CVS log for src/sys/arch/i386/i386/gdt.c

Up to [cvs.NetBSD.org] / src / sys / arch / i386 / i386

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

x86: Sprinkle extensive commentary about %fs/%gs initialization.

Plus some other side quests like the three-stage GDT metamorphosis
lifecycle.

No functional change intended.

x86: Split most of pmap.h into pmap_private.h or vmparam.h.

This way pmap.h only contains the MD definition of the MI pmap(9)
API, which loads of things in the kernel rely on, so changing x86
pmap internals no longer requires recompiling the entire kernel every
time.

Callers needing these internals must now use machine/pmap_private.h.
Note: This is not x86/pmap_private.h because it contains three parts:

1. CPU-specific (different for i386/amd64) definitions used by...

2. common definitions, including Xenisms like xpmap_ptetomach,
   further used by...

3. more CPU-specific inlines for pmap_pte_* operations

So {amd64,i386}/pmap_private.h defines 1, includes x86/pmap_private.h
for 2, and then defines 3.  Maybe we should split that out into a new
pmap_pte.h to reduce this trouble.

No functional change intended, other than that some .c files must
include machine/pmap_private.h when previously uvm/uvm_pmap.h
polluted the namespace with pmap internals.

Note: This migrates part of i386/pmap.h into i386/vmparam.h --
specifically the parts that are needed for several constants defined
in vmparam.h:

VM_MAXUSER_ADDRESS
VM_MAX_ADDRESS
VM_MAX_KERNEL_ADDRESS
VM_MIN_KERNEL_ADDRESS

Since i386 needs PDP_SIZE in vmparam.h, I added it there on amd64
too, just to keep things parallel.

Sync with HEAD.

Merge the x86 gdt function and constant definitions

Mostly merge changes from HEAD upto 20200411

Switch to new PTE bits.

Sync with HEAD

Start replacing the x86 PTE bits.

We reorganise definitions for XEN source support as follows:

XEN - common sources required for baseline XEN support.
XENPV - sources required for support of XEN in PV mode.
XENPVHVM - sources required for support for XEN in HVM mode.
XENPVH - sources required for support for XEN in PVH mode.

Declare gdt_size as const, simplifies.

update from HEAD

The last GDT slots are reserved for Xen, so use FIRST_RESERVED_GDT_BYTE
instead of MAXGDTSIZ for gdt_size.
Xen kernels should boot again.

Fix a subtle ring0 escalation vulnerability in amd64, and implement a
mitigation against similar bugs.

The operations on segment registers can generate a page fault if there is
an issue when touching the in-memory gdt. Theoretically, it is never
supposed to happen, since the gdt is mapped correctly. However, in the
kernel we allow the gdt to be resized, and to do that, we allocate the
maximum amount of va needed by it, but only kenter a few pages until we
need more. Moreover, to avoid reloading the gdt each time we grow it, the
'size' field of gdtr is set to the maximum value. All of this means that
if a mov or iretq is done with a segment register whose index hits a page
that has not been kentered, a page fault is sent.

Such a page fault, if received in kernel mode, does not trigger a swapgs
on amd64; in other words, the kernel would be re-entered with the userland
tls.

And there just happens to be a place in compat_linux32 where the index of
%cs is controlled by userland, making it easy to trigger the page fault
and get kernel privileges.

The mitigation simply consists in abandoning the gdt_grow mechanism and
allocating/kentering the maximum size right away, in such a way that no
page fault can be triggered because of segment registers.

Sync with HEAD

gdt_size is now in bytes, but the HYPERVISOR_set_gdt() expects a number
of entries and has not been converted.
Xen/i386 now boots again.

Put some ()s in the macro (kre).

Use a bitmap-based allocator for i386, same as amd64. Several functions
are now identical - or nearly identical - on both sides. I couldn't test
this change on xen, because I'm having some unrelated issues with my VM
and I've spent enough time not understanding what's wrong with it.

Define MINGDTSIZ/MAXGDTSIZ in bytes, not in number of slots; same as
amd64.

Remove dead and broken code. It is not a bad idea to implement USER_LDT on
Xen, but it certainly shouldn't be done this way.

Sync with HEAD

Sync with HEAD

Analagous fix to that just committed to the equiv amd64 sources.
Note: this one has yet to be compile tested, so anything is possible...

Remove gdt_reload_cpu. GDTR takes a VA as base, and in our x86
implementation this VA is per-cpu and does not change; there is therefore
no need to remotely reload GDTR.

