From: Andre Przywara <andre.przywara@arm.com>
To: Alexandru Elisei <alexandru.elisei@arm.com>
Cc: Will Deacon <will@kernel.org>,
Julien Thierry <julien.thierry.kdev@gmail.com>,
kvm@vger.kernel.org, kvmarm@lists.cs.columbia.edu,
Marc Zyngier <maz@kernel.org>,
Sami Mujawar <sami.mujawar@arm.com>
Subject: Re: [PATCH kvmtool v2 20/22] arm: Reorganise and document memory map
Date: Mon, 15 Mar 2021 12:27:02 +0000 [thread overview]
Message-ID: <20210315122702.5f5c629f@slackpad.fritz.box> (raw)
In-Reply-To: <deb3e029-e634-c28c-2a9a-e461041bb249@arm.com>
On Tue, 9 Mar 2021 15:46:29 +0000
Alexandru Elisei <alexandru.elisei@arm.com> wrote:
Hi,
> Hi Andre,
>
> This is a really good idea, thank you for implementing it!
>
> Some comments below.
>
> On 2/25/21 12:59 AM, Andre Przywara wrote:
> > The hardcoded memory map we expose to a guest is currently described
> > using a series of partially interconnected preprocessor constants,
> > which is hard to read and follow.
> >
> > In preparation for moving the UART and RTC to some different MMIO
> > region, document the current map with some ASCII art, and clean up the
> > definition of the sections.
> >
> > No functional change.
> >
> > Signed-off-by: Andre Przywara <andre.przywara@arm.com>
> > ---
> > arm/include/arm-common/kvm-arch.h | 41 ++++++++++++++++++++++---------
> > 1 file changed, 29 insertions(+), 12 deletions(-)
> >
> > diff --git a/arm/include/arm-common/kvm-arch.h b/arm/include/arm-common/kvm-arch.h
> > index d84e50cd..b12255b0 100644
> > --- a/arm/include/arm-common/kvm-arch.h
> > +++ b/arm/include/arm-common/kvm-arch.h
> > @@ -7,14 +7,33 @@
> >
> > #include "arm-common/gic.h"
> >
> > +/*
> > + * The memory map used for ARM guests (not to scale):
> > + *
> > + * 0 64K 16M 32M 48M 1GB 2GB
> > + * +-------+-..-+-------+-------+--....--+-----+--.....--+---......
> > + * | (PCI) |////| int. | | | | |
> > + * | I/O |////| MMIO: | Flash | virtio | GIC | PCI | DRAM
> > + * | ports |////| UART, | | MMIO | | (AXI) |
> > + * | |////| RTC | | | | |
> > + * +-------+-..-+-------+-------+--....--+-----+--.....--+---......
> > + */
>
> Nitpick: I searched the PCI Local Bus Specification revision 3.0 (which kvmtool
> currently implements) for the term I/O ports, and found one mention in a schematic
> for an add-in card. The I/O region is called in the spec I/O Space.
Right, will change that.
> I don't know what "int." means in the region for the UART and RTC.
It was meant to mean "internal", but this is really nonsensical, so I
will replace it with "plat MMIO".
> The comment says that the art is not to scale, so I don't think there's any need
> for the "..." between the corners of the regions. To my eyes, it makes the ASCII
> art look crooked.
fixed.
> The next patches add the UART and RTC outside the first 64K, I think the region
> should be documented in the patches where the changes are made, not here. Another
> alternative would be to move this patch to the end of the series instead of
> incrementally changing the memory ASCII art (which I imagine is time consuming).
You passed the scrutiny test ;-)
I noticed this last minute, but didn't feel like changing it then.
Have done it now.
>
> Otherwise, the numbers look OK.
