From: Jan Kiszka <jan.kiszka@web.de>
To: Blue Swirl <blauwirbel@gmail.com>
Cc: Anthony Liguori <aliguori@us.ibm.com>,
kvm@vger.kernel.org, Gleb Natapov <gleb@redhat.com>,
Marcelo Tosatti <mtosatti@redhat.com>,
qemu-devel <qemu-devel@nongnu.org>, Avi Kivity <avi@redhat.com>
Subject: Re: [PATCH v2 1/8] kvm: Set cpu_single_env only once
Date: Sat, 11 Feb 2012 11:06:42 +0100 [thread overview]
Message-ID: <4F363DB2.3080908@web.de> (raw)
In-Reply-To: <CAAu8pHt4dR5vp3wM4nVdnwoyxBUBi5L-LkskV=CQ9P0fNOLD9Q@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 2290 bytes --]
On 2012-02-11 11:02, Blue Swirl wrote:
> On Fri, Feb 10, 2012 at 18:31, Jan Kiszka <jan.kiszka@siemens.com> wrote:
>> As we have thread-local cpu_single_env now and KVM uses exactly one
>> thread per VCPU, we can drop the cpu_single_env updates from the loop
>> and initialize this variable only once during setup.
>
> I don't think this is correct. Maybe you missed the part that sets
> cpu_single_env to NULL, which I think is to annoy broken code that
> assumes that some CPU state is always globally available. This is not
> true for monitor context.
I did check this before changing, and I see no such need. Particularly
as this old debugging help prevents valid use case.
Jan
>
>> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>> ---
>> cpus.c | 1 +
>> kvm-all.c | 5 -----
>> 2 files changed, 1 insertions(+), 5 deletions(-)
>>
>> diff --git a/cpus.c b/cpus.c
>> index f45a438..d0c8340 100644
>> --- a/cpus.c
>> +++ b/cpus.c
>> @@ -714,6 +714,7 @@ static void *qemu_kvm_cpu_thread_fn(void *arg)
>> qemu_mutex_lock(&qemu_global_mutex);
>> qemu_thread_get_self(env->thread);
>> env->thread_id = qemu_get_thread_id();
>> + cpu_single_env = env;
>>
>> r = kvm_init_vcpu(env);
>> if (r < 0) {
>> diff --git a/kvm-all.c b/kvm-all.c
>> index c4babda..e2cbc03 100644
>> --- a/kvm-all.c
>> +++ b/kvm-all.c
>> @@ -1118,8 +1118,6 @@ int kvm_cpu_exec(CPUState *env)
>> return EXCP_HLT;
>> }
>>
>> - cpu_single_env = env;
>> -
>> do {
>> if (env->kvm_vcpu_dirty) {
>> kvm_arch_put_registers(env, KVM_PUT_RUNTIME_STATE);
>> @@ -1136,13 +1134,11 @@ int kvm_cpu_exec(CPUState *env)
>> */
>> qemu_cpu_kick_self();
>> }
>> - cpu_single_env = NULL;
>> qemu_mutex_unlock_iothread();
>>
>> run_ret = kvm_vcpu_ioctl(env, KVM_RUN, 0);
>>
>> qemu_mutex_lock_iothread();
>> - cpu_single_env = env;
>> kvm_arch_post_run(env, run);
>>
>> kvm_flush_coalesced_mmio_buffer();
>> @@ -1206,7 +1202,6 @@ int kvm_cpu_exec(CPUState *env)
>> }
>>
>> env->exit_request = 0;
>> - cpu_single_env = NULL;
>> return ret;
>> }
>>
>> --
>> 1.7.3.4
>>
>>
>
>
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]
WARNING: multiple messages have this Message-ID (diff)
From: Jan Kiszka <jan.kiszka@web.de>
To: Blue Swirl <blauwirbel@gmail.com>
Cc: Anthony Liguori <aliguori@us.ibm.com>,
kvm@vger.kernel.org, Gleb Natapov <gleb@redhat.com>,
Marcelo Tosatti <mtosatti@redhat.com>,
qemu-devel <qemu-devel@nongnu.org>, Avi Kivity <avi@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v2 1/8] kvm: Set cpu_single_env only once
Date: Sat, 11 Feb 2012 11:06:42 +0100 [thread overview]
Message-ID: <4F363DB2.3080908@web.de> (raw)
In-Reply-To: <CAAu8pHt4dR5vp3wM4nVdnwoyxBUBi5L-LkskV=CQ9P0fNOLD9Q@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 2290 bytes --]
On 2012-02-11 11:02, Blue Swirl wrote:
> On Fri, Feb 10, 2012 at 18:31, Jan Kiszka <jan.kiszka@siemens.com> wrote:
>> As we have thread-local cpu_single_env now and KVM uses exactly one
>> thread per VCPU, we can drop the cpu_single_env updates from the loop
>> and initialize this variable only once during setup.
