QEMU CPU Emulator Reference Documentation
QEMU is a FAST! processor emulator. By using dynamic translation it achieves a reasonnable speed while being easy to port on new host CPUs.
QEMU has two operating modes:
As QEMU requires no host kernel patches to run, it is very safe and easy to use.
QEMU generic features:
libqemu) which can be used
in other projects.
QEMU user mode emulation features:
QEMU x86 target features:
Current QEMU limitations:
doubles are used instead of the non standard
10 byte long doubles of x86 for floating point emulation to get
maximum performances.
If you need to compile QEMU, please read the `README' which gives the related information.
In order to launch a Linux process, QEMU needs the process executable itself and all the target (x86) dynamic libraries used by it.
qemu -L / /bin/ls
-L / tells that the x86 dynamic linker must be searched with a
`/' prefix.
qemu -L / qemu -L / /bin/ls
LD_LIBRARY_PATH is not set:
unset LD_LIBRARY_PATHThen you can launch the precompiled `ls' x86 executable:
qemu /usr/local/qemu-i386/bin/ls-i386You can look at `/usr/local/qemu-i386/bin/qemu-conf.sh' so that QEMU is automatically launched by the Linux kernel when you try to launch x86 executables. It requires the
binfmt_misc module in the
Linux kernel.
qemu /usr/local/qemu-i386/bin/qemu-i386 /usr/local/qemu-i386/bin/ls-i386
qemu /usr/local/qemu-i386/bin/ls-i386
${HOME}/.wine directory is saved to ${HOME}/.wine.org.
qemu /usr/local/qemu-i386/wine/bin/wine /usr/local/qemu-i386/wine/c/Program\ Files/putty.exe
usage: qemu [-h] [-d] [-L path] [-s size] program [arguments...]
Debug options:
This section explains how to launch a Linux kernel inside QEMU.
sudo so that the command ifconfig contained in
`vl-ifup' can be executed as root. You must verify that your host
kernel supports the TUN/TAP network interfaces: the device
`/dev/net/tun' must be present.
When network is enabled, there is a virtual network connection between
the host kernel and the emulated kernel. The emulated kernel is seen
from the host kernel at IP address 172.20.0.2 and the host kernel is
seen from the emulated kernel at IP address 172.20.0.1.
vl.sh. You should have the following output:
> ./vl.sh connected to host network interface: tun0 Uncompressing Linux... Ok, booting the kernel. Linux version 2.4.20 (bellard@voyager) (gcc version 2.95.2 20000220 (Debian GNU/Linux)) #42 Wed Jun 25 14:16:12 CEST 2003 BIOS-provided physical RAM map: BIOS-88: 0000000000000000 - 000000000009f000 (usable) BIOS-88: 0000000000100000 - 0000000002000000 (usable) 32MB LOWMEM available. On node 0 totalpages: 8192 zone(0): 4096 pages. zone(1): 4096 pages. zone(2): 0 pages. Kernel command line: root=/dev/ram ramdisk_size=6144 Initializing CPU#0 Detected 501.785 MHz processor. Calibrating delay loop... 973.20 BogoMIPS Memory: 24776k/32768k available (725k kernel code, 7604k reserved, 151k data, 48k init, 0k highmem) Dentry cache hash table entries: 4096 (order: 3, 32768 bytes) Inode cache hash table entries: 2048 (order: 2, 16384 bytes) Mount-cache hash table entries: 512 (order: 0, 4096 bytes) Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) Page-cache hash table entries: 8192 (order: 3, 32768 bytes) CPU: Intel Pentium Pro stepping 03 Checking 'hlt' instruction... OK. POSIX conformance testing by UNIFIX Linux NET4.0 for Linux 2.4 Based upon Swansea University Computer Society NET3.039 Initializing RT netlink socket apm: BIOS not found. Starting kswapd pty: 256 Unix98 ptys configured Serial driver version 5.05c (2001-07-08) with no serial options enabled ttyS00 at 0x03f8 (irq = 4) is a 16450 ne.c:v1.10 9/23/94 Donald Becker (becker@scyld.com) Last modified Nov 1, 2000 by Paul Gortmaker NE*000 ethercard probe at 0x300: 52 54 00 12 34 56 eth0: NE2000 found at 0x300, using IRQ 9. RAMDISK driver initialized: 16 RAM disks of 6144K size 1024 blocksize NET4: Linux TCP/IP 1.0 for NET4.0 IP Protocols: ICMP, UDP, TCP, IGMP IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 2048 bind 2048) NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. RAMDISK: ext2 filesystem found at block 0 RAMDISK: Loading 6144 blocks [1 disk] into ram disk... done. Freeing initrd memory: 6144k freed VFS: Mounted root (ext2 filesystem). Freeing unused kernel memory: 48k freed sh: can't access tty; job control turned off #
ls for example. Type Ctrl-a h to have an help
about the keys you can type inside the virtual serial console. In
particular, use Ctrl-a x to exit QEMU and use Ctrl-a b as
the Magic SysRq key.
