This patch is a rollup of all the core pieces of the Xen
implementation, including:
- booting and setup
- pagetable setup
- privileged instructions
- segmentation
- interrupt flags
- upcalls
- multicall batching
BOOTING AND SETUP
The vmlinux image is decorated with ELF notes which tell the Xen
domain builder what the kernel's requirements are; the domain builder
then constructs the address space accordingly and starts the kernel.
Xen has its own entrypoint for the kernel (contained in an ELF note).
The ELF notes are set up by xen-head.S, which is included into head.S.
In principle it could be linked separately, but it seems to provoke
lots of binutils bugs.
Because the domain builder starts the kernel in a fairly sane state
(32-bit protected mode, paging enabled, flat segments set up), there's
not a lot of setup needed before starting the kernel proper. The main
steps are:
1. Install the Xen paravirt_ops, which is simply a matter of a
structure assignment.
2. Set init_mm to use the Xen-supplied pagetables (analogous to the
head.S generated pagetables in a native boot).
3. Reserve address space for Xen, since it takes a chunk at the top
of the address space for its own use.
4. Call start_kernel()
PAGETABLE SETUP
Once we hit the main kernel boot sequence, it will end up calling back
via paravirt_ops to set up various pieces of Xen specific state. One
of the critical things which requires a bit of extra care is the
construction of the initial init_mm pagetable. Because Xen places
tight constraints on pagetables (an active pagetable must always be
valid, and must always be mapped read-only to the guest domain), we
need to be careful when constructing the new pagetable to keep these
constraints in mind. It turns out that the easiest way to do this is
use the initial Xen-provided pagetable as a template, and then just
insert new mappings for memory where a mapping doesn't already exist.
This means that during pagetable setup, it uses a special version of
xen_set_pte which ignores any attempt to remap a read-only page as
read-write (since Xen will map its own initial pagetable as RO), but
lets other changes to the ptes happen, so that things like NX are set
properly.
PRIVILEGED INSTRUCTIONS AND SEGMENTATION
When the kernel runs under Xen, it runs in ring 1 rather than ring 0.
This means that it is more privileged than user-mode in ring 3, but it
still can't run privileged instructions directly. Non-performance
critical instructions are dealt with by taking a privilege exception
and trapping into the hypervisor and emulating the instruction, but
more performance-critical instructions have their own specific
paravirt_ops. In many cases we can avoid having to do any hypercalls
for these instructions, or the Xen implementation is quite different
from the normal native version.
The privileged instructions fall into the broad classes of:
Segmentation: setting up the GDT and the GDT entries, LDT,
TLS and so on. Xen doesn't allow the GDT to be directly
modified; all GDT updates are done via hypercalls where the new
entries can be validated. This is important because Xen uses
segment limits to prevent the guest kernel from damaging the
hypervisor itself.
Traps and exceptions: Xen uses a special format for trap entrypoints,
so when the kernel wants to set an IDT entry, it needs to be
converted to the form Xen expects. Xen sets int 0x80 up specially
so that the trap goes straight from userspace into the guest kernel
without going via the hypervisor. sysenter isn't supported.
Kernel stack: The esp0 entry is extracted from the tss and provided to
Xen.
TLB operations: the various TLB calls are mapped into corresponding
Xen hypercalls.
Control registers: all the control registers are privileged. The most
important is cr3, which points to the base of the current pagetable,
and we handle it specially.
Another instruction we treat specially is CPUID, even though its not
privileged. We want to control what CPU features are visible to the
rest of the kernel, and so CPUID ends up going into a paravirt_op.
Xen implements this mainly to disable the ACPI and APIC subsystems.
INTERRUPT FLAGS
Xen maintains its own separate flag for masking events, which is
contained within the per-cpu vcpu_info structure. Because the guest
kernel runs in ring 1 and not 0, the IF flag in EFLAGS is completely
ignored (and must be, because even if a guest domain disables
interrupts for itself, it can't disable them overall).
(A note on terminology: "events" and interrupts are effectively
synonymous. However, rather than using an "enable flag", Xen uses a
"mask flag", which blocks event delivery when it is non-zero.)
There are paravirt_ops for each of cli/sti/save_fl/restore_fl, which
are implemented to manage the Xen event mask state. The only thing
worth noting is that when events are unmasked, we need to explicitly
see if there's a pending event and call into the hypervisor to make
sure it gets delivered.
UPCALLS
Xen needs a couple of upcall (or callback) functions to be implemented
by each guest. One is the event upcalls, which is how events
(interrupts, effectively) are delivered to the guests. The other is
the failsafe callback, which is used to report errors in either
reloading a segment register, or caused by iret. These are
implemented in i386/kernel/entry.S so they can jump into the normal
iret_exc path when necessary.
