[mirror_edk2.git] / OvmfPkg / Microvm / README

\r
This is an *experimental* port of OVMF for the QEMU microvm\r
machine type.\r
\r
microvm background info\r
-----------------------\r
\r
microvm is designed for modern, virtio-based workloads.  Most legacy\r
lpc/isa devices like pit and pic can be turned off.  virtio-mmio\r
(i.e. '-device virtio-{blk,net,scsi,...}-device') is used for\r
storage/network/etc.\r
\r
Optional pcie support is available and any pcie device supported by\r
QEMU can be plugged in (including virtio-pci if you prefer that over\r
virtio-mmio).\r
\r
https://qemu.readthedocs.io/en/latest/system/i386/microvm.html\r
https://www.kraxel.org/blog/2020/10/qemu-microvm-acpi/\r
\r
design issues\r
-------------\r
\r
Not fully clear yet how to do hardware detection best.  Right now\r
using device tree to find virtio-mmio devices and pcie host bridge,\r
can reuse existing ArmVirtPkg code that way.  Needs patched QEMU.\r
\r
features\r
--------\r
 [working] serial console\r
 [working] direct kernel boot\r
 [working] virtio-mmio support\r
 [working] pcie support\r
\r
known limitations\r
-----------------\r
 * rtc=on is required for now.\r
 * can't use separate code/vars (actually an microvm limitation,\r
   there is no pflash support).\r
 * transitional virtio-pci devices do not work.  microvm doesn't\r
   support ioports on pcie, and ovmf doesn't initialize pcie devices\r
   with ioports if there is no address space for them (even though\r
   pcie devices are required to be functional without ioports).\r
\r
usage\r
-----\r
qemu-system-x86_64 \\r
    -nographic \\r
    -machine microvm,acpi=on,pit=off,pic=off,rtc=on \\r
    -bios /path/to/MICROVM.fd \\r
    [ ... more args here ... ]\r
Commit	Line	Data
e07d27e2 GH	1	\r
	2	This is an experimental port of OVMF for the QEMU microvm\r
	3	machine type.\r
	4	\r
	5	microvm background info\r
	6	-----------------------\r
	7	\r
	8	microvm is designed for modern, virtio-based workloads. Most legacy\r
	9	lpc/isa devices like pit and pic can be turned off. virtio-mmio\r
	10	(i.e. '-device virtio-{blk,net,scsi,...}-device') is used for\r
	11	storage/network/etc.\r
	12	\r
	13	Optional pcie support is available and any pcie device supported by\r
	14	QEMU can be plugged in (including virtio-pci if you prefer that over\r
	15	virtio-mmio).\r
	16	\r
	17	https://qemu.readthedocs.io/en/latest/system/i386/microvm.html\r
	18	https://www.kraxel.org/blog/2020/10/qemu-microvm-acpi/\r
	19	\r
	20	design issues\r
	21	-------------\r
	22	\r
	23	Not fully clear yet how to do hardware detection best. Right now\r
	24	using device tree to find virtio-mmio devices and pcie host bridge,\r
	25	can reuse existing ArmVirtPkg code that way. Needs patched QEMU.\r
	26	\r
	27	features\r
	28	--------\r
	29	[working] serial console\r
	30	[working] direct kernel boot\r
	31	[working] virtio-mmio support\r
632574ce	32	[working] pcie support\r
e07d27e2 GH	33	\r
	34	known limitations\r
	35	-----------------\r
	36	* rtc=on is required for now.\r
	37	* can't use separate code/vars (actually an microvm limitation,\r
	38	there is no pflash support).\r
	39	* transitional virtio-pci devices do not work. microvm doesn't\r
	40	support ioports on pcie, and ovmf doesn't initialize pcie devices\r
	41	with ioports if there is no address space for them (even though\r
	42	pcie devices are required to be functional without ioports).\r
	43	\r
	44	usage\r
	45	-----\r
	46	qemu-system-x86_64 \\r
	47	-nographic \\r
	48	-machine microvm,acpi=on,pit=off,pic=off,rtc=on \\r
	49	-bios /path/to/MICROVM.fd \\r
	50	[ ... more args here ... ]\r