>
> > +
> > #define ARM_IOPORT_AREA _AC(0x0000000000000000, UL)
> > -#define ARM_FLASH_AREA _AC(0x0000000002000000, UL)
> > -#define ARM_MMIO_AREA _AC(0x0000000003000000, UL)
> > +#define ARM_MMIO_AREA _AC(0x0000000001000000, UL)
>
> The patch says it is *documenting* the memory layout, but here it is *changing*
> the layout. Other than that, I like the shuffling of definitions so the kvmtool
> global defines are closer to the arch values.
I amended the commit message to mention that I change the value of
ARM_MMIO_AREA, but that stays internal to that file and doesn't affect
the other definitions.
In fact I imported this header into a C file and printed all
externally used names, before and after: it didn't show any changes.
Cheers,
Andre
> > #define ARM_AXI_AREA _AC(0x0000000040000000, UL)
> > #define ARM_MEMORY_AREA _AC(0x0000000080000000, UL)
> >
> > -#define ARM_LOMAP_MAX_MEMORY ((1ULL << 32) - ARM_MEMORY_AREA)
> > -#define ARM_HIMAP_MAX_MEMORY ((1ULL << 40) - ARM_MEMORY_AREA)
> > +#define KVM_IOPORT_AREA ARM_IOPORT_AREA
> > +#define ARM_IOPORT_SIZE (1U << 16)
> > +
> > +
> > +#define KVM_FLASH_MMIO_BASE (ARM_MMIO_AREA + 0x1000000)
> > +#define KVM_FLASH_MAX_SIZE 0x1000000
> > +
> > +#define KVM_VIRTIO_MMIO_AREA (KVM_FLASH_MMIO_BASE + KVM_FLASH_MAX_SIZE)
> > +#define ARM_VIRTIO_MMIO_SIZE (ARM_AXI_AREA - \
> > + (KVM_VIRTIO_MMIO_AREA + ARM_GIC_SIZE))
> >
> > #define ARM_GIC_DIST_BASE (ARM_AXI_AREA - ARM_GIC_DIST_SIZE)
> > #define ARM_GIC_CPUI_BASE (ARM_GIC_DIST_BASE - ARM_GIC_CPUI_SIZE)
> > @@ -22,19 +41,17 @@
> > #define ARM_GIC_DIST_SIZE 0x10000
> > #define ARM_GIC_CPUI_SIZE 0x20000
> >
> > -#define KVM_FLASH_MMIO_BASE ARM_FLASH_AREA
> > -#define KVM_FLASH_MAX_SIZE (ARM_MMIO_AREA - ARM_FLASH_AREA)
> >
> > -#define ARM_IOPORT_SIZE (1U << 16)
> > -#define ARM_VIRTIO_MMIO_SIZE (ARM_AXI_AREA - (ARM_MMIO_AREA + ARM_GIC_SIZE))
> > +#define KVM_PCI_CFG_AREA ARM_AXI_AREA
> > #define ARM_PCI_CFG_SIZE (1ULL << 24)
> > +#define KVM_PCI_MMIO_AREA (KVM_PCI_CFG_AREA + ARM_PCI_CFG_SIZE)
> > #define ARM_PCI_MMIO_SIZE (ARM_MEMORY_AREA - \
> > (ARM_AXI_AREA + ARM_PCI_CFG_SIZE))
> >
> > -#define KVM_IOPORT_AREA ARM_IOPORT_AREA
> > -#define KVM_PCI_CFG_AREA ARM_AXI_AREA
> > -#define KVM_PCI_MMIO_AREA (KVM_PCI_CFG_AREA + ARM_PCI_CFG_SIZE)
> > -#define KVM_VIRTIO_MMIO_AREA ARM_MMIO_AREA
> > +
> > +#define ARM_LOMAP_MAX_MEMORY ((1ULL << 32) - ARM_MEMORY_AREA)
> > +#define ARM_HIMAP_MAX_MEMORY ((1ULL << 40) - ARM_MEMORY_AREA)
> > +
> >
> > #define KVM_IOEVENTFD_HAS_PIO 0
> >
next prev parent reply other threads:[~2021-03-15 12:28 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-02-25 0:58 [PATCH kvmtool v2 00/22] Unify I/O port and MMIO trap handling Andre Przywara
2021-02-25 0:58 ` [PATCH kvmtool v2 01/22] ioport: Remove ioport__setup_arch() Andre Przywara
2021-03-02 18:02 ` Alexandru Elisei
2021-02-25 0:58 ` [PATCH kvmtool v2 02/22] hw/serial: Use device abstraction for FDT generator function Andre Przywara
