There is no particular reason for keeping the "sub 0, %rsp" insn within
the BPF's x64 JIT prologue.
When tail call code was skipping the whole prologue section these 7
bytes that represent the rsp subtraction could not be simply discarded
as the jump target address would be broken. An option to address that
would be to substitute it with nop7.
Right now tail call is skipping only first 11 bytes of target program's
prologue and "sub X, %rsp" is the first insn that is processed, so if
stack depth is zero then this insn could be omitted without the need for
nop7 swap.
Therefore, do not emit the "sub 0, %rsp" in prologue when program is not
making use of R10 register. Also, make the emission of "add X, %rsp"
conditional in tail call code logic and take into account the presence
of mentioned insn when calculating the jump offsets.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200929204653.4325-3-maciej.fijalkowski@intel.com
Back when all of the callee-saved registers where always pushed to stack
in x64 JIT prologue, tail call counter was placed at the bottom of the
BPF program's stack frame that had a following layout:
+-------------+
| ret addr |
+-------------+
| rbp | <- rbp
+-------------+
| |
| free space |
| from: |
| sub $x,%rsp |
| |
+-------------+
| rbx |
+-------------+
| r13 |
+-------------+
| r14 |
+-------------+
| r15 |
+-------------+
| tail call | <- rsp
| counter |
+-------------+
In order to restore the callee saved registers, epilogue needed to
explicitly toss away the tail call counter via "pop %rbx" insn, so that
%rsp would be back at the place where %r15 was stored.
Currently, the tail call counter is placed on stack *before* the callee
saved registers (brackets on rbx through r15 mean that they are now
pushed to stack only if they are used):
+-------------+
| ret addr |
+-------------+
| rbp | <- rbp
+-------------+
| |
| free space |
| from: |
| sub $x,%rsp |
| |
+-------------+
| tail call |
| counter |
+-------------+
( rbx )
+-------------+
( r13 )
+-------------+
( r14 )
+-------------+
( r15 ) <- rsp
+-------------+
For the record, the epilogue insns consist of (assuming all of the
callee saved registers are used by program):
pop %r15
pop %r14
pop %r13
pop %rbx
pop %rcx
leaveq
retq
"pop %rbx" for getting rid of tail call counter was not an option
anymore as it would overwrite the restored value of %rbx register, so it
was changed to use the %rcx register.
Since epilogue can start popping the callee saved registers right away
without any additional work, the "pop %rcx" could be dropped altogether
as "leave" insn will simply move the %rbp to %rsp. IOW, tail call
counter does not need the explicit handling.
Having in mind the explanation above and the actual reason for that,
let's piggy back on "leave" insn for discarding the tail call counter
from stack and remove the "pop %rcx" from epilogue.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200929204653.4325-2-maciej.fijalkowski@intel.com
Alexei Starovoitov says:
====================
pull-request: bpf-next 2020-09-23
The following pull-request contains BPF updates for your *net-next* tree.
We've added 95 non-merge commits during the last 22 day(s) which contain
a total of 124 files changed, 4211 insertions(+), 2040 deletions(-).
The main changes are:
1) Full multi function support in libbpf, from Andrii.
2) Refactoring of function argument checks, from Lorenz.
3) Make bpf_tail_call compatible with functions (subprograms), from Maciej.
4) Program metadata support, from YiFei.
5) bpf iterator optimizations, from Yonghong.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Two minor conflicts:
1) net/ipv4/route.c, adding a new local variable while
moving another local variable and removing it's
initial assignment.
2) drivers/net/dsa/microchip/ksz9477.c, overlapping changes.
One pretty prints the port mode differently, whilst another
changes the driver to try and obtain the port mode from
the port node rather than the switch node.
Signed-off-by: David S. Miller <davem@davemloft.net>
- fix fault on page table writes during instruction fetch
s390:
- doc improvement
x86:
- The obvious patches are always the ones that turn out to be
completely broken. /me hangs his head in shame.
