[Qemu-devel] [PATCH] configure: require glib 2.22

[Qemu-devel] [PATCH] configure: require glib 2.22

[Paolo Bonzini]

This provides g_ptr_array_new_with_free_func, as well as a few
other functions that we've been hacking around in glib-compat.h.
Cleaning up the compatibility headers will come later.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 configure |    8 +-------
 1 files changed, 1 insertions(+), 7 deletions(-)

diff --git a/configure b/configure
index 4d9a2a7..833bd69 100755
--- a/configure
+++ b/configure
@@ -2745,13 +2745,7 @@ fi
 ##########################################
 # glib support probe
 
-if test "$mingw32" = yes; then
-    # g_poll is required in order to integrate with the glib main loop.
-    glib_req_ver=2.20
-else
-    glib_req_ver=2.12
-fi
-
+glib_req_ver=2.22
 glib_pkg_config()
 {
   if $pkg_config --atleast-version=$glib_req_ver $1; then
-- 
1.7.1

Re: [Qemu-devel] [PATCH] configure: require glib 2.22

[Igor Mammedov]

On Tue,  3 Mar 2015 16:24:15 +0100
Paolo Bonzini <pbonzini@redhat.com> wrote:

> This provides g_ptr_array_new_with_free_func, as well as a few
> other functions that we've been hacking around in glib-compat.h.
> Cleaning up the compatibility headers will come later.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
You were faster :)

Reviewed-by: Igor Mammedov <imammedo@redhat.com>

I've already updated changelog
---
/me didn't new that Peter is RedHat person now :)
CCing to a correct address

> ---
>  configure |    8 +-------
>  1 files changed, 1 insertions(+), 7 deletions(-)
> 
> diff --git a/configure b/configure
> index 4d9a2a7..833bd69 100755
> --- a/configure
> +++ b/configure
> @@ -2745,13 +2745,7 @@ fi
>  ##########################################
>  # glib support probe
>  
> -if test "$mingw32" = yes; then
> -    # g_poll is required in order to integrate with the glib main loop.
> -    glib_req_ver=2.20
> -else
> -    glib_req_ver=2.12
> -fi
> -
> +glib_req_ver=2.22
>  glib_pkg_config()
>  {
>    if $pkg_config --atleast-version=$glib_req_ver $1; then

Re: [Qemu-devel] [PATCH] configure: require glib 2.22

[Michael S. Tsirkin]

On Tue, Mar 03, 2015 at 04:24:15PM +0100, Paolo Bonzini wrote:
> This provides g_ptr_array_new_with_free_func, as well as a few
> other functions that we've been hacking around in glib-compat.h.
> Cleaning up the compatibility headers will come later.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

At least the g_ptr_array_new_with_free_func is easy to work around, I
fixed that one up.  If we are decided, maybe defer this until after 2.3,
we have enough noise in 2.3 already ...

> ---
>  configure |    8 +-------
>  1 files changed, 1 insertions(+), 7 deletions(-)
> 
> diff --git a/configure b/configure
> index 4d9a2a7..833bd69 100755
> --- a/configure
> +++ b/configure
> @@ -2745,13 +2745,7 @@ fi
>  ##########################################
>  # glib support probe
>  
> -if test "$mingw32" = yes; then
> -    # g_poll is required in order to integrate with the glib main loop.
> -    glib_req_ver=2.20
> -else
> -    glib_req_ver=2.12
> -fi
> -
> +glib_req_ver=2.22
>  glib_pkg_config()
>  {
>    if $pkg_config --atleast-version=$glib_req_ver $1; then
> -- 
> 1.7.1

Re: [Qemu-devel] [PATCH] configure: require glib 2.22

[Markus Armbruster]

"Michael S. Tsirkin" <mst@redhat.com> writes:

> On Tue, Mar 03, 2015 at 04:24:15PM +0100, Paolo Bonzini wrote:
>> This provides g_ptr_array_new_with_free_func, as well as a few
>> other functions that we've been hacking around in glib-compat.h.
>> Cleaning up the compatibility headers will come later.
>> 
>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>
> At least the g_ptr_array_new_with_free_func is easy to work around, I
> fixed that one up.  If we are decided, maybe defer this until after 2.3,
> we have enough noise in 2.3 already ...

Nah, let's go for it.  The additional noise is a drop in the bucket.
Moreover, I trust our ability to test a transition to more recent GLib a
whole lot more than GLib compatibility code for unfortunate users stuck
with ancient OS versions that developers generally don't even compile,
let alone use.

Re: [Qemu-devel] [PATCH] configure: require glib 2.22

[Peter Maydell]

On 4 March 2015 at 16:42, Markus Armbruster <armbru@redhat.com> wrote:
> "Michael S. Tsirkin" <mst@redhat.com> writes:
>> At least the g_ptr_array_new_with_free_func is easy to work around, I
>> fixed that one up.  If we are decided, maybe defer this until after 2.3,
>> we have enough noise in 2.3 already ...
>
> Nah, let's go for it.  The additional noise is a drop in the bucket.
> Moreover, I trust our ability to test a transition to more recent GLib a
> whole lot more than GLib compatibility code for unfortunate users stuck
> with ancient OS versions that developers generally don't even compile,
> let alone use.

Not sure what you have in mind here -- Michael's change to his
patch doesn't add separate code for old glib versions, it just
switches to a different API so the same code works on all versions.

I would have been willing to bump the dependency at this point
if it was hard to work around this one, but since we already
have a patch I'd rather use that and defer the dependency
bump til after 2.3.

Incidentally we do at least compile test this, or we wouldn't be
having this conversation. 2.12 can't compile the 'make check'
tests, though, so it is only a compile test... Still in general
I don't expect much of QEMU to be all that dependent on the
underlying OS, if you ignore the KVM parts. About the only
thing with a bad case of API churn is glib, everything else
(libc, X11, SDL, etc) is pretty stable.

-- PMM

Re: [Qemu-devel] [PATCH] configure: require glib 2.22

[Daniel P. Berrange]

On Tue, Mar 03, 2015 at 04:24:15PM +0100, Paolo Bonzini wrote:
> This provides g_ptr_array_new_with_free_func, as well as a few
> other functions that we've been hacking around in glib-compat.h.
> Cleaning up the compatibility headers will come later.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  configure |    8 +-------
>  1 files changed, 1 insertions(+), 7 deletions(-)

Fine with me - doesn't really help or hurt the I/O channels
work I'm doing to bring encryption to migration/chardev/etc.
It is certainly nice to not have different min versions for
native vs mingw though.

> 
> diff --git a/configure b/configure
> index 4d9a2a7..833bd69 100755
> --- a/configure
> +++ b/configure
> @@ -2745,13 +2745,7 @@ fi
>  ##########################################
>  # glib support probe
>  
> -if test "$mingw32" = yes; then
> -    # g_poll is required in order to integrate with the glib main loop.
> -    glib_req_ver=2.20
> -else
> -    glib_req_ver=2.12
> -fi
> -
> +glib_req_ver=2.22
>  glib_pkg_config()
>  {
>    if $pkg_config --atleast-version=$glib_req_ver $1; then

Regards,
Daniel
-- 
|: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org              -o-             http://virt-manager.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|