* [RFC] QEMU KVM target
@ 2007-02-20 1:08 Anthony Liguori
[not found] ` <45DA4A16.4050804-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
0 siblings, 1 reply; 13+ messages in thread
From: Anthony Liguori @ 2007-02-20 1:08 UTC (permalink / raw)
To: kvm-devel
[-- Attachment #1: Type: text/plain, Size: 529 bytes --]
Howdy,
Here's a tiny patch that adds a i386-kvm target. The main difference
between the i386-kvm and i386-softmmu target is that the -kvm target
does not have any of the dyngen infrastructure. This means that it will
build with gcc-4. I know you can do --cc=gcc to use gcc-4 but quite a
few versions of gcc-4 have trouble with compiling dyngen.
I also suspect this may prove useful down the road. What do other
people think? I'm not terribly tied to the i386-kvm name for what it's
worth.
Regards,
Anthony Liguori
[-- Attachment #2: kvm-no-gcc4.diff --]
[-- Type: text/x-patch, Size: 5894 bytes --]
diff -r 64fb395dc8f1 configure
--- a/configure Thu Feb 15 14:36:40 2007 +0000
+++ b/configure Mon Feb 19 18:22:08 2007 -0600
@@ -3,8 +3,6 @@ prefix=/usr/kvm
prefix=/usr/kvm
kerneldir=/lib/modules/$(uname -r)/build
want_module=1
-qemu_cc=$(ls /usr/bin/gcc3* /usr/bin/gcc-3* 2>/dev/null | tail -n1)
-disable_gcc_check=
usage() {
cat <<-EOF
@@ -15,20 +13,11 @@ usage() {
--prefix=PREFIX where to install things ($prefix)
--with-patched-kernel don't use external module
--kerneldir=DIR kernel build directory ($kerneldir)
- --qemu-cc="$qemu_cc" compiler for qemu (needs gcc3.x) ($qemu_cc)
- --disable-gcc-check don't insist on gcc-3.x
- this will break running without kvm
EOF
exit 1
}
-
-# prefer gcc if its version is 3.* ( over a compat-gcc )
-# do it before parsing command line arguments to enable the user
-# to specify a specific gcc he/she likes.
-if gcc -v 2>&1 | grep -q 'gcc *version *3\.[2-4]\.[0-9]'; then
- qemu_cc=gcc
-fi
while [[ "$1" = -* ]]; do
opt="$1"; shift
@@ -47,12 +36,6 @@ while [[ "$1" = -* ]]; do
--with-patched-kernel)
want_module=
;;
- --qemu-cc)
- qemu_cc="$arg"
- ;;
- --disable-gcc-check)
- disable_gcc_check=1
- ;;
--help)
usage
;;
@@ -61,11 +44,6 @@ while [[ "$1" = -* ]]; do
;;
esac
done
-
-if [[ -z "$qemu_cc" ]]; then
- echo "$0: cannot locate gcc 3.x. please install it or specify with --qemu-cc"
- exit 1
-fi
libkvm_kerneldir="$kerneldir"
if (( want_module )); then
@@ -81,12 +59,12 @@ target_cpu() {
}
(cd user; ./configure --prefix="$prefix" --kerneldir="$libkvm_kerneldir")
-(cd qemu; ./configure --target-list=$(target_cpu)-softmmu --cc="$qemu_cc" \
+(cd qemu; ./configure --target-list=i386-kvm \
--disable-kqemu --extra-cflags="-I $PWD/../user" \
--extra-ldflags="-L $PWD/../user" \
--enable-kvm --kernel-path="$libkvm_kerneldir" \
--enable-alsa \
- ${disable_gcc_check:+"--disable-gcc-check"} \
+ --disable-gcc-check \
--prefix="$prefix"
)
diff -r 64fb395dc8f1 qemu/Makefile.target
--- a/qemu/Makefile.target Thu Feb 15 14:36:40 2007 +0000
+++ b/qemu/Makefile.target Mon Feb 19 18:15:02 2007 -0600
@@ -52,7 +52,11 @@ QEMU_USER=qemu-$(TARGET_ARCH2)
# system emulator name
ifdef CONFIG_SOFTMMU
ifeq ($(TARGET_ARCH), i386)
+ifdef CONFIG_KVM
+QEMU_SYSTEM=qemu-kvm$(EXESUF)
+else
QEMU_SYSTEM=qemu$(EXESUF)
+endif
else
QEMU_SYSTEM=qemu-system-$(TARGET_ARCH2)$(EXESUF)
endif
@@ -231,8 +235,10 @@ OBJS+= libqemu.a
OBJS+= libqemu.a
