Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

[Ian Campbell]

On Mon, 2012-05-14 at 17:33 +0100, Xen patchbot-unstable wrote:
> # HG changeset patch
> # User Roger Pau Monne <roger.pau@citrix.com>
> # Date 1337008959 -3600
> # Node ID dfe39bd65137a97d18f0ee7d155d3755ae5530b4
> # Parent  49ce39c88aeeb0ba58e4f0e2bf865f6981f6e99d
> autoconf: check for dev86 and iasl on x86* only
> 
> Check for this tools on x86 systems only.
> 
> Signed-off-by: Roger Pau Monne <roger.pau@citrix.com>
> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
> Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>
> ---
> 
> 
> diff -r 49ce39c88aee -r dfe39bd65137 tools/configure
> --- a/tools/configure	Mon May 14 16:20:33 2012 +0100
> +++ b/tools/configure	Mon May 14 16:22:39 2012 +0100
> @@ -2622,9 +2622,14 @@ LDFLAGS="$PREPEND_LDFLAGS $LDFLAGS $APPE
>  
> 
> 
> -
> -
> -
> +case "$host_cpu" in
> +i[3456]86|x86_64)
> +
> +
> +
> +
> +    ;;
> +esac

This seems like a strange result from the change below, or is XC_ARG_VAR
weird in some way? On ARM I still see the configure option:
        ianc@army:xen-unstable$ ./configure --help | grep iasl
          IASL        Path to iasl tool
        ianc@army:xen-unstable$ uname -m
        armv7l

The thing I actually noticed was that we still have 
        AX_PATH_PROG_OR_FAIL([AS86], [as86])
        AX_PATH_PROG_OR_FAIL([LD86], [ld86])
        AX_PATH_PROG_OR_FAIL([BCC], [bcc])
        AX_PATH_PROG_OR_FAIL([IASL], [iasl])

Is that expected? I don't think it is... Should these not be the ones
which are conditional?

>  
>  # Checks for programs.
>  ac_ext=c
> diff -r 49ce39c88aee -r dfe39bd65137 tools/configure.ac
> --- a/tools/configure.ac	Mon May 14 16:20:33 2012 +0100
> +++ b/tools/configure.ac	Mon May 14 16:22:39 2012 +0100
> @@ -67,10 +67,16 @@ AC_ARG_VAR([CURL], [Path to curl-config 
>  AC_ARG_VAR([XML], [Path to xml2-config tool])
>  AC_ARG_VAR([BASH], [Path to bash shell])
>  AC_ARG_VAR([XGETTEXT], [Path to xgetttext tool])
> -AC_ARG_VAR([AS86], [Path to as86 tool])
> -AC_ARG_VAR([LD86], [Path to ld86 tool])
> -AC_ARG_VAR([BCC], [Path to bcc tool])
> -AC_ARG_VAR([IASL], [Path to iasl tool])
> +
> +dnl as86, ld86, bcc and iasl are only present in x86* systems
> +case "$host_cpu" in
> +i[[3456]]86|x86_64)
> +    AC_ARG_VAR([AS86], [Path to as86 tool])
> +    AC_ARG_VAR([LD86], [Path to ld86 tool])
> +    AC_ARG_VAR([BCC], [Path to bcc tool])
> +    AC_ARG_VAR([IASL], [Path to iasl tool])
> +    ;;
> +esac
>  
>  # Checks for programs.
>  AC_PROG_CC
> 
> _______________________________________________
> Xen-staging mailing list
> Xen-staging@lists.xen.org
> http://lists.xensource.com/xen-staging

Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

[Roger Pau Monne]