-----BEGIN PGP SIGNATURE-----
iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAl9nyjsUHHBib256aW5p
QHJlZGhhdC5jb20ACgkQv/vSX3jHroNpcAf/bsW1B+Q8QJnyfU4RSiX28lG8Ki9F
9A0aVJPW4U/x7COZuhldrQGkbHDA5agavCevghMuOqWkz2gs6ihpAGgzfG+FVIm7
2yi4k9A90kPrMSBf8qaLgvybGNO6uxGpJmv54MjHpkLPUEz+J1MuB9D6eEqBZkWz
ncOSsGS2eeUFpqulA9DCN3O3PbaFeAXPNJnDNGqxrGjV7CriosRlbK02PVxTQzvT
nuGzDgaOmmRXntIQ7hrk9DJlHm7gH2jH8TK9gB2xm0yuVm2/nNlpkY6rP6NDUdLs
OrJOzxWOcSO8HRgBhlFhED/8heTqHCJS1vMUI3M6Z62p324TjRDjRC+4ow==
=tpZU
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm fixes from Paolo Bonzini:
"ARM:
- fix fault on page table writes during instruction fetch
s390:
- doc improvement
x86:
- The obvious patches are always the ones that turn out to be
completely broken. /me hangs his head in shame"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
Revert "KVM: Check the allocation of pv cpu mask"
KVM: arm64: Remove S1PTW check from kvm_vcpu_dabt_iswrite()
KVM: arm64: Assume write fault on S1PTW permission fault on instruction fetch
docs: kvm: add documentation for KVM_CAP_S390_DIAG318
* Disable relocation relaxation for the compressed kernel when not built
as -pie as in that case kernels built with clang and linked with LLD
fail to boot due to the linker optimizing some instructions in non-PIE
form; the gory details in the commit message, from Arvind Sankar.
* A fix for the "bad bp value" warning issued by the frame-pointer
unwinder, from Josh Poimboeuf.
-----BEGIN PGP SIGNATURE-----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=og1U
-----END PGP SIGNATURE-----
Merge tag 'x86_urgent_for_v5.9_rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Borislav Petkov:
- A defconfig fix (Daniel Díaz)
- Disable relocation relaxation for the compressed kernel when not
built as -pie as in that case kernels built with clang and linked
with LLD fail to boot due to the linker optimizing some instructions
in non-PIE form; the gory details in the commit message (Arvind
Sankar)
- A fix for the "bad bp value" warning issued by the frame-pointer
unwinder (Josh Poimboeuf)
* tag 'x86_urgent_for_v5.9_rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/unwind/fp: Fix FP unwinding in ret_from_fork
x86/boot/compressed: Disable relocation relaxation
x86/defconfigs: Explicitly unset CONFIG_64BIT in i386_defconfig
The commit 0f99022210 ("KVM: Check the allocation of pv cpu mask") we
have in 5.9-rc5 has two issue:
1) Compilation fails for !CONFIG_SMP, see:
https://bugzilla.kernel.org/show_bug.cgi?id=209285
2) This commit completely disables PV TLB flush, see
https://lore.kernel.org/kvm/87y2lrnnyf.fsf@vitty.brq.redhat.com/
The allocation problem is likely a theoretical one, if we don't
have memory that early in boot process we're likely doomed anyway.
Let's solve it properly later.
This reverts commit 0f99022210.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
- Add support for the Lakefield chip to the RAPL power capping
driver (Ricardo Neri).
- Modify the ACPI processor idle driver to prevent it from triggering
RCU-lockdep complaints which has started to happen after recent
changes in that area (Peter Zijlstra).
-----BEGIN PGP SIGNATURE-----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=gBKv
-----END PGP SIGNATURE-----
Merge tag 'pm-5.9-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management updates from Rafael Wysocki:
"These add a new CPU ID to the RAPL power capping driver and prevent
the ACPI processor idle driver from triggering RCU-lockdep complaints.
Specifics:
- Add support for the Lakefield chip to the RAPL power capping driver
(Ricardo Neri).