# cpu emulator library
-LIBOBJS=exec.o kqemu.o qemu-kvm.o translate-op.o translate-all.o cpu-exec.o\
- translate.o op.o
+LIBOBJS=exec.o kqemu.o qemu-kvm.o cpu-exec.o
+ifndef CONFIG_KVM
+LIBOBJS+=translate.o op.o translate-op.o translate-all.o
+endif
ifdef CONFIG_SOFTFLOAT
LIBOBJS+=fpu/softfloat.o
else
diff -r 64fb395dc8f1 qemu/configure
--- a/qemu/configure Thu Feb 15 14:36:40 2007 +0000
+++ b/qemu/configure Mon Feb 19 18:14:17 2007 -0600
@@ -819,6 +819,11 @@ target_user_only="no"
target_user_only="no"
if expr $target : '.*-user' > /dev/null ; then
target_user_only="yes"
+fi
+target_kvm="no"
+if expr $target : '.*-kvm' > /dev/null ; then
+ target_kvm="yes"
+ target_softmmu="yes"
fi
target_linux_user="no"
@@ -943,6 +948,10 @@ if test "$target_softmmu" = "yes" ; then
echo "CONFIG_SOFTMMU=yes" >> $config_mak
echo "#define CONFIG_SOFTMMU 1" >> $config_h
fi
+if test "$target_kvm" = "yes" ; then
+ echo "CONFIG_KVM=yes" >> $config_mak
+ echo "#define CONFIG_KVM 1" >> $config_h
+fi
if test "$target_user_only" = "yes" ; then
echo "CONFIG_USER_ONLY=yes" >> $config_mak
echo "#define CONFIG_USER_ONLY 1" >> $config_h
diff -r 64fb395dc8f1 qemu/qemu-kvm.c
--- a/qemu/qemu-kvm.c Thu Feb 15 14:36:40 2007 +0000
+++ b/qemu/qemu-kvm.c Mon Feb 19 18:21:03 2007 -0600
@@ -705,5 +705,40 @@ int kvm_update_debugger(CPUState *env)
return kvm_guest_debug(kvm_context, 0, &dbg);
}
-
-#endif
+#ifdef CONFIG_KVM
+
+CCTable cc_table[CC_OP_NB];
+
+int code_copy_enabled;
+
+static int nop_flags(void)
+{
+ return 0;
+}
+
+int cpu_gen_code(CPUState *env, TranslationBlock *tb,
+ int max_code_size, int *gen_code_size_ptr)
+{
+ return -1;
+}
+
+int cpu_restore_state(TranslationBlock *tb,
+ CPUState *env, unsigned long searched_pc,
+ void *puc)
+{
+ return -1;
+}
+
+void optimize_flags_init(void)
+{
+ int i;
+
+ for (i = 0; i < CC_OP_NB; i++) {
+ cc_table[i].compute_c = nop_flags;
+ cc_table[i].compute_all = nop_flags;
+ }
+}
+
+#endif
+
+#endif
diff -r 64fb395dc8f1 qemu/target-i386/helper.c
--- a/qemu/target-i386/helper.c Thu Feb 15 14:36:40 2007 +0000
+++ b/qemu/target-i386/helper.c Mon Feb 19 18:27:35 2007 -0600
@@ -1594,6 +1594,9 @@ void helper_idivl_EAX_T0(void)
EDX = (uint32_t)r;
}
+/* GCC 4 has a hard time with this helper */
+
+#ifndef CONFIG_KVM
void helper_cmpxchg8b(void)
{
uint64_t d;
@@ -1611,6 +1614,7 @@ void helper_cmpxchg8b(void)
}
CC_SRC = eflags;
}
+#endif
void helper_cpuid(void)
{
@@ -2968,6 +2972,9 @@ void helper_verw(void)
CC_SRC = eflags | CC_Z;
}
+/* GCC 4 has a hard time with the FPU helpers */
+#ifndef CONFIG_KVM
+
/* FPU helpers */
void helper_fldt_ST0_A0(void)
@@ -3485,6 +3492,7 @@ void helper_fxrstor(target_ulong ptr, in
}
}
}
+#endif
#ifndef USE_X86LDOUBLE
diff -r 64fb395dc8f1 qemu/vl.c
--- a/qemu/vl.c Thu Feb 15 14:36:40 2007 +0000
+++ b/qemu/vl.c Mon Feb 19 19:03:34 2007 -0600
@@ -7341,8 +7341,13 @@ int main(int argc, char **argv)
#if USE_KVM
if (kvm_allowed) {
if (kvm_qemu_init() < 0) {
+#ifdef CONFIG_KVM
+ fprintf(stderr, "Could not initialize KVM\n");
+ exit(1);
+#else
fprintf(stderr, "Could not initialize KVM, will disable KVM support\n");
kvm_allowed = 0;
+#endif
}
}
#endif
[-- Attachment #3: Type: text/plain, Size: 345 bytes --]
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
[-- Attachment #4: Type: text/plain, Size: 186 bytes --]
