qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Pranith Kumar <bobby.prani@gmail.com>
To: "Alex Bennée" <alex.bennee@linaro.org>
Cc: qemu-devel@nongnu.org, Richard Henderson <rth@twiddle.net>,
	Peter Maydell <peter.maydell@linaro.org>
Subject: Re: [Qemu-devel] [PATCH v2] mttcg/i386: Patch instruction using async_safe_* framework
Date: Fri, 24 Feb 2017 09:10:14 -0500	[thread overview]
Message-ID: <87zihbaddl.fsf@gmail.com> (raw)
In-Reply-To: <87y3wvlwhx.fsf@linaro.org>


Hi Alex,

Alex Bennée writes:

> Pranith Kumar <bobby.prani@gmail.com> writes:
>
>> In mttcg, calling pause_all_vcpus() during execution from the
>> generated TBs causes a deadlock if some vCPU is waiting for exclusive
>> execution in start_exclusive(). Fix this by using the aync_safe_*
>> framework instead of pausing vcpus for patching instructions.
>>

<...>

>> diff --git a/hw/i386/kvmvapic.c b/hw/i386/kvmvapic.c
>> index 82a4955..11b0d49 100644
>> --- a/hw/i386/kvmvapic.c
>>  
>> -    resume_all_vcpus();
>> +    g_free(info);
>> +}
>> +
>> +static void patch_instruction(VAPICROMState *s, X86CPU *cpu, target_ulong ip)
>> +{
>> +    CPUState *cs = CPU(cpu);
>> +    VAPICHandlers *handlers;
>> +    struct PatchInfo *info;
>> +
>> +    if (smp_cpus == 1) {
>> +        handlers = &s->rom_state.up;
>> +    } else {
>> +        handlers = &s->rom_state.mp;
>> +    }
>> +
>> +    info  = g_new(struct PatchInfo, 1);
>> +    info->handler = handlers;
>> +    info->ip = ip;
>>  
>>      if (!kvm_enabled()) {
>> -        /* Both tb_lock and iothread_mutex will be reset when
>> -         *  longjmps back into the cpu_exec loop. */
>> -        tb_lock();
>> -        tb_gen_code(cs, current_pc, current_cs_base, current_flags, 1);
>> -        cpu_loop_exit_noexc(cs);
>> +        const run_on_cpu_func fn = do_patch_instruction;
>> +
>> +        async_safe_run_on_cpu(cs, fn, RUN_ON_CPU_HOST_PTR(info));
>> +        cs->exception_index = EXCP_INTERRUPT;
>> +        cpu_loop_exit(cs);
>>      }
>> +
>> +    pause_all_vcpus();
>> +
>> +    do_patch_instruction(cs, RUN_ON_CPU_HOST_PTR(info));
>> +
>> +    resume_all_vcpus();
>> +    g_free(info);
>
> I don't know if there is any benefit scheduling this as async work for
> KVM but I'll leave that up to Paolo to decide. From a TCG point of view
> I think its good:
>

We are scheduling this as async work only for non-KVM cases. For KVM, we use
go to the pause/resume path above and patch it there itself.

Thanks,
-- 
Pranith

  reply	other threads:[~2017-02-24 14:10 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-02-24  5:42 [Qemu-devel] [PATCH v2] mttcg/i386: Patch instruction using async_safe_* framework Pranith Kumar
2017-02-24  6:54 ` Richard Henderson
2017-02-24 10:21 ` Alex Bennée
2017-02-24 14:10   ` Pranith Kumar [this message]
2017-02-24 14:36     ` Alex Bennée
2017-06-07 16:40 ` Pranith Kumar
2017-06-07 18:09   ` Alex Bennée
2017-06-07 18:46     ` Pranith Kumar

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=87zihbaddl.fsf@gmail.com \
    --to=bobby.prani@gmail.com \
    --cc=alex.bennee@linaro.org \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=rth@twiddle.net \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).