MULTICALL BATCHING
Xen provides a multicall mechanism, which allows multiple hypercalls
to be issued at once in order to mitigate the cost of trapping into
the hypervisor. This is particularly useful for context switches,
since the 4-5 hypercalls they would normally need (reload cr3, update
TLS, maybe update LDT) can be reduced to one. This patch implements a
generic batching mechanism for hypercalls, which gets used in many
places in the Xen code.
Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>
Cc: Ian Pratt <ian.pratt@xensource.com>
Cc: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
Cc: Adrian Bunk <bunk@stusta.de>
Add function helper, fb_is_primary_device(). Given struct fb_info, it will
return a nonzero value if the device is the primary display.
Currently, only the i386 is supported where the function checks for the
IORESOURCE_ROM_SHADOW flag.
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Cc: David Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reverts commit c8fdd24725.
It turns out the kernel was correct, and the gcc complaint was a gcc
bug. The preferred stack boundary is expressed not in bytes, but in the
the log2() of the preferred boundary, so "-mpreferred-stack-boundary=2"
is in fact exactly what we want, but a gcc that is compiled for x86-64
will consider it an error (because the 64-bit calling sequence says that
the stack should be 16-byte aligned) even if we are then using "-m32" to
generate 32-bit code.
Noted-by: Mikulas Patocka <mikulas@artax.karlin.mff.cuni.cz>
Cc: Jan Hubicka <jh@suse.cz>
Acked-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The kernel only supports gcc 3.2+ now so it doesn't make sense
anymore to explicitely check for options this compiler version
already has.
This actually fixes a bug. The -mprefered-stack-boundary check
never worked because gcc rightly complains
CC arch/i386/kernel/asm-offsets.s
cc1: -mpreferred-stack-boundary=2 is not between 4 and 12
We just never saw the error because of cc-options.
I changed it to 4 to actually work.
Tested by compiling i386 and x86-64 defconfig with gcc 3.2.
Should speed up the build time a tiny bit and improve
stack usage on i386 slightly.
Signed-off-by: Andi Kleen <ak@suse.de>
Jeremy Fitzhardinge suggested the use of -freg-struct-return, which does
structure-returns (such as when using pte_t) in registers instead of on
the stack.
that is indeed so, and this option reduced the kernel size a bit:
text data bss dec hex filename
4799506 543456 3760128 9103090 8ae6f2 vmlinux.before
4798117 543456 3760128 9101701 8ae185 vmlinux.after
the resulting kernel booted fine on my testbox. Lets go for it.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-mregparm=3 has been enabled by default for some time on i386, and AFAIK
there aren't any problems with it left.
This patch removes the REGPARM config option and sets -mregparm=3
unconditionally.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andi Kleen <ak@suse.de>
This patch modifies the i386 kernel so that if CONFIG_RELOCATABLE is
selected it will be able to be loaded at any 4K aligned address below
1G. The technique used is to compile the decompressor with -fPIC and
modify it so the decompressor is fully relocatable. For the main
kernel relocations are generated. Resulting in a kernel that is relocatable
with no runtime overhead and no need to modify the source code.
A reserved 32bit word in the parameters has been assigned
to serve as a stack so we figure out where are running.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Vivek Goyal <vgoyal@in.ibm.com>
Signed-off-by: Andi Kleen <ak@suse.de>
This avoids some problems with gcc 4.x and earlier generating
invalid unwind information. In 4.1 the option is default
when unwind information is enabled.
And it seems to generate smaller code too, so it's probably
a good thing on its own. With gcc 4.0:
i386:
4683198 902112 480868 6066178 5c9002 vmlinux (before)
4449895 902112 480868 5832875 5900ab vmlinux (after)
x86-64:
4939761 1449584 648216 7037561 6b6279 vmlinux (before)
4854193 1449584 648216 6951993 6a1439 vmlinux (after)
On 4.1 it shouldn't make much difference because it is
default when unwind is enabled anyways.
Suggested by Michael Matz and Jan Beulich
Cc: jbeulich@novell.com
Signed-off-by: Andi Kleen <ak@suse.de>
This was copied, pasted but not edited.
Cc: Andi Kleen <ak@muc.de>
Cc: Jan Beulich <jbeulich@novell.com>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Andi Kleen <ak@suse.de>
Current gcc generates calls not jumps to noreturn functions. When that happens the
return address can point to the next function, which confuses the unwinder.
This patch works around it by marking asynchronous exception
frames in contrast normal call frames in the unwind information. Then teach
the unwinder to decode this.
For normal call frames the unwinder now subtracts one from the address which avoids
this problem. The standard libgcc unwinder uses the same trick.
It doesn't include adjustment of the printed address (i.e. for the original
example, it'd still be kernel_math_error+0 that gets displayed, but the
unwinder wouldn't get confused anymore.
This only works with binutils 2.6.17+ and some versions of H.J.Lu's 2.6.16
unfortunately because earlier binutils don't support .cfi_signal_frame
[AK: added automatic detection of the new binutils and wrote description]
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Andi Kleen <ak@suse.de>
... instead of using a CONFIG option. The config option still controls
if the resulting executable actually has unwind information.