_______________________________________________
kvm-devel mailing list
kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/kvm-devel
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DA4A16.4050804-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
@ 2007-02-20 2:55 ` Jun Koi
[not found] ` <fdaac4d50702191855gc2fc393ga83787bbf2d967da-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2007-02-20 7:19 ` Avi Kivity
1 sibling, 1 reply; 13+ messages in thread
From: Jun Koi @ 2007-02-20 2:55 UTC (permalink / raw)
To: Anthony Liguori; +Cc: kvm-devel
Howdy Anthony,
> Here's a tiny patch that adds a i386-kvm target. The main difference
> between the i386-kvm and i386-softmmu target is that the -kvm target
> does not have any of the dyngen infrastructure. This means that it will
> build with gcc-4. I know you can do --cc=gcc to use gcc-4 but quite a
> few versions of gcc-4 have trouble with compiling dyngen.
I understand that your approach is to remove the "binary translation"
part of the vanilla qemu, because we dont need it. We only need the
device emulation code of qemu for kvm. Do I understand correctly?
If so, we do not need to depend on gcc3 tricks anymore. Removing the
dependent on gcc3 is really an important achievement.
Thanks,
Jun
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <fdaac4d50702191855gc2fc393ga83787bbf2d967da-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
@ 2007-02-20 3:51 ` Anthony Liguori
2007-02-20 23:34 ` Rusty Russell
1 sibling, 0 replies; 13+ messages in thread
From: Anthony Liguori @ 2007-02-20 3:51 UTC (permalink / raw)
To: Jun Koi; +Cc: kvm-devel
Jun Koi wrote:
> Howdy Anthony,
>
>> Here's a tiny patch that adds a i386-kvm target. The main difference
>> between the i386-kvm and i386-softmmu target is that the -kvm target
>> does not have any of the dyngen infrastructure. This means that it will
>> build with gcc-4. I know you can do --cc=gcc to use gcc-4 but quite a
>> few versions of gcc-4 have trouble with compiling dyngen.
>
> I understand that your approach is to remove the "binary translation"
> part of the vanilla qemu, because we dont need it. We only need the
> device emulation code of qemu for kvm. Do I understand correctly?
Yup.
Regards,
Anthony Liguori
> If so, we do not need to depend on gcc3 tricks anymore. Removing the
> dependent on gcc3 is really an important achievement.
>
> Thanks,
> Jun
>
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DA4A16.4050804-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
2007-02-20 2:55 ` Jun Koi
@ 2007-02-20 7:19 ` Avi Kivity
[not found] ` <45DAA108.8000501-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
1 sibling, 1 reply; 13+ messages in thread
From: Avi Kivity @ 2007-02-20 7:19 UTC (permalink / raw)
To: Anthony Liguori; +Cc: kvm-devel
Anthony Liguori wrote:
> Howdy,
>
> Here's a tiny patch that adds a i386-kvm target. The main difference
> between the i386-kvm and i386-softmmu target is that the -kvm target
> does not have any of the dyngen infrastructure. This means that it
> will build with gcc-4. I know you can do --cc=gcc to use gcc-4 but
> quite a few versions of gcc-4 have trouble with compiling dyngen.
>
> I also suspect this may prove useful down the road. What do other
> people think? I'm not terribly tied to the i386-kvm name for what
> it's worth.
