From: Aurelien Jarno <aurelien@aurel32.net>
To: James Hogan <james.hogan@imgtec.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>,
Gleb Natapov <gleb@redhat.com>,
qemu-devel@nongnu.org, kvm@vger.kernel.org,
Sanjay Lal <sanjayl@kymasys.com>
Subject: Re: [Qemu-devel] [PATCH v2 06/10] target-mips: Set target page size to 16K in KVM mode
Date: Sat, 21 Dec 2013 16:34:22 +0100 [thread overview]
Message-ID: <20131221153422.GA9265@ohm.rr44.fr> (raw)
In-Reply-To: <1387203165-5553-7-git-send-email-james.hogan@imgtec.com>
On Mon, Dec 16, 2013 at 02:12:41PM +0000, James Hogan wrote:
> From: Sanjay Lal <sanjayl@kymasys.com>
>
> With larger set associative caches KVM can open the possibility of cache
> aliasing between the memory that QEMU allocates with mmap and the
> mapping into the guest address space. Therefore increase the target page
> size to 16K when KVM is configured.
>
> Signed-off-by: Sanjay Lal <sanjayl@kymasys.com>
> Signed-off-by: James Hogan <james.hogan@imgtec.com>
> Cc: Aurelien Jarno <aurelien@aurel32.net>
> ---
> Changes in v2:
> - Expand commit message
> ---
> target-mips/mips-defs.h | 5 +++++
> 1 file changed, 5 insertions(+)
>
> diff --git a/target-mips/mips-defs.h b/target-mips/mips-defs.h
> index bf094a3..473ddf8 100644
> --- a/target-mips/mips-defs.h
> +++ b/target-mips/mips-defs.h
> @@ -5,7 +5,12 @@
> //#define USE_HOST_FLOAT_REGS
>
> /* Real pages are variable size... */
> +#ifdef CONFIG_KVM
> +/* For KVM/MIPS the minimum page size is 16K due to cache aliasing issues */
> +#define TARGET_PAGE_BITS 14
> +#else
> #define TARGET_PAGE_BITS 12
> +#endif
> #define MIPS_TLB_MAX 128
>
> #if defined(TARGET_MIPS64)
I am not sure it is something correct. It means that the emulated CPU
won't support 4K pages anymore, even when running in TCG mode without
KVM. If the kernel maps a 4K page, in practice a 16K page is going to be
mapped by the qemu tlb.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
next prev parent reply other threads:[~2013-12-21 15:34 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-12-16 14:12 [Qemu-devel] [PATCH v2 00/10] KVM Support for MIPS32 Processors James Hogan
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 01/10] hw/mips/cputimer: Don't start periodic timer in KVM mode James Hogan
2013-12-21 15:34 ` Aurelien Jarno
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 02/10] hw/mips: Add API to convert KVM guest KSEG0 <-> GPA James Hogan
2013-12-21 15:34 ` Aurelien Jarno
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 03/10] target-mips: get_physical_address: Add defines for segment bases James Hogan
2013-12-21 15:34 ` Aurelien Jarno
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 04/10] target-mips: get_physical_address: Add KVM awareness James Hogan
2013-12-21 15:34 ` Aurelien Jarno
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 05/10] kvm: Set sigmask length to 16 for MIPS targets James Hogan
2013-12-21 15:34 ` Aurelien Jarno
2013-12-21 20:18 ` Peter Maydell
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 06/10] target-mips: Set target page size to 16K in KVM mode James Hogan
2013-12-21 15:34 ` Aurelien Jarno [this message]
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 07/10] target-mips: kvm: Add main KVM support for MIPS James Hogan
2013-12-21 15:35 ` Aurelien Jarno
2014-02-10 14:07 ` Andreas Färber
2014-02-11 10:54 ` James Hogan
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 08/10] hw/mips: In KVM mode, inject IRQ2 (I/O) interupts via ioctls James Hogan
2013-12-21 15:41 ` Aurelien Jarno
2013-12-21 20:40 ` Peter Maydell
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 09/10] hw/mips: malta: Add KVM support James Hogan
2013-12-21 15:37 ` Aurelien Jarno
2013-12-21 20:42 ` Peter Maydell
2014-02-12 15:07 ` James Hogan
2013-12-16 14:12 ` [Qemu-devel] [PATCH v2 10/10] target-mips: Enable KVM support in build system James Hogan
2013-12-21 15:38 ` Aurelien Jarno
2013-12-21 18:59 ` Peter Maydell
2013-12-23 12:31 ` James Hogan
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=20131221153422.GA9265@ohm.rr44.fr \
--to=aurelien@aurel32.net \
--cc=gleb@redhat.com \
--cc=james.hogan@imgtec.com \
--cc=kvm@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=sanjayl@kymasys.com \
/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).