- Modify the ACPI processor idle driver to prevent it from triggering
RCU-lockdep complaints which has started to happen after recent
changes in that area (Peter Zijlstra)"
* tag 'pm-5.9-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI: processor: Take over RCU-idle for C3-BM idle
cpuidle: Allow cpuidle drivers to take over RCU-idle
ACPI: processor: Use CPUIDLE_FLAG_TLB_FLUSHED
ACPI: processor: Use CPUIDLE_FLAG_TIMER_STOP
powercap: RAPL: Add support for Lakefield
There have been some reports of "bad bp value" warnings printed by the
frame pointer unwinder:
WARNING: kernel stack regs at 000000005bac7112 in sh:1014 has bad 'bp' value 0000000000000000
This warning happens when unwinding from an interrupt in
ret_from_fork(). If entry code gets interrupted, the state of the
frame pointer (rbp) may be undefined, which can confuse the unwinder,
resulting in warnings like the above.
There's an in_entry_code() check which normally silences such
warnings for entry code. But in this case, ret_from_fork() is getting
interrupted. It recently got moved out of .entry.text, so the
in_entry_code() check no longer works.
It could be moved back into .entry.text, but that would break the
noinstr validation because of the call to schedule_tail().
Instead, initialize each new task's RBP to point to the task's entry
regs via an encoded frame pointer. That will allow the unwinder to
reach the end of the stack gracefully.
Fixes: b9f6976bfb ("x86/entry/64: Move non entry code into .text section")
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Reported-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/f366bbf5a8d02e2318ee312f738112d0af74d16f.1600103007.git.jpoimboe@redhat.com
This commit serves two things:
1) it optimizes BPF prologue/epilogue generation
2) it makes possible to have tailcalls within BPF subprogram
Both points are related to each other since without 1), 2) could not be
achieved.
In [1], Alexei says:
"The prologue will look like:
nop5
xor eax,eax // two new bytes if bpf_tail_call() is used in this
// function
push rbp
mov rbp, rsp
sub rsp, rounded_stack_depth
push rax // zero init tail_call counter
variable number of push rbx,r13,r14,r15
Then bpf_tail_call will pop variable number rbx,..
and final 'pop rax'
Then 'add rsp, size_of_current_stack_frame'
jmp to next function and skip over 'nop5; xor eax,eax; push rpb; mov
rbp, rsp'
This way new function will set its own stack size and will init tail
call
counter with whatever value the parent had.
If next function doesn't use bpf_tail_call it won't have 'xor eax,eax'.
Instead it would need to have 'nop2' in there."
Implement that suggestion.
Since the layout of stack is changed, tail call counter handling can not
rely anymore on popping it to rbx just like it have been handled for
constant prologue case and later overwrite of rbx with actual value of
rbx pushed to stack. Therefore, let's use one of the register (%rcx) that
is considered to be volatile/caller-saved and pop the value of tail call
counter in there in the epilogue.
Drop the BUILD_BUG_ON in emit_prologue and in
emit_bpf_tail_call_indirect where instruction layout is not constant
anymore.
Introduce new poke target, 'tailcall_bypass' to poke descriptor that is
dedicated for skipping the register pops and stack unwind that are
generated right before the actual jump to target program.
For case when the target program is not present, BPF program will skip
the pop instructions and nop5 dedicated for jmpq $target. An example of
such state when only R6 of callee saved registers is used by program:
ffffffffc0513aa1: e9 0e 00 00 00 jmpq 0xffffffffc0513ab4
ffffffffc0513aa6: 5b pop %rbx
ffffffffc0513aa7: 58 pop %rax
ffffffffc0513aa8: 48 81 c4 00 00 00 00 add $0x0,%rsp
ffffffffc0513aaf: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
ffffffffc0513ab4: 48 89 df mov %rbx,%rdi
When target program is inserted, the jump that was there to skip
pops/nop5 will become the nop5, so CPU will go over pops and do the
actual tailcall.
One might ask why there simply can not be pushes after the nop5?