2021-02-25 0:58 ` [PATCH kvmtool v2 03/22] ioport: Retire .generate_fdt_node functionality Andre Przywara
2021-02-25 0:58 ` [PATCH kvmtool v2 04/22] mmio: Extend handling to include ioport emulation Andre Przywara
2021-03-03 17:58 ` Alexandru Elisei
2021-03-15 12:24 ` Andre Przywara
2021-02-25 0:58 ` [PATCH kvmtool v2 05/22] hw/i8042: Clean up data types Andre Przywara
2021-02-25 0:58 ` [PATCH kvmtool v2 06/22] hw/i8042: Refactor trap handler Andre Przywara
2021-03-08 17:16 ` Alexandru Elisei
2021-02-25 0:59 ` [PATCH kvmtool v2 07/22] hw/i8042: Switch to new trap handlers Andre Przywara
2021-03-08 17:22 ` Alexandru Elisei
2021-02-25 0:59 ` [PATCH kvmtool v2 08/22] x86/ioport: Refactor " Andre Przywara
2021-03-09 11:49 ` Alexandru Elisei
2021-03-15 12:26 ` Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 09/22] x86/ioport: Switch to new " Andre Przywara
2021-03-09 12:09 ` Alexandru Elisei
2021-03-15 12:26 ` Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 10/22] hw/rtc: Refactor " Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 11/22] hw/rtc: Switch to new trap handler Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 12/22] hw/vesa: Switch trap handling to use MMIO handler Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 13/22] hw/serial: Refactor trap handler Andre Przywara
2021-03-12 11:29 ` Alexandru Elisei
2021-03-15 12:26 ` Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 14/22] hw/serial: Switch to new trap handlers Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 15/22] vfio: Refactor ioport trap handler Andre Przywara
2021-03-09 14:21 ` Alexandru Elisei
2021-02-25 0:59 ` [PATCH kvmtool v2 16/22] vfio: Switch to new ioport trap handlers Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 17/22] virtio: Switch trap handling to use MMIO handler Andre Przywara
2021-03-09 14:43 ` Alexandru Elisei
2021-02-25 0:59 ` [PATCH kvmtool v2 18/22] pci: " Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 19/22] Remove ioport specific routines Andre Przywara
2021-03-09 15:13 ` Alexandru Elisei
2021-02-25 0:59 ` [PATCH kvmtool v2 20/22] arm: Reorganise and document memory map Andre Przywara
2021-03-09 15:46 ` Alexandru Elisei
2021-03-15 12:27 ` Andre Przywara [this message]
2021-02-25 0:59 ` [PATCH kvmtool v2 21/22] hw/serial: ARM/arm64: Use MMIO at higher addresses Andre Przywara
2021-03-09 16:02 ` Alexandru Elisei
2021-03-15 12:27 ` Andre Przywara
2021-02-25 0:59 ` [PATCH kvmtool v2 22/22] hw/rtc: " Andre Przywara
2021-03-09 16:06 ` Alexandru Elisei
2021-03-02 17:57 ` [PATCH kvmtool v2 00/22] Unify I/O port and MMIO trap handling Alexandru Elisei
2021-03-03 18:41 ` Andre Przywara
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=20210315122702.5f5c629f@slackpad.fritz.box \
--to=andre.przywara@arm.com \
--cc=alexandru.elisei@arm.com \
--cc=julien.thierry.kdev@gmail.com \
--cc=kvm@vger.kernel.org \
--cc=kvmarm@lists.cs.columbia.edu \
--cc=maz@kernel.org \
--cc=sami.mujawar@arm.com \
--cc=will@kernel.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