QEMU fw_cfg DMA interface

QEMU fw_cfg DMA interface

[Marc Marí]

Implementation of the FW CFG DMA interface.

When running a Linux guest on top of QEMU, using the -kernel options, this
is the timing improvement for x86:

QEMU commit 16a1b6e and SeaBIOS commit e4d2b8c
QEMU startup time: .080
BIOS startup time: .060
Kernel setup time: .586
Total time: .726

QEMU with this patch series and SeaBIOS with this patch series
QEMU startup time: .080
BIOS startup time: .039
Kernel setup time: .002
Total time: .121

QEMU startup time is the time between the start and the first kvm_entry.
BIOS startup time is the time between the first kvm_entry and the start of
function do_boot, in SeaBIOS.
Kernel setup time is the time between the start of the function do_boot in
SeaBIOS and the jump to the Linux kernel.

As you can see, both the BIOS (because of ACPI tables and other configurations)
and the Linux kernel boot (because of the copy to memory) are greatly
improved with this new interface.

Also, this new interface is an addon to the old interface. Both interfaces
are compatible and interchangeable.

Changes from v1:
 - Take into account order of fields in the FWCfgDmaAccess structure
 - Check and change endianness of FWCfgDmaAccess fields
 - Change order of fields in the FWCfgDmaAccess structure
 - Add FW_CFG_DMA_CTL_SKIP feature for control field
 - Split FW_CFG_SIZE in QEMU
 - Make FW_CFG_ID a bitmap of features
 - Add 64 bit address support for the transfer. Trigger when writing the low
   address, and address is 0 by default and at the end of each transfer.
 - Align ports and addresses.
 - Preserve old fw_cfg_comb_valid behaviour in QEMU
 - Update documentation to reflect all these changes

Changes from v2:
 - Make IOports fw_cfg DMA region a different IO region.
 - Reuse everything for MMIO and IOport DMA regions
 - Make transfer status only based on control field
 - Use DMA helpers instead of direct map/unmap
 - Change ARM fw_cfg DMA address space
 - Change Linux boot process to match linuxboot.S
 - Add select capabilities in the FWCfgDmaAccess struct
 - Update documentation to reflect all these changes

[PATCH v3] QEMU fw_cfg DMA interface documentation

[Marc Marí]

Add fw_cfg DMA interface specfication in the fw_cfg documentation.

Signed-off-by: Marc Marí <markmb@redhat.com>
---
 Documentation/devicetree/bindings/arm/fw-cfg.txt | 49 +++++++++++++++++++++++-
 1 file changed, 48 insertions(+), 1 deletion(-)

diff --git a/Documentation/devicetree/bindings/arm/fw-cfg.txt b/Documentation/devicetree/bindings/arm/fw-cfg.txt
index 953fb64..ba6945e 100644
--- a/Documentation/devicetree/bindings/arm/fw-cfg.txt
+++ b/Documentation/devicetree/bindings/arm/fw-cfg.txt
@@ -45,6 +45,51 @@ blob to be read from the data register has size 4, and it is to be interpreted
 as a uint32_t value in little endian byte order. The current value
 (corresponding to the above outer protocol) is zero.
 