. /etc/linuxrcThen enable X11 connections on your PC from the emulated Linux:
xhost +172.20.0.2You can now launch `xterm' or `xlogo' and verify that you have a real Virtual Linux system !
NOTES:
You can use any Linux kernel within QEMU provided it is mapped at address 0x90000000 (the default is 0xc0000000). You must modify only two lines in the kernel source:
In asm/page.h, replace
#define __PAGE_OFFSET (0xc0000000)
by
#define __PAGE_OFFSET (0x90000000)
And in arch/i386/vmlinux.lds, replace
. = 0xc0000000 + 0x100000;
by
. = 0x90000000 + 0x100000;
The file config-2.4.20 gives the configuration of the example kernel.
Just type
make bzImage
As you would do to make a real kernel. Then you can use with QEMU exactly the same kernel as you would boot on your PC (in `arch/i386/boot/bzImage').
If you are not using a 2.5 kernel as host kernel but if you use a target 2.5 kernel, you must also ensure that the 'HZ' define is set to 100 (1000 is the default) as QEMU cannot currently emulate timers at frequencies greater than 100 Hz on host Linux systems < 2.5. In asm/param.h, replace:
# define HZ 1000 /* Internal kernel timer frequency */
by
# define HZ 100 /* Internal kernel timer frequency */
QEMU emulates the following PC peripherials:
Uncompressing Linux message)
Like bochs [3], QEMU emulates an x86 CPU. But QEMU is much faster than bochs as it uses dynamic compilation and because it uses the host MMU to simulate the x86 MMU. The downside is that currently the emulation is not as accurate as bochs (for example, you cannot currently run Windows inside QEMU).
Like Valgrind [2], QEMU does user space emulation and dynamic translation. Valgrind is mainly a memory debugger while QEMU has no support for it (QEMU could be used to detect out of bound memory accesses as Valgrind, but it has no support to track uninitialised data as Valgrind does). The Valgrind dynamic translator generates better code than QEMU (in particular it does register allocation) but it is closely tied to an x86 host and target and has no support for precise exceptions and system emulation.
EM86 [4] is the closest project to user space QEMU (and QEMU still uses some of its code, in particular the ELF file loader). EM86 was limited to an alpha host and used a proprietary and slow interpreter (the interpreter part of the FX!32 Digital Win32 code translator [5]).
TWIN [6] is a Windows API emulator like Wine. It is less accurate than Wine but includes a protected mode x86 interpreter to launch x86 Windows executables. Such an approach as greater potential because most of the Windows API is executed natively but it is far more difficult to develop because all the data structures and function parameters exchanged between the API and the x86 code must be converted.
User mode Linux [7] was the only solution before QEMU to launch a Linux kernel as a process while not needing any host kernel patches. However, user mode Linux requires heavy kernel patches while QEMU accepts unpatched Linux kernels. It would be interesting to compare the performance of the two approaches.
The new Plex86 [8] PC virtualizer is done in the same spirit as the QEMU system emulator. It requires a patched Linux kernel to work (you cannot launch the same kernel on your PC), but the patches are really small. As it is a PC virtualizer (no emulation is done except for some priveledged instructions), it has the potential of being faster than QEMU. The downside is that a complicated (and potentially unsafe) host kernel patch is needed.
QEMU is a dynamic translator. When it first encounters a piece of code, it converts it to the host instruction set. Usually dynamic translators are very complicated and highly CPU dependent. QEMU uses some tricks which make it relatively easily portable and simple while achieving good performances.
The basic idea is to split every x86 instruction into fewer simpler instructions. Each simple instruction is implemented by a piece of C code (see `op-i386.c'). Then a compile time tool (`dyngen') takes the corresponding object file (`op-i386.o') to generate a dynamic code generator which concatenates the simple instructions to build a function (see `op-i386.h:dyngen_code()').
In essence, the process is similar to [1], but more work is done at compile time.
A key idea to get optimal performances is that constant parameters can be passed to the simple operations. For that purpose, dummy ELF relocations are generated with gcc for each constant parameter. Then, the tool (`dyngen') can locate the relocations and generate the appriopriate C code to resolve them when building the dynamic code.
That way, QEMU is no more difficult to port than a dynamic linker.
To go even faster, GCC static register variables are used to keep the state of the virtual CPU.
Since QEMU uses fixed simple instructions, no efficient register allocation can be done. However, because RISC CPUs have a lot of register, most of the virtual CPU state can be put in registers without doing complicated register allocation.
Good CPU condition codes emulation (EFLAGS register on x86) is a
critical point to get good performances. QEMU uses lazy condition code
evaluation: instead of computing the condition codes after each x86
instruction, it just stores one operand (called CC_SRC), the
result (called CC_DST) and the type of operation (called
CC_OP).
CC_OP is almost never explicitely set in the generated code
because it is known at translation time.
In order to increase performances, a backward pass is performed on the
generated simple instructions (see
translate-i386.c:optimize_flags()). When it can be proved that
the condition codes are not needed by the next instructions, no
condition codes are computed at all.
The x86 CPU has many internal states which change the way it evaluates instructions. In order to achieve a good speed, the translation phase considers that some state information of the virtual x86 CPU cannot change in it. For example, if the SS, DS and ES segments have a zero base, then the translator does not even generate an addition for the segment base.
