qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Anthony Liguori <anthony@codemonkey.ws>
To: Avi Kivity <avi@redhat.com>
Cc: qemu-devel@nongnu.org, kvm@vger.kernel.org
Subject: Re: [Qemu-devel] [PATCH 14/23] memory: transaction API
Date: Mon, 25 Jul 2011 14:16:26 -0500	[thread overview]
Message-ID: <4E2DC10A.4040005@codemonkey.ws> (raw)
In-Reply-To: <1311602584-23409-15-git-send-email-avi@redhat.com>

On 07/25/2011 09:02 AM, Avi Kivity wrote:
> Allow changes to the memory hierarchy to be accumulated and
> made visible all at once.  This reduces computational effort,
> especially when an accelerator (e.g. kvm) is involved.
>
> Useful when a single register update causes multiple changes
> to an address space.
>
> Signed-off-by: Avi Kivity<avi@redhat.com>

What's the motivation for this?  Was this just because it seemed neat to 
do or did you run into a performance issue you were trying to solve?


> ---
>   memory.c |   20 ++++++++++++++++++++
>   memory.h |    8 ++++++++
>   2 files changed, 28 insertions(+), 0 deletions(-)
>
> diff --git a/memory.c b/memory.c
> index 009ad33..370e189 100644
> --- a/memory.c
> +++ b/memory.c
> @@ -18,6 +18,8 @@
>   #include "kvm.h"
>   #include<assert.h>
>
> +unsigned memory_region_transaction_depth = 0;

Shouldn't this be static?

Regards,

Anthony Liguori

> +
>   typedef struct AddrRange AddrRange;
>
>   struct AddrRange {
> @@ -623,6 +625,10 @@ static void address_space_update_topology(AddressSpace *as)
>
>   static void memory_region_update_topology(void)
>   {
> +    if (memory_region_transaction_depth) {
> +        return;
> +    }
> +
>       if (address_space_memory.root) {
>           address_space_update_topology(&address_space_memory);
>       }
> @@ -631,6 +637,20 @@ static void memory_region_update_topology(void)
>       }
>   }
>
> +void memory_region_transaction_begin(void)
> +{
> +    ++memory_region_transaction_depth;
> +}
> +
> +void memory_region_transaction_commit(void)
> +{
> +    if (!memory_region_transaction_depth) {
> +        abort();
> +    }
> +    --memory_region_transaction_depth;
> +    memory_region_update_topology();
> +}
> +
>   void memory_region_init(MemoryRegion *mr,
>                           const char *name,
>                           uint64_t size)
> diff --git a/memory.h b/memory.h
> index e4c0ad1..cb3a9b6 100644
> --- a/memory.h
> +++ b/memory.h
> @@ -246,6 +246,14 @@ void memory_region_add_subregion_overlap(MemoryRegion *mr,
>   void memory_region_del_subregion(MemoryRegion *mr,
>                                    MemoryRegion *subregion);
>
> +/* Start a transaction; changes will be accumulated and made visible only
> + * when the transaction ends.
> + */
> +void memory_region_transaction_begin(void);
> +/* Commit a transaction and make changes visible to the guest.
> + */
> +void memory_region_transaction_commit(void);
> +
>   #endif
>
>   #endif

  reply	other threads:[~2011-07-25 19:16 UTC|newest]