+If bit 1 of the feature bitmap is set, the DMA interface is present. This
+can be used through the 64-bit wide address register.
+
+The address register is in big-endian format. The value for the register is 0
+at startup and after an operation. A write to the lower half triggers an
+operation. This means, that operations with 32-bit addresses can be triggered
+with just one write, whereas operations with 64-bit addresses can be triggered
+with one 64-bit write or two 32-bit writes, starting with the higher part.
+
+In this register, the physical address of a FWCfgDmaAccess structure in RAM
+should be written. This is the format of the FWCfgDmaAccess structure:
+
+typedef struct FWCfgDmaAccess {
+    uint32_t control;
+    uint32_t length;
+    uint64_t address;
+} FWCfgDmaAccess;
+
+The fields of the structure are in big endian mode, and the field at the lowest
+address is the "control" field.
+
+The "control" field has the following bits:
+ - Bit 0: Error
+ - Bit 1: Read
+ - Bit 2: Skip
+ - Bit 3: Select. The upper 16 bits are the selected index.
+
+When an operation is triggered, if the "control" field has bit 3 set, the
+upper 16 bits are interpreted as an index of a firmware configuration item.
+This has the same effect as writing the selector register.
+
+If the "control" field has bit 1 set, a read operation will be performed.
+"length" bytes for the current selector and offset will be copied into the
+physical RAM address specified by the "address" field.
+
+If the "control" field has bit 2 set (and not bit 1), a skip operation will be
+performed. The offset for the current selector will be advanced "length" bytes.
+
+To check result, read the "control" field:
+   error bit set        ->  something went wrong.
+   all bits cleared     ->  transfer finished successfully.
+   otherwise            ->  transfer still in progress (doesn't happen
+                            today due to implementation not being async,
+                            but may in the future).
+
 The guest kernel is not expected to use these registers (although it is
 certainly allowed to); the device tree bindings are documented here because
 this is where device tree bindings reside in general.
@@ -56,6 +101,8 @@ Required properties:
 - reg: the MMIO region used by the device.
   * Bytes 0x0 to 0x7 cover the data register.
   * Bytes 0x8 to 0x9 cover the selector register.
+  * With DMA interface enabled: Bytes 0xc to 0x13 cover the DMA address
+    register.
   * Further registers may be appended to the region in case of future interface
     revisions / feature bits.
 
@@ -66,7 +113,7 @@ Example:
 	#address-cells = <0x2>;
 
 	fw-cfg@9020000 {
+		reg = <0x0 0x9020000 0x0 0x14>;
 		compatible = "qemu,fw-cfg-mmio";
-		reg = <0x0 0x9020000 0x0 0xa>;
 	};
 };
-- 
2.4.3

QEMU fw_cfg DMA interface

[Marc Marí]

Implementation of the FW CFG DMA interface.

When running a Linux guest on top of QEMU, using the -kernel options, this
is the timing improvement for x86:

QEMU commit b2312c6 and SeaBIOS commit 423542e
QEMU startup time: .080
BIOS startup time: .060
Kernel setup time: .586
Total time: .726

QEMU with this patch series and SeaBIOS with this patch series
QEMU startup time: .080
BIOS startup time: .039
Kernel setup time: .005
Total time: .126

QEMU startup time is the time between the start and the first kvm_entry.
BIOS startup time is the time between the first kvm_entry and the start of
function do_boot, in SeaBIOS.
Kernel setup time is the time between the start of the function do_boot in
SeaBIOS and the jump to the Linux kernel.

As you can see, both the BIOS (because of ACPI tables and other configurations)
and the Linux kernel boot (because of the copy to memory) are greatly
improved with this new interface.

Also, this new interface is an addon to the old interface. Both interfaces
are compatible and interchangeable.

Changes from v1:
 - Take into account order of fields in the FWCfgDmaAccess structure
 - Check and change endianness of FWCfgDmaAccess fields
 - Change order of fields in the FWCfgDmaAccess structure
 - Add FW_CFG_DMA_CTL_SKIP feature for control field
 - Split FW_CFG_SIZE in QEMU
 - Make FW_CFG_ID a bitmap of features
 - Add 64 bit address support for the transfer. Trigger when writing the low
   address, and address is 0 by default and at the end of each transfer.
 - Align ports and addresses.
 - Preserve old fw_cfg_comb_valid behaviour in QEMU
 - Update documentation to reflect all these changes

Changes from v2:
 - Make IOports fw_cfg DMA region a different IO region.
 - Reuse everything for MMIO and IOport DMA regions
 - Make transfer status only based on control field
 - Use DMA helpers instead of direct map/unmap
 - Change ARM fw_cfg DMA address space
 - Change Linux boot process to match linuxboot.S
 - Add select capabilities in the FWCfgDmaAccess struct
 - Update documentation to reflect all these changes

Changes from v3:
 - Set properly fw_cfg DMA fields in ARM
 - Set fw_cfg DMA boot process properly (by Laszlo Ersek)
 - Add signature to fw_cfg DMA address field (by Kevin O'Connor)
 - Minor nitpicks

QEMU fw_cfg DMA interface

[Marc Marí]

Implementation of the FW CFG DMA interface.