Localify, add a comment and merge some others.

Sync with HEAD

Rename ldt->ldtstore and gdt->gdtstore on i386. It reduces the diff with
amd64, and makes it easier to track down these variables on nxr - 'ldt'
and 'gdt' being common keywords.

Sync with HEAD.  (Note that most of these changes are simply $NetBSD$
tag issues.)

Put a limit in the percpu segment, so we can detect overflows on %fs.

Sync with HEAD

fix the build

Localify.

sync with head

Apply patch (requested by bouyer in ticket #1738).

Do not special-case XEN and always use the proper selectors for %fs and %gs
in buildcontext() and setregs(). The consequence was that signal handlers
would have the wrong %fs/%gs. Found by running atf tests under Xen/i386.

sync to latest -current.

(xen) - remove the (*xpq_cpu)() shim.We hasten the %fs/%gs setup process during boot.Although this is hacky, it lets us use the non-xen specificpmap_pte_xxx() functions in pmap code (and others).

sync with head

[merging from cherry-xenmp] Make the xen MMU op queue locking api private. Implement per-cpu queues.

Remove the "xpq lock", since we have per-cpu mmu queues now. This may need further testing. Also add some preliminary locking around queue-ops in the network backend driver

Sync with HEAD. Most notably: uvm/pmap work done by rmind@, and MP Xen
work of cherry@.

No regression observed on suspend/restore.

per-cpu gdt support for xen

gdtgrow() must extend the gdt size on all cpus, not just the primary

grow MP support for i386. boots to single user

Sync with HEAD

sync with head

Use lwp_getpcb() on x86 MD code, clean from struct user usage.

Add a flags argument to pmap_kenter_pa(9).
Patch showed on tech-kern@ http://mail-index.netbsd.org/tech-kern/2009/11/04/msg006434.html
No objections.

Sync with HEAD.

sync with head

Sync with HEAD.

on Xen the GDT has to be updated though HYPERVISOR_update_descriptor().
Export i386/i386/gdt.c:update_descriptor() and use it in x86_set_sdbase(),
as a direct write to the GDT will cause a kernel trap.
Fix PR port-xen/41401.

Sync with HEAD.

Commit is split, to avoid a "too many arguments" protocol error.

sync with head.

Sync with HEAD.

Pull up following revision(s) (requested by ad in ticket #656):
	sys/arch/amd64/amd64/gdt.c: revision 1.21 via patch
	sys/arch/amd64/amd64/machdep.c: revision 1.129 via patch
	sys/arch/i386/i386/gdt.c: revision 1.47 via patch
	sys/arch/i386/i386/kvm86.c: revision 1.17 via patch
	sys/arch/i386/i386/locore.S: revision 1.85 via patch
	sys/arch/i386/i386/machdep.c: revision 1.666 via patch
	sys/arch/i386/i386/vector.S: revision 1.45 via patch
	sys/arch/i386/include/pcb.h: revision 1.47 via patch
	sys/arch/x86/include/pmap.h: revision 1.22 via patch
	sys/arch/x86/include/sysarch.h: revision 1.8 via patch
	sys/arch/x86/x86/pmap.c: revision 1.80 via patch
	sys/arch/x86/x86/sys_machdep.c: revision 1.17 via patch
	sys/compat/linux/arch/i386/linux_machdep.c: revision 1.143 via patch
	sys/kern/init_main.c: revision 1.384 via patch
PR port-i386/40143 Viewing an mpeg transport stream with mplayer causes crash
Fix numerous problems:
1. LDT updates are not atomic.
2. Number of processes running with private LDTs and/or I/O bitmaps
   is not capped. System with high maxprocs can be paniced.
3. LDTR can be leaked over context switch.
4. GDT slot allocations can race, giving the same LDT slot to two procs.
5. Incomplete interrupt/trap frames can be stacked.
6. In some rare cases segment faults are not handled correctly.

PR port-i386/40143 Viewing an mpeg transport stream with mplayer causes crash

Fix numerous problems:

1. LDT updates are not atomic.

2. Number of processes running with private LDTs and/or I/O bitmaps
   is not capped. System with high maxprocs can be paniced.

3. LDTR can be leaked over context switch.

4. GDT slot allocations can race, giving the same LDT slot to two procs.

5. Incomplete interrupt/trap frames can be stacked.

6. In some rare cases segment faults are not handled correctly.

ansify function definitions

Sync with HEAD.

sync with head.

sync with head.