>
That kills the -no-kvm switch, which allows a single binary to be used
both with and without kvm. Or do you think both target-i386+kvm and
target-kvm ought to be kept?
My thinking about qemu integration is that kqemu/kvm code needs to be
abstracted into an API to reduce the #ifdefing in qemu, and that API
could call kqemu or kvm as appropriate.
--
Do not meddle in the internals of kernels, for they are subtle and quick to panic.
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DAA108.8000501-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
@ 2007-02-20 9:48 ` Avi Kivity
[not found] ` <45DAC3FF.1090208-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-02-20 15:13 ` Anthony Liguori
1 sibling, 1 reply; 13+ messages in thread
From: Avi Kivity @ 2007-02-20 9:48 UTC (permalink / raw)
To: Anthony Liguori; +Cc: kvm-devel
Avi Kivity wrote:
> Anthony Liguori wrote:
>
>> Howdy,
>>
>> Here's a tiny patch that adds a i386-kvm target. The main difference
>> between the i386-kvm and i386-softmmu target is that the -kvm target
>> does not have any of the dyngen infrastructure. This means that it
>> will build with gcc-4. I know you can do --cc=gcc to use gcc-4 but
>> quite a few versions of gcc-4 have trouble with compiling dyngen.
>>
>> I also suspect this may prove useful down the road. What do other
>> people think? I'm not terribly tied to the i386-kvm name for what
>> it's worth.
>>
>>
>
> That kills the -no-kvm switch, which allows a single binary to be used
> both with and without kvm. Or do you think both target-i386+kvm and
> target-kvm ought to be kept?
>
>
I meant here: target-i386 (with kvm support) and target-i386-kvm.
> My thinking about qemu integration is that kqemu/kvm code needs to be
> abstracted into an API to reduce the #ifdefing in qemu, and that API
> could call kqemu or kvm as appropriate.
>
>
>
--
error compiling committee.c: too many arguments to function
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DAA108.8000501-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-02-20 9:48 ` Avi Kivity
@ 2007-02-20 15:13 ` Anthony Liguori
1 sibling, 0 replies; 13+ messages in thread
From: Anthony Liguori @ 2007-02-20 15:13 UTC (permalink / raw)
To: Avi Kivity; +Cc: kvm-devel
Avi Kivity wrote:
> Anthony Liguori wrote:
>> Howdy,
>>
>> Here's a tiny patch that adds a i386-kvm target. The main difference
>> between the i386-kvm and i386-softmmu target is that the -kvm target
>> does not have any of the dyngen infrastructure. This means that it
>> will build with gcc-4. I know you can do --cc=gcc to use gcc-4 but
>> quite a few versions of gcc-4 have trouble with compiling dyngen.
>>
>> I also suspect this may prove useful down the road. What do other
>> people think? I'm not terribly tied to the i386-kvm name for what
>> it's worth.
>>
>
> That kills the -no-kvm switch, which allows a single binary to be used
> both with and without kvm. Or do you think both target-i386+kvm and
> target-kvm ought to be kept?
I think both should be kept. I think configure ought to be smart enough
to figure out if gcc-3 is available and build the appropriate targets.
Adding the new target is a small enough patch that it won't be offensive
and should make a lot of users happy.
> My thinking about qemu integration is that kqemu/kvm code needs to be
> abstracted into an API to reduce the #ifdefing in qemu, and that API
> could call kqemu or kvm as appropriate.
That's definitely not a bad idea.
Regards,
Anthony Liguori
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DAC3FF.1090208-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
@ 2007-02-20 15:19 ` Anthony Liguori
[not found] ` <45DB119A.8000809-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
0 siblings, 1 reply; 13+ messages in thread
From: Anthony Liguori @ 2007-02-20 15:19 UTC (permalink / raw)
To: Avi Kivity; +Cc: kvm-devel
Avi Kivity wrote:
> Avi Kivity wrote:
>> Anthony Liguori wrote:
>>
>>> Howdy,
>>>
>>> Here's a tiny patch that adds a i386-kvm target. The main
>>> difference between the i386-kvm and i386-softmmu target is that the
>>> -kvm target does not have any of the dyngen infrastructure. This
>>> means that it will build with gcc-4. I know you can do --cc=gcc to
>>> use gcc-4 but quite a few versions of gcc-4 have trouble with
>>> compiling dyngen.