In the following example snippet:
ffffffffc037030c: 48 89 fb mov %rdi,%rbx
(...)
ffffffffc0370332: 5b pop %rbx
ffffffffc0370333: 58 pop %rax
ffffffffc0370334: 48 81 c4 00 00 00 00 add $0x0,%rsp
ffffffffc037033b: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
ffffffffc0370340: 48 81 ec 00 00 00 00 sub $0x0,%rsp
ffffffffc0370347: 50 push %rax
ffffffffc0370348: 53 push %rbx
ffffffffc0370349: 48 89 df mov %rbx,%rdi
ffffffffc037034c: e8 f7 21 00 00 callq 0xffffffffc0372548
There is the bpf2bpf call (at ffffffffc037034c) right after the tailcall
and jump target is not present. ctx is in %rbx register and BPF
subprogram that we will call into on ffffffffc037034c is relying on it,
e.g. it will pick ctx from there. Such code layout is therefore broken
as we would overwrite the content of %rbx with the value that was pushed
on the prologue. That is the reason for the 'bypass' approach.
Special care needs to be taken during the install/update/remove of
tailcall target. In case when target program is not present, the CPU
must not execute the pop instructions that precede the tailcall.
To address that, the following states can be defined:
A nop, unwind, nop
B nop, unwind, tail
C skip, unwind, nop
D skip, unwind, tail
A is forbidden (lead to incorrectness). The state transitions between
tailcall install/update/remove will work as follows:
First install tail call f: C->D->B(f)
* poke the tailcall, after that get rid of the skip
Update tail call f to f': B(f)->B(f')
* poke the tailcall (poke->tailcall_target) and do NOT touch the
poke->tailcall_bypass
Remove tail call: B(f')->C(f')
* poke->tailcall_bypass is poked back to jump, then we wait the RCU
grace period so that other programs will finish its execution and
after that we are safe to remove the poke->tailcall_target
Install new tail call (f''): C(f')->D(f'')->B(f'').
* same as first step
This way CPU can never be exposed to "unwind, tail" state.
Last but not least, when tailcalls get mixed with bpf2bpf calls, it
would be possible to encounter the endless loop due to clearing the
tailcall counter if for example we would use the tailcall3-like from BPF
selftests program that would be subprogram-based, meaning the tailcall
would be present within the BPF subprogram.
This test, broken down to particular steps, would do:
entry -> set tailcall counter to 0, bump it by 1, tailcall to func0
func0 -> call subprog_tail
(we are NOT skipping the first 11 bytes of prologue and this subprogram
has a tailcall, therefore we clear the counter...)
subprog -> do the same thing as entry
and then loop forever.
To address this, the idea is to go through the call chain of bpf2bpf progs
and look for a tailcall presence throughout whole chain. If we saw a single
tail call then each node in this call chain needs to be marked as a subprog
that can reach the tailcall. We would later feed the JIT with this info
and:
- set eax to 0 only when tailcall is reachable and this is the entry prog
- if tailcall is reachable but there's no tailcall in insns of currently
JITed prog then push rax anyway, so that it will be possible to
propagate further down the call chain
- finally if tailcall is reachable, then we need to precede the 'call'
insn with mov rax, [rbp - (stack_depth + 8)]
Tail call related cases from test_verifier kselftest are also working
fine. Sample BPF programs that utilize tail calls (sockex3, tracex5)
work properly as well.
[1]: https://lore.kernel.org/bpf/20200517043227.2gpq22ifoq37ogst@ast-mbp.dhcp.thefacebook.com/
Suggested-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reflect the actual purpose of poke->ip and rename it to
poke->tailcall_target so that it will not the be confused with another
poke target that will be introduced in next commit.
While at it, do the same thing with poke->ip_stable - rename it to
poke->tailcall_target_stable.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Currently, %rax is used to store the jump target when BPF program is
emitting the retpoline instructions that are handling the indirect
tailcall.
There is a plan to use %rax for different purpose, which is storing the
tail call counter. In order to preserve this value across the tailcalls,
adjust the BPF indirect tailcalls so that the target program will reside
in %rcx and teach the retpoline instructions about new location of jump
target.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Make acpi_processor_idle() use the generic TLB flushing code.