[The FPU stack pointer register is not handled that way yet].
A 2MByte cache holds the most recently used translations. For simplicity, it is completely flushed when it is full. A translation unit contains just a single basic block (a block of x86 instructions terminated by a jump or by a virtual CPU state change which the translator cannot deduce statically).
After each translated basic block is executed, QEMU uses the simulated Program Counter (PC) and other cpu state informations (such as the CS segment base value) to find the next basic block.
In order to accelerate the most common cases where the new simulated PC is known, QEMU can patch a basic block so that it jumps directly to the next one.
The most portable code uses an indirect jump. An indirect jump makes it
easier to make the jump target modification atomic. On some
architectures (such as PowerPC), the JUMP opcode is directly
patched so that the block chaining has no overhead.
Self-modifying code is a special challenge in x86 emulation because no instruction cache invalidation is signaled by the application when code is modified.
When translated code is generated for a basic block, the corresponding
host page is write protected if it is not already read-only (with the
system call mprotect()). Then, if a write access is done to the
page, Linux raises a SEGV signal. QEMU then invalidates all the
translated code in the page and enables write accesses to the page.
Correct translated code invalidation is done efficiently by maintaining a linked list of every translated block contained in a given page. Other linked lists are also maintained to undo direct block chaining.
Althought the overhead of doing mprotect() calls is important,
most MSDOS programs can be emulated at reasonnable speed with QEMU and
DOSEMU.
Note that QEMU also invalidates pages of translated code when it detects
that memory mappings are modified with mmap() or munmap().
longjmp() is used when an exception such as division by zero is encountered.
The host SIGSEGV and SIGBUS signal handlers are used to get invalid memory accesses. The exact CPU state can be retrieved because all the x86 registers are stored in fixed host registers. The simulated program counter is found by retranslating the corresponding basic block and by looking where the host program counter was at the exception point.
The virtual CPU cannot retrieve the exact EFLAGS register because
in some cases it is not computed because of condition code
optimisations. It is not a big concern because the emulated code can
still be restarted in any cases.
QEMU includes a generic system call translator for Linux. It means that the parameters of the system calls can be converted to fix the endianness and 32/64 bit issues. The IOCTLs are converted with a generic type description system (see `ioctls.h' and `thunk.c').
QEMU supports host CPUs which have pages bigger than 4KB. It records all
the mappings the process does and try to emulated the mmap()
system calls in cases where the host mmap() call would fail
because of bad page alignment.
Normal and real-time signals are queued along with their information
(siginfo_t) as it is done in the Linux kernel. Then an interrupt
request is done to the virtual CPU. When it is interrupted, one queued
signal is handled by generating a stack frame in the virtual CPU as the
Linux kernel does. The sigreturn() system call is emulated to return
from the virtual signal handler.
Some signals (such as SIGALRM) directly come from the host. Other
signals are synthetized from the virtual CPU exceptions such as SIGFPE
when a division by zero is done (see main.c:cpu_loop()).
The blocked signal mask is still handled by the host Linux kernel so
that most signal system calls can be redirected directly to the host
Linux kernel. Only the sigaction() and sigreturn() system
calls need to be fully emulated (see `signal.c').
The Linux clone() system call is usually used to create a thread. QEMU uses the host clone() system call so that real host threads are created for each emulated thread. One virtual CPU instance is created for each thread.
The virtual x86 CPU atomic operations are emulated with a global lock so that their semantic is preserved.
Note that currently there are still some locking issues in QEMU. In particular, the translated cache flush is not protected yet against reentrancy.
QEMU was conceived so that ultimately it can emulate itself. Althought it is not very useful, it is an important test to show the power of the emulator.
Achieving self-virtualization is not easy because there may be address space conflicts. QEMU solves this problem by being an executable ELF shared object as the ld-linux.so ELF interpreter. That way, it can be relocated at load time.
For system emulation, QEMU uses the mmap() system call to emulate the target CPU MMU. It works as long the emulated OS does not use an area reserved by the host OS (such as the area above 0xc0000000 on x86 Linux).
It is planned to add a slower but more precise MMU emulation with a software MMU.
In the directory `tests/', various interesting testing programs are available. There are used for regression testing.
Very simple statically linked x86 program, just to test QEMU during a port to a new host CPU.
Very simple statically linked ARM program, just to test QEMU during a port to a new host CPU.
This program executes most of the 16 bit and 32 bit x86 instructions and
generates a text output. It can be compared with the output obtained with
a real CPU or another emulator. The target make test runs this
program and a diff on the generated output.
The Linux system call modify_ldt() is used to create x86 selectors
to test some 16 bit addressing and 32 bit with segmentation cases.
The Linux system call vm86() is used to test vm86 emulation.
Various exceptions are raised to test most of the x86 user space exception reporting.
It is a simple benchmark. Care must be taken to interpret the results
because it mostly tests the ability of the virtual CPU to optimize the
rol x86 instruction and the condition code computations.
This document was generated on 27 June 2003 using texi2html 1.56k.