When running a Linux guest on top of QEMU, using the -kernel options, this
is the timing improvement for x86:

QEMU commit 090d0bf and SeaBIOS commit 2fc20dc
QEMU startup time: .078
BIOS startup time: .060
Kernel setup time: .578
Total time: .716

QEMU with this patch series and SeaBIOS with this patch series
QEMU startup time: .080
BIOS startup time: .039
Kernel setup time: .002
Total time: .121

QEMU startup time is the time between the start and the first kvm_entry.
BIOS startup time is the time between the first kvm_entry and the start of
function do_boot, in SeaBIOS.
Kernel setup time is the time between the start of the function do_boot in
SeaBIOS and the jump to the Linux kernel.

As you can see, both the BIOS (because of ACPI tables and other configurations)
and the Linux kernel boot (because of the copy to memory) are greatly
improved with this new interface.

Also, this new interface is an addon to the old interface. Both interfaces
are compatible and interchangeable.

Changes from v1:
 - Take into account order of fields in the FWCfgDmaAccess structure
 - Check and change endianness of FWCfgDmaAccess fields
 - Change order of fields in the FWCfgDmaAccess structure
 - Add FW_CFG_DMA_CTL_SKIP feature for control field
 - Split FW_CFG_SIZE in QEMU
 - Make FW_CFG_ID a bitmap of features
 - Add 64 bit address support for the transfer. Trigger when writing the low
   address, and address is 0 by default and at the end of each transfer.
 - Align ports and addresses.
 - Preserve old fw_cfg_comb_valid behaviour in QEMU
 - Update documentation to reflect all these changes

Re: QEMU fw_cfg DMA interface

[Marc Marí]

[-- Attachment #1: Type: text/plain, Size: 2332 bytes --]

On Thu, 6 Aug 2015 11:30:43 -0400
"Kevin O'Connor" <kevin@koconnor.net> wrote:

> On Thu, Aug 06, 2015 at 02:37:45PM +0200, Marc Marí wrote:
> > On Thu, 6 Aug 2015 13:27:16 +0100
> > Stefan Hajnoczi <stefanha@gmail.com> wrote:
> > 
> > > On Thu, Aug 6, 2015 at 12:00 PM, Marc Marí <markmb@redhat.com>
> > > wrote:
> > > > When running a Linux guest on top of QEMU, using the -kernel
> > > > options, this is the timing improvement for x86:
> > > >
> > > > QEMU commit 2be4f242b50a8 and SeaBIOS commit 908a58c1d5ff
> > > > QEMU startup time: .078
> > > > BIOS startup time: .060
> > > > Kernel setup time: .578
> > > > Total time: .716
> > > >
> > > > QEMU with this patch series and SeaBIOS with this patch series
> > > > QEMU startup time: .080
> > > > BIOS startup time: .039
> > > > Kernel setup time: .002
> > > > Total time: .121
> > > 
> > > Impressive results!
> > > 
> > > Is this a fully-featured QEMU build or did you disable things?
> > > 
> > > Is this the default SeaBIOS build or did you disable things?
> > > 
> > 
> > This is the default QEMU configuration I get for my system. It's
> > not a fully-featured QEMU, but it has a lot of things enabled.
> > SeaBIOS has a default configuration (with debugging disabled).
> 
> Thanks!
> 
> What qemu command-line did you use during testing?  Also, do you have
> a quick primer on how to use the kvm trace stuff to obtain timings?
> 

The command line I used is:

x86_64-softmmu/qemu-system-x86_64 --enable-kvm \
    -kernel /boot/vmlinuz-4.0.7-300.fc22.x86_64 \
    -L pc-bios/optionrom/ \
    -bios roms/seabios/out/bios.bin -nographic

And I used perf (and two out instructions in the BIOS) to measure the
times:

perf record -a -e kvm:\* -e sched:sched_process_exec

And searching for sched:sched_process_exec, kvm:kvm_entry, pio_write at
0xf5 and pio_write at 0xf4. Out at 0xf5 is the one in "do_boot"
function, and out at 0xf4 is the one just before the jump to the Linux
kernel.

