
+
+AC_SUBST(DOCDIR)
+

Re: make doc-dir configurable

On Fri, 2005-07-01 at 19:45 +0200, Kay Sievers wrote:
> This makes the place where the documentation gets installed
> configurable.
> 
> Ok to commit?

Go ahead.

Thanks,
David

storage icon name

Hi

I read elsewhere (gnome-vfs) that the libhal-storage icon name policy
stuff in Hal was to be deprecated.

Does this mean I will lose the ability to specify exactly the icon I
want for each of my devices - or will the mechanism to do this simply
change.

Right now I have a couple of cameras. I have configured hal via my own
fdi files to specify my own icons (each one being an icon representing
the particular camera), so that gnome/nautilus displays my specific
camera-icon for each camera (to help me know which one is which).

I'm worried that if this goes away, I will be left with two of the same
icon on my desktop - not so easy to distinguish between the two.

I tried googling on the hal mailing list, but couldn't find anything
describing the pending change in hal - just the patch submitted for
gnome-vfs.

Karl

Re: [PATCH] include mount.h instead of include linux/fs.h

This commit broke the build for my tinderbox; I get the following error now:

checking for BLKGETSIZE64... no
configure: error: BLKGETSIZE64 is not defined

I'm guessing this is just something dumb I don't have installed on the
tinderbox, but I have no idea what- kernel-headers are installed
(though a bizarre hoary version- 2.5.999?)

Thanks-
Luis

On 7/1/05, David Zeuthen <david <at> fubar.dk> wrote:
> On Fri, 2005-07-01 at 12:54 +0200, Kay Sievers wrote:
> > This should fix pgoff_t no defined with some linux headers.
> >
> > Ok to commit?
> 
> Go for it.
> 
> Thanks,
> David
> 
> 
> _______________________________________________
> hal mailing list
> hal <at> lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/hal
>

Re: [PATCH] include mount.h instead of include linux/fs.h

On Fri, Jul 01, 2005 at 07:53:44PM -0400, Luis Villa wrote:
> This commit broke the build for my tinderbox; I get the following error now:
> 
> checking for BLKGETSIZE64... no
> configure: error: BLKGETSIZE64 is not defined
> 
> I'm guessing this is just something dumb I don't have installed on the
> tinderbox, but I have no idea what- kernel-headers are installed
> (though a bizarre hoary version- 2.5.999?)

That's a mess with the kernel headers, every time you fix one
architecture, some other breaks. But I expect your headers are
simply too old.

What does this print:
  find /usr/include | xargs grep BLKGETSIZE64
  /usr/include/linux/fs.h:#define BLKGETSIZE64 _IOR(0x12,114,size_t) /* return device size in bytes
(u64 *arg) */
  /usr/include/sys/mount.h:#define BLKGETSIZE64 _IOR(0x12,114,size_t) /* return device size.  */

If this fails too, your system should definitely catch up:
  find /usr/include | xargs grep NETLINK_KOBJECT_UEVENT
  /usr/include/linux/netlink.h:#define NETLINK_KOBJECT_UEVENT     15 /* Kernel messages to userspace */

Thanks,
Kay