>>>
>>> I also suspect this may prove useful down the road. What do other
>>> people think? I'm not terribly tied to the i386-kvm name for what
>>> it's worth.
>>>
>>>
>>
>> That kills the -no-kvm switch, which allows a single binary to be
>> used both with and without kvm. Or do you think both target-i386+kvm
>> and target-kvm ought to be kept?
>>
>>
>
> I meant here: target-i386 (with kvm support) and target-i386-kvm.
target-i386 is an architecture. It can currently generate i386-softmmu,
i386-user, x86_64-softmmu, x86_64-user. We're simply adding the ability
to generate i386-kvm and x86_64-kvm. I'd like to find another name.
I still want to be able to support kvm/qemu in the normal i386-softmmu.
Regards,
Anthony Liguori
>> My thinking about qemu integration is that kqemu/kvm code needs to be
>> abstracted into an API to reduce the #ifdefing in qemu, and that API
>> could call kqemu or kvm as appropriate.
>>
>>
>>
>
>
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DB119A.8000809-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
@ 2007-02-20 15:21 ` Avi Kivity
[not found] ` <45DB11FD.1050900-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
0 siblings, 1 reply; 13+ messages in thread
From: Avi Kivity @ 2007-02-20 15:21 UTC (permalink / raw)
To: Anthony Liguori; +Cc: kvm-devel
Anthony Liguori wrote:
>>>
>>> That kills the -no-kvm switch, which allows a single binary to be
>>> used both with and without kvm. Or do you think both
>>> target-i386+kvm and target-kvm ought to be kept?
>>>
>>>
>>
>> I meant here: target-i386 (with kvm support) and target-i386-kvm.
>
> target-i386 is an architecture. It can currently generate
> i386-softmmu, i386-user, x86_64-softmmu, x86_64-user. We're simply
> adding the ability to generate i386-kvm and x86_64-kvm. I'd like to
> find another name.
>
> I still want to be able to support kvm/qemu in the normal i386-softmmu.
Thanks, that clears it up. I'll apply the patch.
--
error compiling committee.c: too many arguments to function
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DB11FD.1050900-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
@ 2007-02-20 15:26 ` Anthony Liguori
[not found] ` <45DB1326.3060705-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
0 siblings, 1 reply; 13+ messages in thread
From: Anthony Liguori @ 2007-02-20 15:26 UTC (permalink / raw)
To: Avi Kivity; +Cc: kvm-devel
Avi Kivity wrote:
> Anthony Liguori wrote:
>>>>
>>>> That kills the -no-kvm switch, which allows a single binary to be
>>>> used both with and without kvm. Or do you think both
>>>> target-i386+kvm and target-kvm ought to be kept?
>>>>
>>>>
>>>
>>> I meant here: target-i386 (with kvm support) and target-i386-kvm.
>>
>> target-i386 is an architecture. It can currently generate
>> i386-softmmu, i386-user, x86_64-softmmu, x86_64-user. We're simply
>> adding the ability to generate i386-kvm and x86_64-kvm. I'd like to
>> find another name.
>>
>> I still want to be able to support kvm/qemu in the normal i386-softmmu.
>
> Thanks, that clears it up. I'll apply the patch.
Can you wait on that? I'll submit a better one tonight. It was just an
RFC :-)
Regards,
Anthony Liguori
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <45DB1326.3060705-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
@ 2007-02-20 15:29 ` Avi Kivity
0 siblings, 0 replies; 13+ messages in thread
From: Avi Kivity @ 2007-02-20 15:29 UTC (permalink / raw)
To: Anthony Liguori; +Cc: kvm-devel
Anthony Liguori wrote:
>
> Can you wait on that? I'll submit a better one tonight. It was just
> an RFC :-)
>
Ok.