Ian Campbell wrote:
> On Mon, 2012-05-14 at 17:33 +0100, Xen patchbot-unstable wrote:
>> # HG changeset patch
>> # User Roger Pau Monne<roger.pau@citrix.com>
>> # Date 1337008959 -3600
>> # Node ID dfe39bd65137a97d18f0ee7d155d3755ae5530b4
>> # Parent  49ce39c88aeeb0ba58e4f0e2bf865f6981f6e99d
>> autoconf: check for dev86 and iasl on x86* only
>>
>> Check for this tools on x86 systems only.
>>
>> Signed-off-by: Roger Pau Monne<roger.pau@citrix.com>
>> Acked-by: Ian Jackson<ian.jackson@eu.citrix.com>
>> Committed-by: Ian Jackson<ian.jackson@eu.citrix.com>
>> ---
>>
>>
>> diff -r 49ce39c88aee -r dfe39bd65137 tools/configure
>> --- a/tools/configure	Mon May 14 16:20:33 2012 +0100
>> +++ b/tools/configure	Mon May 14 16:22:39 2012 +0100
>> @@ -2622,9 +2622,14 @@ LDFLAGS="$PREPEND_LDFLAGS $LDFLAGS $APPE
>>
>>
>>
>> -
>> -
>> -
>> +case "$host_cpu" in
>> +i[3456]86|x86_64)
>> +
>> +
>> +
>> +
>> +    ;;
>> +esac
>
> This seems like a strange result from the change below, or is XC_ARG_VAR
> weird in some way? On ARM I still see the configure option:
>          ianc@army:xen-unstable$ ./configure --help | grep iasl
>            IASL        Path to iasl tool
>          ianc@army:xen-unstable$ uname -m
>          armv7l
>
> The thing I actually noticed was that we still have
>          AX_PATH_PROG_OR_FAIL([AS86], [as86])
>          AX_PATH_PROG_OR_FAIL([LD86], [ld86])
>          AX_PATH_PROG_OR_FAIL([BCC], [bcc])
>          AX_PATH_PROG_OR_FAIL([IASL], [iasl])
>
> Is that expected? I don't think it is... Should these not be the ones
> which are conditional?

It is a conditional, but since the script is the same for all 
architectures and arch is not checked when doing a "configure --help", 
all the possible options are printed, even those that don't apply to a 
system. I will try to check if there's a better way to hide them, but 
I'm not sure.

>>
>>   # Checks for programs.
>>   ac_ext=c
>> diff -r 49ce39c88aee -r dfe39bd65137 tools/configure.ac
>> --- a/tools/configure.ac	Mon May 14 16:20:33 2012 +0100
>> +++ b/tools/configure.ac	Mon May 14 16:22:39 2012 +0100
>> @@ -67,10 +67,16 @@ AC_ARG_VAR([CURL], [Path to curl-config
>>   AC_ARG_VAR([XML], [Path to xml2-config tool])
>>   AC_ARG_VAR([BASH], [Path to bash shell])
>>   AC_ARG_VAR([XGETTEXT], [Path to xgetttext tool])
>> -AC_ARG_VAR([AS86], [Path to as86 tool])
>> -AC_ARG_VAR([LD86], [Path to ld86 tool])
>> -AC_ARG_VAR([BCC], [Path to bcc tool])
>> -AC_ARG_VAR([IASL], [Path to iasl tool])
>> +
>> +dnl as86, ld86, bcc and iasl are only present in x86* systems
>> +case "$host_cpu" in
>> +i[[3456]]86|x86_64)
>> +    AC_ARG_VAR([AS86], [Path to as86 tool])
>> +    AC_ARG_VAR([LD86], [Path to ld86 tool])
>> +    AC_ARG_VAR([BCC], [Path to bcc tool])
>> +    AC_ARG_VAR([IASL], [Path to iasl tool])
>> +    ;;
>> +esac
>>
>>   # Checks for programs.
>>   AC_PROG_CC
>>

I don't know why, but I think my previous patch missed to also make the 
actual check conditional, so the applied patch was useless. This should 
fix it:

8<----------------------------------------------------------

autoconf: disable dev86 and iasl checks on arm

Run autogen after applying this patch.

