[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_s
From: |
Juan Quintela |
Subject: |
Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate |
Date: |
Wed, 11 Sep 2013 11:17:22 +0200 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/24.3 (gnu/linux) |
Lei Li <address@hidden> wrote:
> qemu_file_rate_limit() never return negative value since the refactor
> by Commit 1964a39, this patch gets rid of the negative check for it,
> adjust bytes_transferred and return value correspondingly in
> ram_save_iterate().
>
> Signed-off-by: Lei Li <address@hidden>
> Signed-off-by: Paolo Bonzini <address@hidden>
> ---
>
> Change since v1:
> Return fixes and improvement from Paolo Bonzini.
>
> arch_init.c | 15 ++++++++++-----
> 1 files changed, 10 insertions(+), 5 deletions(-)
>
> diff --git a/arch_init.c b/arch_init.c
> index 94d45e1..a26bc89 100644
> --- a/arch_init.c
> +++ b/arch_init.c
> @@ -709,15 +709,20 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
> */
> ram_control_after_iterate(f, RAM_CONTROL_ROUND);
>
> + bytes_transferred += total_sent;
Agreed.
> +
> + /*
> + * Do not count these 8 bytes into total_sent, so that we can
> + * return 0 if no page had been dirtied.
> + */
> + qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
> + bytes_transferred += 8;
> +
> + ret = qemu_file_get_error(f);
> if (ret < 0) {
Not sure this is the right solution.
We are sending anyways RAM_SAVE_FLAG_EOS.
And I think that the right solution is make qemu_get_rate_limit() to
return -1 in case of error (or the error, I don't care). Looking at the
callers:
migration.c::migration_thread()
we check for error when we qemu_file_rate_limit() returns != 0. Well.
second call:
if (qemu_file_rate_limit(s->file)) {
/* usleep expects microseconds */
g_usleep((initial_time + BUFFER_DELAY - current_time)*1000);
}
if should be:
if (qemu_file_rate_limit(s->file) == 1)
block_migration: not correct, we don't check for the error.
arch_init.c:
check is correct, but we need to return -ERROR in case of errors.
hw/ppc/spapr.c:
will work correctly even if changed to -ERROR.
savevm.c: qemu_savevm_state_iterate()
if (qemu_file_rate_limit(f)) {
return 0;
}
check is incorrect again, we should return an error if there is one
error.
I think that returning qemu_rate_limit() to return 0/1/negative makes sense.
Thoughts?
Thanks, Juan.
- [Qemu-devel] [PATCH 0/3 resend v2] Migration fix, Lei Li, 2013/09/04
- [Qemu-devel] [PATCH 1/3 resend v2] savevm: add comments for qemu_file_get_error(), Lei Li, 2013/09/04
- [Qemu-devel] [PATCH 2/3 resend v2] savevm: fix wrong initialization by ram_control_load_hook, Lei Li, 2013/09/04
- [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate, Lei Li, 2013/09/04
- Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate,
Juan Quintela <=
- Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate, Paolo Bonzini, 2013/09/11
- Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate, Juan Quintela, 2013/09/11
- Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate, Paolo Bonzini, 2013/09/11
- Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate, Lei Li, 2013/09/12
- Re: [Qemu-devel] [PATCH 3/3 resend v2] arch_init: right return for ram_save_iterate, Orit Wasserman, 2013/09/12
Re: [Qemu-devel] [PATCH 0/3 resend v2] Migration fix, Orit Wasserman, 2013/09/04