--
error compiling committee.c: too many arguments to function
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <fdaac4d50702191855gc2fc393ga83787bbf2d967da-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2007-02-20 3:51 ` Anthony Liguori
@ 2007-02-20 23:34 ` Rusty Russell
[not found] ` <1172014483.24118.13.camel-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org>
1 sibling, 1 reply; 13+ messages in thread
From: Rusty Russell @ 2007-02-20 23:34 UTC (permalink / raw)
To: Jun Koi; +Cc: kvm-devel
On Tue, 2007-02-20 at 11:55 +0900, Jun Koi wrote:
> Howdy Anthony,
>
> > Here's a tiny patch that adds a i386-kvm target. The main difference
> > between the i386-kvm and i386-softmmu target is that the -kvm target
> > does not have any of the dyngen infrastructure. This means that it will
> > build with gcc-4. I know you can do --cc=gcc to use gcc-4 but quite a
> > few versions of gcc-4 have trouble with compiling dyngen.
>
> I understand that your approach is to remove the "binary translation"
> part of the vanilla qemu, because we dont need it. We only need the
> device emulation code of qemu for kvm. Do I understand correctly?
>
> If so, we do not need to depend on gcc3 tricks anymore. Removing the
> dependent on gcc3 is really an important achievement.
Does your patch still call the binary qemu? It'd be nice for
distributions to have a real qemu still available as 'qemu', and the
kvm-specific one as something else.
Cheers,
Rusty.
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
[not found] ` <1172014483.24118.13.camel-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org>
@ 2007-02-21 0:17 ` Anthony Liguori
0 siblings, 0 replies; 13+ messages in thread
From: Anthony Liguori @ 2007-02-21 0:17 UTC (permalink / raw)
To: Rusty Russell; +Cc: kvm-devel
Rusty Russell wrote:
> On Tue, 2007-02-20 at 11:55 +0900, Jun Koi wrote:
>
>> Howdy Anthony,
>>
>>
>>> Here's a tiny patch that adds a i386-kvm target. The main difference
>>> between the i386-kvm and i386-softmmu target is that the -kvm target
>>> does not have any of the dyngen infrastructure. This means that it will
>>> build with gcc-4. I know you can do --cc=gcc to use gcc-4 but quite a
>>> few versions of gcc-4 have trouble with compiling dyngen.
>>>
>> I understand that your approach is to remove the "binary translation"
>> part of the vanilla qemu, because we dont need it. We only need the
>> device emulation code of qemu for kvm. Do I understand correctly?
>>
>> If so, we do not need to depend on gcc3 tricks anymore. Removing the
>> dependent on gcc3 is really an important achievement.
>>
>
> Does your patch still call the binary qemu? It'd be nice for
> distributions to have a real qemu still available as 'qemu', and the
> kvm-specific one as something else.
>
It calls it qemu-kvm (or qemu-system-x86_64-kvm). I'd like to find a
better name, perhaps qemu-system-i386-accel or something.
Regards,
Anthony Liguori
> Cheers,
> Rusty.
>
>
>
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [RFC] QEMU KVM target
@ 2007-02-20 9:44 Dylan Taft
0 siblings, 0 replies; 13+ messages in thread
From: Dylan Taft @ 2007-02-20 9:44 UTC (permalink / raw)
To: kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f
Sweet deal. I just applied the patch to v14, and compiled with gcc
4.1.1. Thanks a lot!
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
^ permalink raw reply [flat|nested] 13+ messages in thread
end of thread, other threads:[~2007-02-21 0:17 UTC | newest]
Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-02-20 1:08 [RFC] QEMU KVM target Anthony Liguori
[not found] ` <45DA4A16.4050804-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
2007-02-20 2:55 ` Jun Koi
[not found] ` <fdaac4d50702191855gc2fc393ga83787bbf2d967da-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2007-02-20 3:51 ` Anthony Liguori
2007-02-20 23:34 ` Rusty Russell
[not found] ` <1172014483.24118.13.camel-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org>
2007-02-21 0:17 ` Anthony Liguori
2007-02-20 7:19 ` Avi Kivity
[not found] ` <45DAA108.8000501-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-02-20 9:48 ` Avi Kivity
[not found] ` <45DAC3FF.1090208-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-02-20 15:19 ` Anthony Liguori
[not found] ` <45DB119A.8000809-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
2007-02-20 15:21 ` Avi Kivity
[not found] ` <45DB11FD.1050900-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-02-20 15:26 ` Anthony Liguori
[not found] ` <45DB1326.3060705-rdkfGonbjUSkNkDKm+mE6A@public.gmane.org>
2007-02-20 15:29 ` Avi Kivity
2007-02-20 15:13 ` Anthony Liguori
2007-02-20 9:44 Dylan Taft
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.