>
> I don't think this is correct. Maybe you missed the part that sets
> cpu_single_env to NULL, which I think is to annoy broken code that
> assumes that some CPU state is always globally available. This is not
> true for monitor context.
I did check this before changing, and I see no such need. Particularly
as this old debugging help prevents valid use case.
Jan
>
>> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>> ---
>> cpus.c | 1 +
>> kvm-all.c | 5 -----
>> 2 files changed, 1 insertions(+), 5 deletions(-)
>>
>> diff --git a/cpus.c b/cpus.c
>> index f45a438..d0c8340 100644
>> --- a/cpus.c
>> +++ b/cpus.c
>> @@ -714,6 +714,7 @@ static void *qemu_kvm_cpu_thread_fn(void *arg)
>> qemu_mutex_lock(&qemu_global_mutex);
>> qemu_thread_get_self(env->thread);
>> env->thread_id = qemu_get_thread_id();
>> + cpu_single_env = env;
>>
>> r = kvm_init_vcpu(env);
>> if (r < 0) {
>> diff --git a/kvm-all.c b/kvm-all.c
>> index c4babda..e2cbc03 100644
>> --- a/kvm-all.c
>> +++ b/kvm-all.c
>> @@ -1118,8 +1118,6 @@ int kvm_cpu_exec(CPUState *env)
>> return EXCP_HLT;
>> }
>>
>> - cpu_single_env = env;
>> -
>> do {
>> if (env->kvm_vcpu_dirty) {
>> kvm_arch_put_registers(env, KVM_PUT_RUNTIME_STATE);
>> @@ -1136,13 +1134,11 @@ int kvm_cpu_exec(CPUState *env)
>> */
>> qemu_cpu_kick_self();
>> }
>> - cpu_single_env = NULL;
>> qemu_mutex_unlock_iothread();
>>
>> run_ret = kvm_vcpu_ioctl(env, KVM_RUN, 0);
>>
>> qemu_mutex_lock_iothread();
>> - cpu_single_env = env;
>> kvm_arch_post_run(env, run);
>>
>> kvm_flush_coalesced_mmio_buffer();
>> @@ -1206,7 +1202,6 @@ int kvm_cpu_exec(CPUState *env)
>> }
>>
>> env->exit_request = 0;
>> - cpu_single_env = NULL;
>> return ret;
>> }
>>
>> --
>> 1.7.3.4
>>
>>
>
>
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]
next prev parent reply other threads:[~2012-02-11 10:06 UTC|newest]
Thread overview: 90+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-10 18:31 [PATCH v2 0/8] uq/master: TPR access optimization for Windows guests Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-10 18:31 ` [PATCH v2 1/8] kvm: Set cpu_single_env only once Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-11 10:02 ` Blue Swirl
2012-02-11 10:02 ` [Qemu-devel] " Blue Swirl
2012-02-11 10:06 ` Jan Kiszka [this message]
2012-02-11 10:06 ` Jan Kiszka
2012-02-11 11:25 ` Blue Swirl
2012-02-11 11:25 ` [Qemu-devel] " Blue Swirl
2012-02-11 11:49 ` Andreas Färber
2012-02-11 11:49 ` [Qemu-devel] " Andreas Färber
2012-02-11 12:43 ` Jan Kiszka
2012-02-11 12:43 ` Jan Kiszka
2012-02-11 13:06 ` Andreas Färber
2012-02-11 13:06 ` [Qemu-devel] " Andreas Färber
2012-02-11 13:07 ` Jan Kiszka
2012-02-11 13:07 ` Jan Kiszka
2012-02-11 13:21 ` Andreas Färber
2012-02-11 13:21 ` Andreas Färber
2012-02-11 13:35 ` Jan Kiszka
2012-02-11 13:35 ` [Qemu-devel] " Jan Kiszka
2012-02-11 13:59 ` Andreas Färber
2012-02-11 13:59 ` [Qemu-devel] " Andreas Färber
2012-02-11 14:02 ` Jan Kiszka
2012-02-11 14:02 ` [Qemu-devel] " Jan Kiszka
2012-02-11 14:12 ` Andreas Färber