I attach the script I've been using. It can be improved, but it works.
It can be run like this:

sudo ../../results/run_test.sh x86_64-softmmu/qemu-system-x86_64 \
    --enable-kvm -kernel /boot/vmlinuz-4.0.8-300.fc22.x86_64 \
    -L pc-bios/optionrom/ \
    -bios roms/seabios/out/bios.bin -nographic

Thanks
Marc

[-- Attachment #2: run_test.sh --]
[-- Type: application/x-shellscript, Size: 1401 bytes --]

Re: QEMU fw_cfg DMA interface

[Kevin O'Connor]

On Thu, Aug 06, 2015 at 02:37:45PM +0200, Marc Marí wrote:
> On Thu, 6 Aug 2015 13:27:16 +0100
> Stefan Hajnoczi <stefanha@gmail.com> wrote:
> 
> > On Thu, Aug 6, 2015 at 12:00 PM, Marc Marí <markmb@redhat.com> wrote:
> > > When running a Linux guest on top of QEMU, using the -kernel
> > > options, this is the timing improvement for x86:
> > >
> > > QEMU commit 2be4f242b50a8 and SeaBIOS commit 908a58c1d5ff
> > > QEMU startup time: .078
> > > BIOS startup time: .060
> > > Kernel setup time: .578
> > > Total time: .716
> > >
> > > QEMU with this patch series and SeaBIOS with this patch series
> > > QEMU startup time: .080
> > > BIOS startup time: .039
> > > Kernel setup time: .002
> > > Total time: .121
> > 
> > Impressive results!
> > 
> > Is this a fully-featured QEMU build or did you disable things?
> > 
> > Is this the default SeaBIOS build or did you disable things?
> > 
> 
> This is the default QEMU configuration I get for my system. It's not a
> fully-featured QEMU, but it has a lot of things enabled. SeaBIOS
> has a default configuration (with debugging disabled).

Thanks!

What qemu command-line did you use during testing?  Also, do you have
a quick primer on how to use the kvm trace stuff to obtain timings?

-Kevin

Re: QEMU fw_cfg DMA interface

[Stefan Hajnoczi]

On Thu, Aug 6, 2015 at 1:37 PM, Marc Marí <markmb@redhat.com> wrote:
> On Thu, 6 Aug 2015 13:27:16 +0100
> Stefan Hajnoczi <stefanha@gmail.com> wrote:
>
>> On Thu, Aug 6, 2015 at 12:00 PM, Marc Marí <markmb@redhat.com> wrote:
>> > When running a Linux guest on top of QEMU, using the -kernel
>> > options, this is the timing improvement for x86:
>> >
>> > QEMU commit 2be4f242b50a8 and SeaBIOS commit 908a58c1d5ff
>> > QEMU startup time: .078
>> > BIOS startup time: .060
>> > Kernel setup time: .578
>> > Total time: .716
>> >
>> > QEMU with this patch series and SeaBIOS with this patch series
>> > QEMU startup time: .080
>> > BIOS startup time: .039
>> > Kernel setup time: .002
>> > Total time: .121
>>
>> Impressive results!
>>
>> Is this a fully-featured QEMU build or did you disable things?
>>
>> Is this the default SeaBIOS build or did you disable things?
>>
>
> This is the default QEMU configuration I get for my system. It's not a
> fully-featured QEMU, but it has a lot of things enabled. SeaBIOS
> has a default configuration (with debugging disabled).

That's great.

Since SeaBIOS is a default configuration, the remaining BIOS startup
time is amenable to more optimizations in the future.

Stefan

Re: QEMU fw_cfg DMA interface

[Marc Marí]

On Thu, 6 Aug 2015 13:27:16 +0100
Stefan Hajnoczi <stefanha@gmail.com> wrote:

> On Thu, Aug 6, 2015 at 12:00 PM, Marc Marí <markmb@redhat.com> wrote:
> > When running a Linux guest on top of QEMU, using the -kernel
> > options, this is the timing improvement for x86:
> >
> > QEMU commit 2be4f242b50a8 and SeaBIOS commit 908a58c1d5ff
> > QEMU startup time: .078
> > BIOS startup time: .060
> > Kernel setup time: .578
> > Total time: .716
> >
> > QEMU with this patch series and SeaBIOS with this patch series
> > QEMU startup time: .080
> > BIOS startup time: .039
> > Kernel setup time: .002
> > Total time: .121
> 
> Impressive results!
> 
> Is this a fully-featured QEMU build or did you disable things?
> 
> Is this the default SeaBIOS build or did you disable things?
> 

This is the default QEMU configuration I get for my system. It's not a
fully-featured QEMU, but it has a lot of things enabled. SeaBIOS
has a default configuration (with debugging disabled).