Remove clause 3 and 4 from TNF licenses

sync with HEAD

Sync with HEAD.

sync with head

Merge the bouyer-xeni386 branch to head, at tag bouyer-xeni386-merge1 (the
branch is still active and will see i386PAE support developement).
Sumary of changes:
- switch xeni386 to the x86/x86/pmap.c, and the xen/x86/x86_xpmap.c
  pmap bootstrap.
- merge back most of xen/i386/ to i386/i386
- change the build to reduce diffs between i386 and amd64 in file locations
- remove include files that were identical to the i386/amd64 counterparts,
  the build will find them via the xen-ma/machine link.

Merge xen bits to i386/i386/gdt.c. Convert remaining uses of PTE_* macros to
pmap_pte_* macros/inlines.
Fix think-o in pmap.c for native i386.

sync with HEAD

Sync with HEAD

i386:
	- make tss per-cpu.  this considerably speeds up context switch for,
	  at least, pentium4, where ltr instruction seems very slow.
i386, xen:
	- kill cpu_maxproc.
kvm86:
	- adapt to per-cpu tss.
	- cleanup and simplify.
	- move kvm86_mp_lock to more meaningful place.
	- disable preemption during a call.

Sync with HEAD.

Sync with HEAD.

Sync with HEAD

sync with head.

Sync with HEAD.

Set a 4G limit on the per-CPU segment (quicker for the MMU, and no reason
to limit it).

sync with HEAD

Merge the ppcoea-renovation branch to HEAD.

This branch was a major cleanup and rototill of many of the various OEA
cpu based PPC ports that focused on sharing as much code as possible
between the various ports to eliminate near-identical copies of files in
every tree.  Additionally there is a new PIC system that unifies the
interface to interrupt code for all different OEA ppc arches.  The work
for this branch was done by a variety of people, too long to list here.

TODO:
bebox still needs work to complete the transition to -renovation.
ofppc still needs a bunch of work, which I will be looking at.
ev64260 still needs to be renovated
amigappc was not attempted.

NOTES:
pmppc was removed as an arch, and moved to a evbppc target.

Sync with HEAD

sync with head.

Sync with HEAD.

Sync with HEAD.

Sync with HEAD.

Fix typo.

More calls to pmap_update().

Sync with head.

Sync with head.

sync with head.

Fix crappy locking around LDT updates. i386 part of PR kern/36046.

sync with head.

Replace some uses of lockmgr() / simplelocks.

sync with head.

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

merge ktrace-lwp.

Sync with HEAD. Here we go again...

Avoid shadowed variables.

sync with -current

Sync with HEAD.

merge yamt-km branch.
- don't use managed mappings/backing objects for wired memory allocations.
  save some resources like pv_entry.  also fix (most of) PR/27030.
- simplify kernel memory management API.
- simplify pmap bootstrap of some ports.
- some related cleanups.

- convert i386 to new apis.
- remove a pmap bootstrap kludge, which is no longer needed.

Fix the sync with head I botched.

Sync with HEAD.

Sync with HEAD

Uppercase CPU, plural is CPUs.

gdt_alloc_cpu: use unmanaged mappings as gdt_init and gdt_grow.

Make sure to include opt_foo.h if a defflag option FOO is used.

Sync with HEAD.

Fix comment.
De-__P().

Now that gdt_compact() is gone, #if 0'ed gdt_shrink() doesn't need to be
here, either. Remove other #if 0 code as well.

Catch up to -current.

sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work

Remove gdt_compact_slot, it was commented out already.
Expect callers to ldt_alloc and ldt_free to take the corresponding
pmap spinlock.

Merge Bill Sommerfeld's i386 MP branch. This code has some known
caveats, but works quite well in a lot of MP cases, and all
UP cases that I have tested. Parts of this will hopefully be
reworked in the not-too-distant future.

Since we're now allocating stuff out of the gdt earlier on,
copy more of it.

merged with -current as of today.
rearranged gdt init to deal with KVM86 changes; however, KVM86 has no
chance of working yet; good thing it's not part of GENERIC

No longer need to pull in lwp.h; proc.h pulls it in for us.

Initialize gdt on boot cpu earlier in boot instead of in gdt_init().
Put apm back in GENERIC and APM_NO_IDLE in GENERIC.MP.
Subsystems which install gdt entries during autoconf no longer explode.
(Do *not* expect full apm functionality on multiprocessors).