2012-02-11 14:12 ` Andreas Färber
2012-02-11 14:24 ` Jan Kiszka
2012-02-11 14:24 ` Jan Kiszka
2012-02-11 14:49 ` Andreas Färber
2012-02-11 14:49 ` [Qemu-devel] " Andreas Färber
2012-02-13 8:17 ` Paolo Bonzini
2012-02-13 8:17 ` Paolo Bonzini
2012-02-11 13:54 ` Blue Swirl
2012-02-11 13:54 ` Blue Swirl
2012-02-11 14:00 ` Jan Kiszka
2012-02-11 14:00 ` [Qemu-devel] " Jan Kiszka
2012-02-11 14:11 ` Blue Swirl
2012-02-11 14:11 ` Blue Swirl
2012-02-11 14:18 ` Jan Kiszka
2012-02-11 14:18 ` Jan Kiszka
2012-02-11 14:23 ` Blue Swirl
2012-02-11 14:23 ` Blue Swirl
2012-02-11 12:41 ` Jan Kiszka
2012-02-11 12:41 ` [Qemu-devel] " Jan Kiszka
2012-02-10 18:31 ` [PATCH v2 2/8] Allow to use pause_all_vcpus from VCPU context Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-11 14:16 ` Blue Swirl
2012-02-11 14:16 ` [Qemu-devel] " Blue Swirl
2012-02-11 14:31 ` Jan Kiszka
2012-02-11 14:31 ` [Qemu-devel] " Jan Kiszka
2012-02-10 18:31 ` [PATCH v2 3/8] target-i386: Add infrastructure for reporting TPR MMIO accesses Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-11 14:32 ` Blue Swirl
2012-02-11 14:32 ` Blue Swirl
2012-02-10 18:31 ` [PATCH v2 4/8] kvmvapic: Add option ROM Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-10 18:31 ` [PATCH v2 5/8] kvmvapic: Introduce TPR access optimization for Windows guests Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-11 15:25 ` Blue Swirl
2012-02-11 15:25 ` [Qemu-devel] " Blue Swirl
2012-02-13 10:16 ` Jan Kiszka
2012-02-13 10:16 ` Jan Kiszka
2012-02-13 18:50 ` Blue Swirl
2012-02-13 18:50 ` Blue Swirl
2012-02-13 19:11 ` Gleb Natapov
2012-02-13 19:11 ` Gleb Natapov
2012-02-13 19:22 ` Jan Kiszka
2012-02-13 19:22 ` [Qemu-devel] " Jan Kiszka
2012-02-14 7:54 ` Gleb Natapov
2012-02-14 7:54 ` [Qemu-devel] " Gleb Natapov
2012-02-14 8:55 ` Jan Kiszka
2012-02-14 8:55 ` [Qemu-devel] " Jan Kiszka
2012-02-14 8:59 ` Gleb Natapov
2012-02-14 8:59 ` [Qemu-devel] " Gleb Natapov
2012-02-10 18:31 ` [PATCH v2 6/8] kvmvapic: Simplify mp/up_set_tpr Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-10 18:31 ` [PATCH v2 7/8] optionsrom: Reserve space for checksum Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
2012-02-11 11:46 ` Andreas Färber
2012-02-11 11:46 ` [Qemu-devel] " Andreas Färber
2012-02-11 12:45 ` Jan Kiszka
2012-02-11 12:45 ` [Qemu-devel] " Jan Kiszka
2012-02-11 12:51 ` Andreas Färber
2012-02-11 12:51 ` [Qemu-devel] " Andreas Färber
2012-02-11 12:57 ` Jan Kiszka
2012-02-11 12:57 ` [Qemu-devel] " Jan Kiszka
2012-02-10 18:31 ` [PATCH v2 8/8] kvmvapic: Use optionrom helpers Jan Kiszka
2012-02-10 18:31 ` [Qemu-devel] " Jan Kiszka
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4F363DB2.3080908@web.de \
--to=jan.kiszka@web.de \
--cc=aliguori@us.ibm.com \
--cc=avi@redhat.com \
--cc=blauwirbel@gmail.com \
--cc=gleb@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=mtosatti@redhat.com \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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.