Thread overview: 66+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-25 14:02 [Qemu-devel] [PATCH 00/23] Memory API, batch 1 Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 01/23] Hierarchical memory region API Avi Kivity
2011-07-25 18:41   ` Anthony Liguori
2011-07-26  9:35     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 02/23] memory: implement dirty tracking Avi Kivity
2011-07-25 18:43   ` Anthony Liguori
2011-07-25 14:02 ` [Qemu-devel] [PATCH 03/23] memory: merge adjacent segments of a single memory region Avi Kivity
2011-07-25 18:48   ` Anthony Liguori
2011-07-26  9:55     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 04/23] Internal interfaces for memory API Avi Kivity
2011-07-25 18:49   ` Anthony Liguori
2011-07-25 14:02 ` [Qemu-devel] [PATCH 05/23] memory: abstract address space operations Avi Kivity
2011-07-25 18:51   ` Anthony Liguori
2011-07-25 14:02 ` [Qemu-devel] [PATCH 06/23] memory: rename MemoryRegion::has_ram_addr to ::terminates Avi Kivity
2011-07-25 18:56   ` Anthony Liguori
2011-07-26  9:59     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 07/23] memory: late initialization of ram_addr Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 08/23] memory: I/O address space support Avi Kivity
2011-07-25 19:00   ` Anthony Liguori
2011-07-25 14:02 ` [Qemu-devel] [PATCH 09/23] memory: add backward compatibility for old portio registration Avi Kivity
2011-07-25 19:01   ` Anthony Liguori
2011-07-25 14:02 ` [Qemu-devel] [PATCH 10/23] memory: add backward compatibility for old mmio registration Avi Kivity
2011-07-25 19:02   ` Anthony Liguori
2011-07-25 14:02 ` [Qemu-devel] [PATCH 11/23] memory: add ioeventfd support Avi Kivity
2011-07-25 15:16   ` malc
2011-07-25 15:17     ` Avi Kivity
2011-07-25 15:22       ` malc
2011-07-25 15:28         ` Avi Kivity
2011-07-25 15:38           ` malc
2011-07-25 15:43             ` Avi Kivity
2011-07-25 19:08   ` Anthony Liguori
2011-07-26 10:08     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 12/23] memory: separate building the final memory map into two steps Avi Kivity
2011-07-25 19:12   ` Anthony Liguori
2011-07-26 10:43     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 13/23] memory: document the memory API Avi Kivity
2011-07-25 19:15   ` Anthony Liguori
2011-07-26 10:44     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 14/23] memory: transaction API Avi Kivity
2011-07-25 19:16   ` Anthony Liguori [this message]
2011-07-26 10:48     ` Avi Kivity
2011-07-26 11:39       ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 15/23] exec.c: initialize memory map Avi Kivity
2011-07-25 19:17   ` Anthony Liguori
2011-07-26 10:55     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 16/23] ioport: register ranges by byte aligned addresses always Avi Kivity
2011-07-25 19:20   ` Anthony Liguori
2011-07-26 10:59     ` Avi Kivity
2011-07-25 14:02 ` [Qemu-devel] [PATCH 17/23] pc: grab system_memory Avi Kivity
2011-07-25 19:22   ` Anthony Liguori
2011-07-25 14:02 ` [Qemu-devel] [PATCH 18/23] pc: convert pc_memory_init() to memory API Avi Kivity
2011-07-25 19:23   ` Anthony Liguori
2011-07-25 14:03 ` [Qemu-devel] [PATCH 19/23] pc: move global memory map out of pc_init1() and into its callers Avi Kivity
2011-07-25 20:02   ` Anthony Liguori
2011-07-26 11:02     ` Avi Kivity
2011-07-25 14:03 ` [Qemu-devel] [PATCH 20/23] pci: pass address space to pci bus when created Avi Kivity
2011-07-25 20:03   ` Anthony Liguori
2011-07-25 14:03 ` [Qemu-devel] [PATCH 21/23] pci: add MemoryRegion based BAR management API Avi Kivity
2011-07-25 20:20   ` Anthony Liguori
2011-07-26 11:06     ` Avi Kivity
2011-07-25 14:03 ` [Qemu-devel] [PATCH 22/23] sysbus: add MemoryRegion based memory " Avi Kivity
2011-07-25 20:21   ` Anthony Liguori
2011-07-25 14:03 ` [Qemu-devel] [PATCH 23/23] usb-ohci: convert to MemoryRegion Avi Kivity
2011-07-25 20:22   ` Anthony Liguori
2011-07-25 20:23 ` [Qemu-devel] [PATCH 00/23] Memory API, batch 1 Anthony Liguori
2011-07-26 11:32   ` Avi Kivity

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=4E2DC10A.4040005@codemonkey.ws \
    --to=anthony@codemonkey.ws \
    --cc=avi@redhat.com \
    --cc=kvm@vger.kernel.org \
    --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 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).