Signed-off-by: Roger Pau Monne <roger.pau@citrix.com>
---
  tools/configure.ac |   13 +++++++++----
  1 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/tools/configure.ac b/tools/configure.ac
index 706ee13..f7aa9b8 100644
--- a/tools/configure.ac
+++ b/tools/configure.ac
@@ -109,10 +109,15 @@ AS_IF([test "x$pythontools" = "xy"], [
      AX_CHECK_PYTHON_DEVEL()
  ])
  AX_PATH_PROG_OR_FAIL([XGETTEXT], [xgettext])
-AX_PATH_PROG_OR_FAIL([AS86], [as86])
-AX_PATH_PROG_OR_FAIL([LD86], [ld86])
-AX_PATH_PROG_OR_FAIL([BCC], [bcc])
-AX_PATH_PROG_OR_FAIL([IASL], [iasl])
+dnl as86, ld86, bcc and iasl are only present in x86* systems
+case "$host_cpu" in
+i[[3456]]86|x86_64)
+    AX_PATH_PROG_OR_FAIL([AS86], [as86])
+    AX_PATH_PROG_OR_FAIL([LD86], [ld86])
+    AX_PATH_PROG_OR_FAIL([BCC], [bcc])
+    AX_PATH_PROG_OR_FAIL([IASL], [iasl])
+    ;;
+esac
  AX_CHECK_UUID
  AX_CHECK_CURSES
  PKG_CHECK_MODULES(glib, glib-2.0)
-- 

Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

[Ian Campbell]

On Fri, 2012-05-25 at 14:38 +0100, Roger Pau Monne wrote:
> > Is that expected? I don't think it is... Should these not be the ones
> > which are conditional?
> 
> It is a conditional, but since the script is the same for all 
> architectures and arch is not checked when doing a "configure --help", 
> all the possible options are printed, even those that don't apply to a 
> system. I will try to check if there's a better way to hide them, but 
> I'm not sure.

If not then I guess the original patch should be reverted?

> >>
> >>   # Checks for programs.
> >>   ac_ext=c
> >> diff -r 49ce39c88aee -r dfe39bd65137 tools/configure.ac
> >> --- a/tools/configure.ac	Mon May 14 16:20:33 2012 +0100
> >> +++ b/tools/configure.ac	Mon May 14 16:22:39 2012 +0100
> >> @@ -67,10 +67,16 @@ AC_ARG_VAR([CURL], [Path to curl-config
> >>   AC_ARG_VAR([XML], [Path to xml2-config tool])
> >>   AC_ARG_VAR([BASH], [Path to bash shell])
> >>   AC_ARG_VAR([XGETTEXT], [Path to xgetttext tool])
> >> -AC_ARG_VAR([AS86], [Path to as86 tool])
> >> -AC_ARG_VAR([LD86], [Path to ld86 tool])
> >> -AC_ARG_VAR([BCC], [Path to bcc tool])
> >> -AC_ARG_VAR([IASL], [Path to iasl tool])
> >> +
> >> +dnl as86, ld86, bcc and iasl are only present in x86* systems
> >> +case "$host_cpu" in
> >> +i[[3456]]86|x86_64)
> >> +    AC_ARG_VAR([AS86], [Path to as86 tool])
> >> +    AC_ARG_VAR([LD86], [Path to ld86 tool])
> >> +    AC_ARG_VAR([BCC], [Path to bcc tool])
> >> +    AC_ARG_VAR([IASL], [Path to iasl tool])
> >> +    ;;
> >> +esac
> >>
> >>   # Checks for programs.
> >>   AC_PROG_CC
> >>
> 
> I don't know why, but I think my previous patch missed to also make the 
> actual check conditional, so the applied patch was useless. This should 
> fix it:
> 
> 8<----------------------------------------------------------
> 
> autoconf: disable dev86 and iasl checks on arm
> 
> Run autogen after applying this patch.
> 
> Signed-off-by: Roger Pau Monne <roger.pau@citrix.com>

Looks good to me
Acked-by: Ian Campbell <ian.campbell@citrix.com>

> ---
>   tools/configure.ac |   13 +++++++++----
>   1 files changed, 9 insertions(+), 4 deletions(-)
> 
> diff --git a/tools/configure.ac b/tools/configure.ac
> index 706ee13..f7aa9b8 100644
> --- a/tools/configure.ac
> +++ b/tools/configure.ac
> @@ -109,10 +109,15 @@ AS_IF([test "x$pythontools" = "xy"], [
>       AX_CHECK_PYTHON_DEVEL()
>   ])
>   AX_PATH_PROG_OR_FAIL([XGETTEXT], [xgettext])
> -AX_PATH_PROG_OR_FAIL([AS86], [as86])
> -AX_PATH_PROG_OR_FAIL([LD86], [ld86])
> -AX_PATH_PROG_OR_FAIL([BCC], [bcc])
> -AX_PATH_PROG_OR_FAIL([IASL], [iasl])
> +dnl as86, ld86, bcc and iasl are only present in x86* systems
> +case "$host_cpu" in
> +i[[3456]]86|x86_64)
> +    AX_PATH_PROG_OR_FAIL([AS86], [as86])
> +    AX_PATH_PROG_OR_FAIL([LD86], [ld86])
> +    AX_PATH_PROG_OR_FAIL([BCC], [bcc])
> +    AX_PATH_PROG_OR_FAIL([IASL], [iasl])
> +    ;;
> +esac
>   AX_CHECK_UUID
>   AX_CHECK_CURSES
>   PKG_CHECK_MODULES(glib, glib-2.0)
> -- 

Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

[Roger Pau Monne]

Ian Campbell wrote:
> On Fri, 2012-05-25 at 14:38 +0100, Roger Pau Monne wrote:
>>> Is that expected? I don't think it is... Should these not be the ones
>>> which are conditional?
>> It is a conditional, but since the script is the same for all
>> architectures and arch is not checked when doing a "configure --help",
>> all the possible options are printed, even those that don't apply to a
>> system. I will try to check if there's a better way to hide them, but
>> I'm not sure.
>
> If not then I guess the original patch should be reverted?

The original patch doesn't do anything, so yes it can be reverted 
(although it doesn't affect configure in any way).

>>>>    # Checks for programs.
>>>>    ac_ext=c
>>>> diff -r 49ce39c88aee -r dfe39bd65137 tools/configure.ac
>>>> --- a/tools/configure.ac	Mon May 14 16:20:33 2012 +0100
>>>> +++ b/tools/configure.ac	Mon May 14 16:22:39 2012 +0100
>>>> @@ -67,10 +67,16 @@ AC_ARG_VAR([CURL], [Path to curl-config
>>>>    AC_ARG_VAR([XML], [Path to xml2-config tool])
>>>>    AC_ARG_VAR([BASH], [Path to bash shell])
>>>>    AC_ARG_VAR([XGETTEXT], [Path to xgetttext tool])
>>>> -AC_ARG_VAR([AS86], [Path to as86 tool])
>>>> -AC_ARG_VAR([LD86], [Path to ld86 tool])
>>>> -AC_ARG_VAR([BCC], [Path to bcc tool])
>>>> -AC_ARG_VAR([IASL], [Path to iasl tool])
>>>> +
>>>> +dnl as86, ld86, bcc and iasl are only present in x86* systems
>>>> +case "$host_cpu" in
>>>> +i[[3456]]86|x86_64)
>>>> +    AC_ARG_VAR([AS86], [Path to as86 tool])
>>>> +    AC_ARG_VAR([LD86], [Path to ld86 tool])
>>>> +    AC_ARG_VAR([BCC], [Path to bcc tool])
>>>> +    AC_ARG_VAR([IASL], [Path to iasl tool])
>>>> +    ;;
>>>> +esac
>>>>
>>>>    # Checks for programs.
>>>>    AC_PROG_CC
>>>>
>> I don't know why, but I think my previous patch missed to also make the
>> actual check conditional, so the applied patch was useless. This should
>> fix it:
>>
>> 8<----------------------------------------------------------
>>
>> autoconf: disable dev86 and iasl checks on arm
>>
>> Run autogen after applying this patch.
>>
>> Signed-off-by: Roger Pau Monne<roger.pau@citrix.com>
>
> Looks good to me
> Acked-by: Ian Campbell<ian.campbell@citrix.com>

Any news on this one? It's been quite some time since the Ack, but I 
don't think it has been committed (or at least I cannot find it).

Thanks!