The QEMU configuration is:
[...]
tcg debug enabled no
gprof enabled     no
sparse enabled    no
strip binaries    yes
profiler          no
static build      no
pixman            system
SDL support       yes
GTK support       yes
GNUTLS support    yes
GNUTLS hash       yes
GNUTLS gcrypt     no
GNUTLS nettle     yes (2.7.1)
VTE support       yes
curses support    yes
curl support      yes
mingw32 support   no
Audio drivers     oss
Block whitelist (rw)
Block whitelist (ro)
VirtFS support    yes
VNC support       yes
VNC TLS support   yes
VNC SASL support  yes
VNC JPEG support  yes
VNC PNG support   yes
xen support       no
brlapi support    yes
bluez  support    yes
Documentation     no
GUEST_BASE        yes
PIE               yes
vde support       yes
netmap support    no
Linux AIO support yes
ATTR/XATTR support yes
Install blobs     yes
KVM support       yes
RDMA support      yes
TCG interpreter   no
fdt support       yes
preadv support    yes
fdatasync         yes
madvise           yes
posix_madvise     yes
sigev_thread_id   yes
uuid support      yes
libcap-ng support yes
vhost-net support yes
vhost-scsi support yes
Trace backends    nop
spice support     yes (0.12.7/0.12.5)
rbd support       yes
xfsctl support    no
nss used          yes
libusb            yes
usb net redir     yes
OpenGL support    no
libiscsi support  yes
libnfs support    no
build guest agent yes
QGA VSS support   no
QGA w32 disk info no
seccomp support   yes
coroutine backend ucontext
coroutine pool    yes
GlusterFS support yes
Archipelago support no
gcov              gcov
gcov enabled      no
TPM support       yes
libssh2 support   yes
TPM passthrough   yes
QOM debugging     yes
vhdx              yes
lzo support       yes
snappy support    yes
bzip2 support     yes
NUMA host support yes
tcmalloc support  no

Marc

Re: QEMU fw_cfg DMA interface

[Stefan Hajnoczi]

On Thu, Aug 6, 2015 at 12:00 PM, Marc Marí <markmb@redhat.com> wrote:
> When running a Linux guest on top of QEMU, using the -kernel options, this
> is the timing improvement for x86:
>
> QEMU commit 2be4f242b50a8 and SeaBIOS commit 908a58c1d5ff
> QEMU startup time: .078
> BIOS startup time: .060
> Kernel setup time: .578
> Total time: .716
>
> QEMU with this patch series and SeaBIOS with this patch series
> QEMU startup time: .080
> BIOS startup time: .039
> Kernel setup time: .002
> Total time: .121

Impressive results!

Is this a fully-featured QEMU build or did you disable things?

Is this the default SeaBIOS build or did you disable things?

Stefan

QEMU fw_cfg DMA interface

[Marc Marí]

Implementation of the FW CFG DMA interface.

When running a Linux guest on top of QEMU, using the -kernel options, this
is the timing improvement for x86:

QEMU commit 2be4f242b50a8 and SeaBIOS commit 908a58c1d5ff
QEMU startup time: .078
BIOS startup time: .060
Kernel setup time: .578
Total time: .716

QEMU with this patch series and SeaBIOS with this patch series
QEMU startup time: .080
BIOS startup time: .039
Kernel setup time: .002
Total time: .121

QEMU startup time is the time between the start and the first kvm_entry.
BIOS startup time is the time between the first kvm_entry and the start of
function do_boot, in SeaBIOS.
Kernel setup time is the time between the start of the function do_boot in
SeaBIOS and the jump to the Linux kernel.

As you can see, both the BIOS (because of ACPI tables and other configurations)
and the Linux kernel boot (because of the copy to memory) are greatly
improved with this new interface.

Also, this new interface is an addon to the old interface. Both interfaces
are compatible and interchangeable.