Hi all,
modularization of TCG is still incomplete in my view: we need to handle the
error case better.
If we configure --enable-modules --enable-tcg --enable-kvm , but for some
reason the .so is not present or is not loaded successfully,
and kvm fails to be enabled (for example, /dev/kvm not av
Everything is in place now, so re-enable modular tcg builds,
with alot more code actually being built modular now.
Signed-off-by: Gerd Hoffmann
---
meson.build | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meson.build b/meson.build
index e83aa4f5c3c9..bf6378481250 100644
Allows more fine-grained patches
without breaking bisecting.
Signed-off-by: Gerd Hoffmann
---
meson.build | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meson.build b/meson.build
index bf6378481250..e83aa4f5c3c9 100644
--- a/meson.build
+++ b/meson.build
@@ -94,7 +94,7 @@ en
Create tcg-module.[ch] files, with struct TCGModuleOps, empty for now.
Followup patches will add function pointers to the struct and stub
functions to tcg-module.c. That will effectively will switch stubs from
compile-time to runtime, which is needed to build tcg as module.
Signed-off-by: Gerd H
Gerd Hoffmann (7):
plugins: register qemu_plugin_opts using opts_init()
tcg/module: move hmp.c to tcg module
tcg/module: move cpu-exec-common.c from tcg_ss to specific_ss
tcg/module: add some infrastructure for modular tcg.
tcg/module: Add tlb_flush to TCGModuleOps
tcg/module: Add
On 7/30/21 11:05 AM, Gerd Hoffmann wrote:
> Hi,
>
> On Thu, Jul 29, 2021 at 06:40:17PM +0200, Paolo Bonzini wrote:
>> On 29/07/21 11:14, Gerd Hoffmann wrote:
>>> The common functions could be added TCGCPUOps to allow them being called via
>>> CPUClass->tcg_ops->$name instead of a direct symbol r
Hi,
On Thu, Jul 29, 2021 at 06:40:17PM +0200, Paolo Bonzini wrote:
> On 29/07/21 11:14, Gerd Hoffmann wrote:
> > The common functions could be added TCGCPUOps to allow them being called via
> > CPUClass->tcg_ops->$name instead of a direct symbol reference. Not sure
> > this
> > is a good idea
On 29/07/21 11:14, Gerd Hoffmann wrote:
The common functions could be added TCGCPUOps to allow them being called via
CPUClass->tcg_ops->$name instead of a direct symbol reference. Not sure this
is a good idea though. Experimental patch covering tcg_exec_realizefn +
tcg_exec_unrealizefn below.
Hi,
>>>>>
>>>>>> And another comment: I think we should have some progress on ARM with
>>>>>> the kvm/tcg split and with the KConfig of boards, before we continue
>>>>>> here.
>>>>>
>>>>> Why? Thi
e should have some progress on ARM with
>>>>> the kvm/tcg split and with the KConfig of boards, before we continue
>>>>> here.
>>>>
>>>> Why? This can easily be tacked in parallel. We can flip the switch
>>>> for modular tcg per tar
tcg split and with the KConfig of boards, before we continue
>>>> here.
>>>
>>> Why? This can easily be tacked in parallel. We can flip the switch
>>> for modular tcg per target in meson.build.
>>>
>>> take care,
>>> Gerd
>>&
tcg split and with the KConfig of boards, before we continue
>>>> here.
>>>
>>> Why? This can easily be tacked in parallel. We can flip the switch
>>> for modular tcg per target in meson.build.
>>>
>>> take care,
>>> Gerd
>>&
.
>>
>> Why? This can easily be tacked in parallel. We can flip the switch
>> for modular tcg per target in meson.build.
>>
>> take care,
>> Gerd
>>
>
> Because in the end we need to do this for ARM too and for the other archs too
> (s390 is al
On 7/29/21 12:29 PM, Gerd Hoffmann wrote:
> Hi,
>
>> And another comment: I think we should have some progress on ARM with
>> the kvm/tcg split and with the KConfig of boards, before we continue
>> here.
>
> Why? This can easily be tacked in parallel. We can fli
ic parts in my view. So you have a big
>> tcg-arm module, a tcg-i386 module etc.
>>
>> I think I sketched already the idea in the Makefile I shared before?
>
> We have target-specific modules in master branch.
> Used for qtest (all archs) and tcg (i386/x86_64 only, ac
Hi,
> And another comment: I think we should have some progress on ARM with
> the kvm/tcg split and with the KConfig of boards, before we continue
> here.
Why? This can easily be tacked in parallel. We can flip the switch
for modular tcg per target in meson.build.
take care,
Gerd
86/x86_64 only, accel ops only).
The build system changes to build more tcg bits modular are here:
https://git.kraxel.org/cgit/qemu/log/?h=sirius/modules-tcg-meson
Doesn't build due unresolved symbols, but shows which code
changes/cleanups/reorganizations are needed for (more) modular tcg.
take care,
Gerd
On 7/29/21 11:14 AM, Gerd Hoffmann wrote:
> Hi,
>
>>> So we need more work to make this actually work right.
>>
>> Yes. I want have all of tcg in the tcg accel module, not only parts of
>> it, but that needs some more refactoring. I'll go start looking at this
>> once I managed to wade through
On 7/29/21 11:14 AM, Gerd Hoffmann wrote:
> Hi,
>
>>> So we need more work to make this actually work right.
>>
>> Yes. I want have all of tcg in the tcg accel module, not only parts of
>> it, but that needs some more refactoring. I'll go start looking at this
>> once I managed to wade through
Hi,
> > So we need more work to make this actually work right.
>
> Yes. I want have all of tcg in the tcg accel module, not only parts of
> it, but that needs some more refactoring. I'll go start looking at this
> once I managed to wade through my vacation backlog.
So, changed the meson.buil
From: Paolo Bonzini
Accelerator modularity does not work on Darwin:
ld: illegal thread local variable reference to regular symbol _current_cpu for
architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
Fix by avoiding modular TCG builds.
Signed
Accelerator modularity does not work on Darwin:
ld: illegal thread local variable reference to regular symbol _current_cpu for
architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
Fix by avoiding modular TCG builds.
Signed-off-by: Paolo Bonzini
Paolo Bonzini writes:
> Accelerator modularity does not work on Darwin:
>
> ld: illegal thread local variable reference to regular symbol _current_cpu
> for architecture x86_64
> clang: error: linker command failed with exit code 1 (use -v to see
> invocation)
>
> Fix
Accelerator modularity does not work on Darwin:
ld: illegal thread local variable reference to regular symbol _current_cpu for
architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
Fix by avoiding modular TCG builds.
Signed-off-by: Paolo Bonzini
24 matches
Mail list logo