This is useful to prevent compilation errors when users select
CONFIG_STACK_UNWIND on old binutils and also allows to use
CFI in the future for non kernel debugging applications.
Cc: jbeulich@novell.com
Cc: sam@ravnborg.org
Signed-off-by: Andi Kleen <ak@suse.de>
Add a "make isoimage" to i386 and x86-64, which allows the automatic
creation of a bootable CD image. It also adds an option FDINITRD= to
include an initrd of the user's choice in generated floppy- or CD boot
images. Finally, some minor cleanups of the image generation code.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Andi Kleen <ak@muc.de>
Cc: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
gcc should handle this anyways, and it causes problems when
sprintf is turned into strcpy by gcc behind our backs and
the C fallback version of strcpy is actually defining __builtin_strcpy
Then drop -ffreestanding from the main Makefile because it isn't
needed anymore and implies -fno-builtin, which is wrong now.
(it was only added for x86-64, so dropping it should be safe)
Noticed by Roman Zippel
Cc: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The kbuild system takes advantage of an incorrect behavior in GNU make.
Once this behavior is fixed, all files in the kernel rebuild every time,
even if nothing has changed. This patch ensures kbuild works with both
the incorrect and correct behaviors of GNU make.
For more details on the incorrect behavior, see:
http://lists.gnu.org/archive/html/bug-make/2006-03/msg00003.html
Changes in this patch:
- Keep all targets that are to be marked .PHONY in a variable, PHONY.
- Add .PHONY: $(PHONY) to mark them properly.
- Remove any $(PHONY) files from the $? list when determining whether
targets are up-to-date or not.
Signed-off-by: Paul Smith <psmith@gnu.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Since we do no longer support any gcc < 3.0, there's no need to check
for it..
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Cc: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Allow gcc4 compilers to optimize unit-at-a-time.
This flag enables gcc to "see" the entire C file before making optimisation
decisions such as inline, which results in gcc making better decisions. One
of the immediate effects of this is that static functions that are used only
once now get inlined.
gcc 3.4 has this flag as well, however gcc 3.x have a problem with inlining
and stacks and as a result, enabling this flag there would cause excessive and
unacceptable stack use. This problem is fixed in the gcc 4.x series. The
x86-64 architecture already enables this feature so it's well tested already.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@infradead.org>
Acked-by: Jeff Garzik <jgarzik@pobox.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This removes the dependency from vmlinux to install, thus avoiding the
current situation where "make install" has a nasty tendency to leave
root-turds in the working directory.
It also updates x86-64 to be in sync with i386.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
This was causing some ordering problems. Remove the up-front evaluation
and just revaluate the compiler version each time we need it.
(The up-front evaluation was problematic because some architectures modify
the value of $(CC)).
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Make UML share the underlying cpu-specific tuning done on i386.
Actually, for now many config options aren't used a lot - but that can be done
later. Also, UML relies on GCC optimization for things like memcpy and such
more than i386, so specifying the correct -march and -mtune should be enough.
Later, we may want to correct some other stuff.
For instance, since FPU context switching, for us, is done (at least
partially, i.e. between our kernelspace and userspace) by the host, we may
allow usage of FPU operations by GCC. This doesn't hold for kernelspace vs.
kernelspace, but we don't support preemption.
Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Building asm-offsets.h has been moved to a seperate Kbuild file
located in the top-level directory. This allow us to share the
functionality across the architectures.
The old rules in architecture specific Makefiles will die
in subsequent patches.
Furhtermore the usual kbuild dependency tracking is now used
when deciding to rebuild asm-offsets.s. So we no longer risk
to fail a rebuild caused by asm-offsets.c dependencies being touched.
With this common rule-set we now force the same name across
all architectures. Following patches will fix the rest.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
This allows the i386 architecture to be built on a system with a biarch
compiler that defaults to x86-64, merely by specifying ARCH=i386.
As previously discussed, this uses the equivalent logic to the ppc port.
Signed-Off-By: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
- Changed Name/defines from "Geode GX" to "Geode GX1" for clarification
- Dropped "-march=i586" in favor of "-march=i486"
- Dopped X86_OOSTORE support for Geode GX1
Signed-off-by: Kianusch Sayah Karadji <kianusch@sk-tech.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Removing the dependency on vmlinux for the install target raised a few
complaints, so instead a new target i added: kernel_install.
kernel_install will install the kernel just like the ordinary install target.
The only difference is that install has a dependency on vmlinux,
kernel_install does not. Therefore kernel_install is the best choice
when accessing the kernel over a NFS mount or as another user.
kernel_install is similar to modules_install in the fact that neither does
a full kernel compile before performing the install.
In this way they are good for root use. Also added back the
dependency on vmlinux for the install target so peoples scripts are no
longer broken.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