>> ---
>>    tools/configure.ac |   13 +++++++++----
>>    1 files changed, 9 insertions(+), 4 deletions(-)
>>
>> diff --git a/tools/configure.ac b/tools/configure.ac
>> index 706ee13..f7aa9b8 100644
>> --- a/tools/configure.ac
>> +++ b/tools/configure.ac
>> @@ -109,10 +109,15 @@ AS_IF([test "x$pythontools" = "xy"], [
>>        AX_CHECK_PYTHON_DEVEL()
>>    ])
>>    AX_PATH_PROG_OR_FAIL([XGETTEXT], [xgettext])
>> -AX_PATH_PROG_OR_FAIL([AS86], [as86])
>> -AX_PATH_PROG_OR_FAIL([LD86], [ld86])
>> -AX_PATH_PROG_OR_FAIL([BCC], [bcc])
>> -AX_PATH_PROG_OR_FAIL([IASL], [iasl])
>> +dnl as86, ld86, bcc and iasl are only present in x86* systems
>> +case "$host_cpu" in
>> +i[[3456]]86|x86_64)
>> +    AX_PATH_PROG_OR_FAIL([AS86], [as86])
>> +    AX_PATH_PROG_OR_FAIL([LD86], [ld86])
>> +    AX_PATH_PROG_OR_FAIL([BCC], [bcc])
>> +    AX_PATH_PROG_OR_FAIL([IASL], [iasl])
>> +    ;;
>> +esac
>>    AX_CHECK_UUID
>>    AX_CHECK_CURSES
>>    PKG_CHECK_MODULES(glib, glib-2.0)
>> --
>
>

Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

[Ian Campbell]

On Mon, 2012-06-18 at 10:50 +0100, Roger Pau Monne wrote:
> >    AX_PATH_PROG_OR_FAIL([XGETTEXT], [xgettext])
> >> -AX_PATH_PROG_OR_FAIL([AS86], [as86])
> >> -AX_PATH_PROG_OR_FAIL([LD86], [ld86])
> >> -AX_PATH_PROG_OR_FAIL([BCC], [bcc])
> >> -AX_PATH_PROG_OR_FAIL([IASL], [iasl])
> >> +dnl as86, ld86, bcc and iasl are only present in x86* systems
> >> +case "$host_cpu" in

In the meantime AC_CANONICAL_HOST has been removed (25303:078c7d4cde1d)
which breaks this test. I shall re-add it and fold into the commit.

Ian.

Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

[Ian Campbell]

On Tue, 2012-07-24 at 09:28 +0100, Ian Campbell wrote:
> On Mon, 2012-06-18 at 10:50 +0100, Roger Pau Monne wrote:
> > >    AX_PATH_PROG_OR_FAIL([XGETTEXT], [xgettext])
> > >> -AX_PATH_PROG_OR_FAIL([AS86], [as86])
> > >> -AX_PATH_PROG_OR_FAIL([LD86], [ld86])
> > >> -AX_PATH_PROG_OR_FAIL([BCC], [bcc])
> > >> -AX_PATH_PROG_OR_FAIL([IASL], [iasl])
> > >> +dnl as86, ld86, bcc and iasl are only present in x86* systems
> > >> +case "$host_cpu" in
> 
> In the meantime AC_CANONICAL_HOST has been removed (25303:078c7d4cde1d)
> which breaks this test. I shall re-add it and fold into the commit.

I applied this as 25639:ce24b84f6af8 -- please take a look and ensure I
got it right.

Ian.

Re: [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only

[Ian Jackson]

Ian Campbell writes ("Re: [Xen-devel] [Xen-staging] [xen-unstable] autoconf: check for dev86 and iasl on x86* only"):
> On Tue, 2012-07-24 at 09:28 +0100, Ian Campbell wrote:
> > On Mon, 2012-06-18 at 10:50 +0100, Roger Pau Monne wrote:
> > > >    AX_PATH_PROG_OR_FAIL([XGETTEXT], [xgettext])
> > > >> -AX_PATH_PROG_OR_FAIL([AS86], [as86])
> > > >> -AX_PATH_PROG_OR_FAIL([LD86], [ld86])
> > > >> -AX_PATH_PROG_OR_FAIL([BCC], [bcc])
> > > >> -AX_PATH_PROG_OR_FAIL([IASL], [iasl])
> > > >> +dnl as86, ld86, bcc and iasl are only present in x86* systems
> > > >> +case "$host_cpu" in
> > 
> > In the meantime AC_CANONICAL_HOST has been removed (25303:078c7d4cde1d)
> > which breaks this test. I shall re-add it and fold into the commit.
> 
> I applied this as 25639:ce24b84f6af8 -- please take a look and ensure I
> got it right.

Inspecting the diff reveals some probably-unintentional whitespace
changes but aside from that it looks good to me.

Ian.