Sync kqueue branch with -current.

Catch up to -current.

Avoid copying bits of boot-time gdt which might not be there.
From Frank van der Linden

Set up per-cpu GDT.
Use %fs for per-cpu data access.

Yet another mergeup.

New work:
 - Frank van der Linden's tlb shootdown fix.
Mainline functionality merged:
 - IrDA
 - Kernel RCSID's
 - Transmeta CPU support
 - ACPI
 - XMM register access through procfs

allocate and free memory explicitly rather than abusing uvm_map_pageable().

add RCSID

Initial commit of scheduler activations and lightweight process support.

Correct merge botches.
Post-merge kernel now boots multiuser.

Snapshot of merge-in-progress with -current.

[Not expected to build].  Catch up with the last N months worth of
changes to -current.

Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.

Update from trunk:
Move the TSS selector out of the PCB and put it into mdproc.  We
need to access this when we have the proclist locked for reading,
and thus cannot store it in the PCB (which may be swapped out).

As part of this, call pmap_activate() from cpu_switch() to switch
to the new address space, and refresh the PCB's copy of the LDT
selector from the pmap structure (see above paragraph).  We need
to do this for MP support anyhow.

Fixes a "panic: spinlock_switchcheck: CPU 0 has 1 spin locks" via
gdt_compact() reported by Nathan Williams.

Move the TSS selector out of the PCB and put it into mdproc.  We
need to access this when we have the proclist locked for reading,
and thus cannot store it in the PCB (which may be swapped out).

As part of this, call pmap_activate() from cpu_switch() to switch
to the new address space, and refresh the PCB's copy of the LDT
selector from the pmap structure (see above paragraph).  We need
to do this for MP support anyhow.

Fixes a "panic: spinlock_switchcheck: CPU 0 has 1 spin locks" via
gdt_compact() reported by Nathan Williams.

Split creation/allocation of the memory containing the GDT from
loading of the GDT descriptor into the processor; the former happens
once, the latter, once per CPU.

file gdt.c was added on branch sommerfeld_i386mp_1 on 2000-08-12 16:09:24 +0000

remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>

remove/move more mach vm header files:

	<vm/pglist.h> -> <uvm/uvm_pglist.h>
	<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
	<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
	<vm/vm_object.h> -> nothing
	<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

Update from trunk.

Lock the proclist as appropriate.

Use the lock manager to lock the GDT, rather than a non-MP-safe homegrown
method.

Sync w/ -current.

Update a uvm_map_pageable() call that I forgot to commit previously.

Move the user-set LDT out of the PCB and into the pmap.  Applications
which set the LDT and share VM space (e.g. new versions of WINE) expect
the LDT to be logically coupled to the address space.  Use the new pmap_fork()
interface to copy non-shared user-set LDTs when the address space is forked.

completely remove Mach VM support.  all that is left is the all the
header files as UVM still uses (most of) these.

vm_offset_t -> {vaddr_t,paddr_t}, vm_size_t -> vsize_t

bzero->memset, bcopy->memcpy, bcmp->memcmp

- add defopt's for UVM, UVMHIST and PMAP_NEW.
- remove unnecessary UVMHIST_DECL's.

add the i386 MD portions for UVM.

Some rearrangement of the startup code:
* Move some of the memory bootstrapping and the DDB startup earlier.
* Always put the IDT and the initial GDT and LDT in a separate page.
* Various minor changes.

Pull rev 1.10 up from trunk (mycroft)

Some general cleanup:
* Call gdt_init() earlier on.
* Don't bother to check for TSS and LDT descriptors in verr_gdt().
* Nuke the sti special case for trace and breakpoint traps.
* Move some variable declarations into more appropriate locations.

Update marc-pcmcia branch from trunk.

Fix tipo inherited from old version of TNF copyright template.

Prototype and gcc -Wall cleanups

Copyright assigned to the NetBSD Foundation

Bring in changes for mondo patch 2.

Allocate the GDT area in a different way.

fixup copyright notice before the release

Update from main branch.

Rename some variables for consistency.

Maintain a free list.

Go back to statically allocating the GDT entries for proc0.

Various changes from John Kohl and me:
Map kernel stacks only at unique addresses.
Use one TSS per process.
Add sysarch calls for modifying IOPL and the I/O permission bitmap.
Add a compacting GDT entry allocator, for TSS and LDT selectors.
Enable modifying %fs and %gs with PT_SETREGS.
Sanitize various bits of code.