Backwards compat for < 2.6.16?

Backwards compat for < 2.6.16?

[Rusty Russell]

Hi all,

	The lapic code uses ktime_t and hrtimers, which means we no longer
build for kernels 2.6.15 and below.  Is it worth trying to fake this
infrastructure up, building without lapic, or abandoning < 2.6.16?

Thanks,
Rusty.


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

[PATCH] Script to help check compatibility against a range of kernels.

[Rusty Russell]

This makes it easier to check kvm-userspace compatibility layer.

---
 check-compat.sh |   27 +++++++++++++++++++++++++++
 1 files changed, 27 insertions(+), 0 deletions(-)
 create mode 100755 check-compat.sh

diff --git a/check-compat.sh b/check-compat.sh
new file mode 100755
index 0000000..c07be4d
--- /dev/null
+++ b/check-compat.sh
@@ -0,0 +1,27 @@
+#! /bin/sh
+
+die()
+{
+    echo FAILED:
+    cat $LOGFILE
+    exit 1
+}
+
+if [ $# = 0 ]; then
+    echo Usage: ./check-compat.sh kerneldir... >&2
+    echo eg. './check-compat.sh ../linux-2.6.*' >&2
+    exit 1
+fi
+
+LOGFILE=`mktemp -t kvm-check-compat.XXXXXXXXXX`
+trap "rm -f $LOGFILE" 0
+
+for k in "$@"; do
+    [ -d $k ] || continue
+    make clean > /dev/null 2>&1
+    echo -n Checking $k...
+    ./configure --kerneldir=`pwd`/$k > $LOGFILE 2>&1 || die
+    make >> $LOGFILE 2>&1 || die
+    echo ok.
+done
+
-- 
1.5.1.3




-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

Re: Backwards compat for < 2.6.16?

[Avi Kivity]

Rusty Russell wrote:
> Hi all,
>
> 	The lapic code uses ktime_t and hrtimers, which means we no longer
> build for kernels 2.6.15 and below.  Is it worth trying to fake this
> infrastructure up, building without lapic, or abandoning < 2.6.16?
>
>   

I think we can stub the code away, and hack the feature reporting to 
hide the crippled functionality.


-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

Re: [PATCH] Script to help check compatibility against a range of kernels.

[Avi Kivity]

Rusty Russell wrote:
> This makes it easier to check kvm-userspace compatibility layer.
>
>   

I actually started to write something like this, except it pulled older 
kernels from git.

Thanks.

> ---
>  check-compat.sh |   27 +++++++++++++++++++++++++++
>  1 files changed, 27 insertions(+), 0 deletions(-)
>  create mode 100755 check-compat.sh
>
> diff --git a/check-compat.sh b/check-compat.sh
> new file mode 100755
> index 0000000..c07be4d
> --- /dev/null
> +++ b/check-compat.sh
> @@ -0,0 +1,27 @@
> +#! /bin/sh
> +
> +die()
> +{
> +    echo FAILED:
> +    cat $LOGFILE
> +    exit 1
> +}
> +
> +if [ $# = 0 ]; then
> +    echo Usage: ./check-compat.sh kerneldir... >&2
> +    echo eg. './check-compat.sh ../linux-2.6.*' >&2
> +    exit 1
> +fi
> +
> +LOGFILE=`mktemp -t kvm-check-compat.XXXXXXXXXX`
> +trap "rm -f $LOGFILE" 0
> +
> +for k in "$@"; do
> +    [ -d $k ] || continue
> +    make clean > /dev/null 2>&1
> +    echo -n Checking $k...
> +    ./configure --kerneldir=`pwd`/$k > $LOGFILE 2>&1 || die
> +    make >> $LOGFILE 2>&1 || die
> +    echo ok.
> +done
> +
>   


-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/