From: Paolo Bonzini <pbonzini@redhat.com>
To: "Emilio G. Cota" <cota@braap.org>, qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH] translate-all: use bitmap helpers for PageDesc's bitmap
Date: Wed, 22 Apr 2015 15:13:05 +0200 [thread overview]
Message-ID: <55379E61.5080104@redhat.com> (raw)
In-Reply-To: <1426919142-20770-1-git-send-email-cota@braap.org>
On 21/03/2015 07:25, Emilio G. Cota wrote:
> Note that this test
> if (b & ((1 << len) - 1))
> can be simplified to
> if (b & 1)
> , since we know that iff the first bit of a tb is set,
> all other bits from that tb are set too.
I don't think this optimization is valid, unfortunately. It is possible
that say a tb starts at 0x12340001 and tb_invalidate_phys_page_fast is
called for the range starting at 0x12340000 and ending at 0x12340003.
In this case you need the full test that was used before.
Therefore, I am unqueuing this patch. Sorry. :)
Paolo
> Signed-off-by: Emilio G. Cota <cota@braap.org>
> ---
> translate-all.c | 39 +++++----------------------------------
> 1 file changed, 5 insertions(+), 34 deletions(-)
>
> diff --git a/translate-all.c b/translate-all.c
> index 11763c6..749d0f6 100644
> --- a/translate-all.c
> +++ b/translate-all.c
> @@ -59,6 +59,7 @@
>
> #include "exec/cputlb.h"
> #include "translate-all.h"
> +#include "qemu/bitmap.h"
> #include "qemu/timer.h"
>
> //#define DEBUG_TB_INVALIDATE
> @@ -79,7 +80,7 @@ typedef struct PageDesc {
> /* in order to optimize self modifying code, we count the number
> of lookups we do to a given page to use a bitmap */
> unsigned int code_write_count;
> - uint8_t *code_bitmap;
> + unsigned long *code_bitmap;
> #if defined(CONFIG_USER_ONLY)
> unsigned long flags;
> #endif
> @@ -978,39 +979,12 @@ void tb_phys_invalidate(TranslationBlock *tb, tb_page_addr_t page_addr)
> tcg_ctx.tb_ctx.tb_phys_invalidate_count++;
> }
>
> -static inline void set_bits(uint8_t *tab, int start, int len)
> -{
> - int end, mask, end1;
> -
> - end = start + len;
> - tab += start >> 3;
> - mask = 0xff << (start & 7);
> - if ((start & ~7) == (end & ~7)) {
> - if (start < end) {
> - mask &= ~(0xff << (end & 7));
> - *tab |= mask;
> - }
> - } else {
> - *tab++ |= mask;
> - start = (start + 8) & ~7;
> - end1 = end & ~7;
> - while (start < end1) {
> - *tab++ = 0xff;
> - start += 8;
> - }
> - if (start < end) {
> - mask = ~(0xff << (end & 7));
> - *tab |= mask;
> - }
> - }
> -}
> -
> static void build_page_bitmap(PageDesc *p)
> {
> int n, tb_start, tb_end;
> TranslationBlock *tb;
>
> - p->code_bitmap = g_malloc0(TARGET_PAGE_SIZE / 8);
> + p->code_bitmap = bitmap_new(TARGET_PAGE_SIZE);
>
> tb = p->first_tb;
> while (tb != NULL) {
> @@ -1029,7 +1003,7 @@ static void build_page_bitmap(PageDesc *p)
> tb_start = 0;
> tb_end = ((tb->pc + tb->size) & ~TARGET_PAGE_MASK);
> }
> - set_bits(p->code_bitmap, tb_start, tb_end - tb_start);
> + bitmap_set(p->code_bitmap, tb_start, tb_end - tb_start);
> tb = tb->page_next[n];
> }
> }
> @@ -1219,7 +1193,6 @@ void tb_invalidate_phys_page_range(tb_page_addr_t start, tb_page_addr_t end,
> void tb_invalidate_phys_page_fast(tb_page_addr_t start, int len)
> {
> PageDesc *p;
> - int offset, b;
>
> #if 0
> if (1) {
> @@ -1235,9 +1208,7 @@ void tb_invalidate_phys_page_fast(tb_page_addr_t start, int len)
> return;
> }
> if (p->code_bitmap) {
> - offset = start & ~TARGET_PAGE_MASK;
> - b = p->code_bitmap[offset >> 3] >> (offset & 7);
> - if (b & ((1 << len) - 1)) {
> + if (test_bit(start & ~TARGET_PAGE_MASK, p->code_bitmap)) {
> goto do_invalidate;
> }
> } else {
>
next prev parent reply other threads:[~2015-04-22 13:13 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-21 6:25 [Qemu-devel] [PATCH] translate-all: use bitmap helpers for PageDesc's bitmap Emilio G. Cota
2015-04-03 0:08 ` Emilio G. Cota
2015-04-03 14:18 ` Paolo Bonzini
2015-04-22 13:13 ` Paolo Bonzini [this message]
2015-04-22 16:47 ` Emilio G. Cota
2015-04-22 16:53 ` [Qemu-devel] [PATCH v2] " Emilio G. Cota
2015-04-22 20:30 ` Paolo Bonzini
2015-04-22 21:47 ` Emilio G. Cota
2015-04-22 21:50 ` [Qemu-devel] [PATCH v3] " Emilio G. Cota
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=55379E61.5080104@redhat.com \
--to=pbonzini@redhat.com \
--cc=cota@braap.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).