From: Peter Xu <peterx@redhat.com>
To: Joao Martins <joao.m.martins@oracle.com>
Cc: qemu-devel@nongnu.org,
Alex Williamson <alex.williamson@redhat.com>,
Cedric Le Goater <clg@redhat.com>,
Paolo Bonzini <pbonzini@redhat.com>,
David Hildenbrand <david@redhat.com>,
Philippe Mathieu-Daude <philmd@linaro.org>,
Avihai Horon <avihaih@nvidia.com>
Subject: Re: [PATCH v2 1/2] exec/ram_addr: return nr of dirty pages in cpu_physical_memory_set_dirty_lebitmap()
Date: Thu, 25 May 2023 09:16:42 -0400 [thread overview]
Message-ID: <ZG9fungk+x/F0vD+@x1n> (raw)
In-Reply-To: <20230525114321.71066-2-joao.m.martins@oracle.com>
On Thu, May 25, 2023 at 12:43:20PM +0100, Joao Martins wrote:
> In preparation for including the number of dirty pages in the
> vfio_get_dirty_bitmap() tracepoint, return the number of dirty pages in
> cpu_physical_memory_set_dirty_lebitmap() similar to
> cpu_physical_memory_sync_dirty_bitmap().
The patch itself looks good to me, but it's slightly different from sync
version because that was only for MIGRATION bitmap, meanwhile it counts
newly dirtied ones (so exclude already dirtied ones even if re-dirtied in
the MIGRATION bitmap), while this one counts any dirty bits in *bitmap.
Shall we perhaps state it somewhere explicitly? A comment for retval might
be suitable above the function?
Thanks,
>
> To avoid counting twice when GLOBAL_DIRTY_RATE is enabled, stash the
> number of bits set per bitmap quad in a variable (@nbits) and reuse it
> there.
>
> Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
> ---
> include/exec/ram_addr.h | 21 +++++++++++++++------
> 1 file changed, 15 insertions(+), 6 deletions(-)
>
> diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
> index f4fb6a211175..8b8f271d0731 100644
> --- a/include/exec/ram_addr.h
> +++ b/include/exec/ram_addr.h
> @@ -333,14 +333,16 @@ static inline void cpu_physical_memory_set_dirty_range(ram_addr_t start,
> }
>
> #if !defined(_WIN32)
> -static inline void cpu_physical_memory_set_dirty_lebitmap(unsigned long *bitmap,
> - ram_addr_t start,
> - ram_addr_t pages)
> +static inline
> +uint64_t cpu_physical_memory_set_dirty_lebitmap(unsigned long *bitmap,
> + ram_addr_t start,
> + ram_addr_t pages)
> {
> unsigned long i, j;
> - unsigned long page_number, c;
> + unsigned long page_number, c, nbits;
> hwaddr addr;
> ram_addr_t ram_addr;
> + uint64_t num_dirty = 0;
> unsigned long len = (pages + HOST_LONG_BITS - 1) / HOST_LONG_BITS;
> unsigned long hpratio = qemu_real_host_page_size() / TARGET_PAGE_SIZE;
> unsigned long page = BIT_WORD(start >> TARGET_PAGE_BITS);
> @@ -368,6 +370,7 @@ static inline void cpu_physical_memory_set_dirty_lebitmap(unsigned long *bitmap,
> if (bitmap[k]) {
> unsigned long temp = leul_to_cpu(bitmap[k]);
>
> + nbits = ctpopl(temp);
> qatomic_or(&blocks[DIRTY_MEMORY_VGA][idx][offset], temp);
>
> if (global_dirty_tracking) {
> @@ -376,10 +379,12 @@ static inline void cpu_physical_memory_set_dirty_lebitmap(unsigned long *bitmap,
> temp);
> if (unlikely(
> global_dirty_tracking & GLOBAL_DIRTY_DIRTY_RATE)) {
> - total_dirty_pages += ctpopl(temp);
> + total_dirty_pages += nbits;
> }
> }
>
> + num_dirty += nbits;
> +
> if (tcg_enabled()) {
> qatomic_or(&blocks[DIRTY_MEMORY_CODE][idx][offset],
> temp);
> @@ -408,9 +413,11 @@ static inline void cpu_physical_memory_set_dirty_lebitmap(unsigned long *bitmap,
> for (i = 0; i < len; i++) {
> if (bitmap[i] != 0) {
> c = leul_to_cpu(bitmap[i]);
> + nbits = ctpopl(c);
> if (unlikely(global_dirty_tracking & GLOBAL_DIRTY_DIRTY_RATE)) {
> - total_dirty_pages += ctpopl(c);
> + total_dirty_pages += nbits;
> }
> + num_dirty += nbits;
> do {
> j = ctzl(c);
> c &= ~(1ul << j);
> @@ -423,6 +430,8 @@ static inline void cpu_physical_memory_set_dirty_lebitmap(unsigned long *bitmap,
> }
> }
> }
> +
> + return num_dirty;
> }
> #endif /* not _WIN32 */
>
> --
> 2.31.1
>
--
Peter Xu
next prev parent reply other threads:[~2023-05-25 13:17 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-25 11:43 [PATCH v2 0/2] hw/vfio: Improve vfio_get_dirty_bitmap() tracepoint Joao Martins
2023-05-25 11:43 ` [PATCH v2 1/2] exec/ram_addr: return nr of dirty pages in cpu_physical_memory_set_dirty_lebitmap() Joao Martins
2023-05-25 13:16 ` Peter Xu [this message]
2023-05-25 13:47 ` Joao Martins
2023-05-25 13:54 ` Peter Xu
2023-05-25 15:02 ` Joao Martins
2023-05-25 11:43 ` [PATCH v2 2/2] hw/vfio: Add nr of dirty pages to vfio_get_dirty_bitmap tracepoint Joao Martins
2023-05-25 11:52 ` Philippe Mathieu-Daudé
2023-05-25 13:36 ` Joao Martins
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=ZG9fungk+x/F0vD+@x1n \
--to=peterx@redhat.com \
--cc=alex.williamson@redhat.com \
--cc=avihaih@nvidia.com \
--cc=clg@redhat.com \
--cc=david@redhat.com \
--cc=joao.m.martins@oracle.com \
--cc=pbonzini@redhat.com \
--cc=philmd@linaro.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).