This again removes RCU usage after rcu_idle_enter().
(XXX make every C3 invalidate TLBs, not just C3-BM)
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Yunhai Zhang recently fixed a VGA software scrollback bug in commit
ebfdfeeae8 ("vgacon: Fix for missing check in scrollback handling"),
but that then made people look more closely at some of this code, and
there were more problems on the vgacon side, but also the fbcon software
scrollback.
We don't really have anybody who maintains this code - probably because
nobody actually _uses_ it any more. Sure, people still use both VGA and
the framebuffer consoles, but they are no longer the main user
interfaces to the kernel, and haven't been for decades, so these kinds
of extra features end up bitrotting and not really being used.
So rather than try to maintain a likely unused set of code, I'll just
aggressively remove it, and see if anybody even notices. Maybe there
are people who haven't jumped on the whole GUI badnwagon yet, and think
it's just a fad. And maybe those people use the scrollback code.
If that turns out to be the case, we can resurrect this again, once
we've found the sucker^Wmaintainer for it who actually uses it.
Reported-by: NopNop Nop <nopitydays@gmail.com>
Tested-by: Willy Tarreau <w@1wt.eu>
Cc: 张云海 <zhangyunhai@nsfocus.com>
Acked-by: Andy Lutomirski <luto@amacapital.net>
Acked-by: Willy Tarreau <w@1wt.eu>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The x86-64 psABI [0] specifies special relocation types
(R_X86_64_[REX_]GOTPCRELX) for indirection through the Global Offset
Table, semantically equivalent to R_X86_64_GOTPCREL, which the linker
can take advantage of for optimization (relaxation) at link time. This
is supported by LLD and binutils versions 2.26 onwards.
The compressed kernel is position-independent code, however, when using
LLD or binutils versions before 2.27, it must be linked without the -pie
option. In this case, the linker may optimize certain instructions into
a non-position-independent form, by converting foo@GOTPCREL(%rip) to $foo.
This potential issue has been present with LLD and binutils-2.26 for a
long time, but it has never manifested itself before now:
- LLD and binutils-2.26 only relax
movq foo@GOTPCREL(%rip), %reg
to
leaq foo(%rip), %reg
which is still position-independent, rather than
mov $foo, %reg
which is permitted by the psABI when -pie is not enabled.
- GCC happens to only generate GOTPCREL relocations on mov instructions.
- CLang does generate GOTPCREL relocations on non-mov instructions, but
when building the compressed kernel, it uses its integrated assembler
(due to the redefinition of KBUILD_CFLAGS dropping -no-integrated-as),
which has so far defaulted to not generating the GOTPCRELX
relocations.
Nick Desaulniers reports [1,2]:
"A recent change [3] to a default value of configuration variable
(ENABLE_X86_RELAX_RELOCATIONS OFF -> ON) in LLVM now causes Clang's
integrated assembler to emit R_X86_64_GOTPCRELX/R_X86_64_REX_GOTPCRELX
relocations. LLD will relax instructions with these relocations based
on whether the image is being linked as position independent or not.
When not, then LLD will relax these instructions to use absolute
addressing mode (R_RELAX_GOT_PC_NOPIC). This causes kernels built with
Clang and linked with LLD to fail to boot."
Patch series [4] is a solution to allow the compressed kernel to be
linked with -pie unconditionally, but even if merged is unlikely to be
backported. As a simple solution that can be applied to stable as well,
prevent the assembler from generating the relaxed relocation types using
the -mrelax-relocations=no option. For ease of backporting, do this
unconditionally.
[0] https://gitlab.com/x86-psABIs/x86-64-ABI/-/blob/master/x86-64-ABI/linker-optimization.tex#L65
[1] https://lore.kernel.org/lkml/20200807194100.3570838-1-ndesaulniers@google.com/
[2] https://github.com/ClangBuiltLinux/linux/issues/1121
[3] https://reviews.llvm.org/rGc41a18cf61790fc898dcda1055c3efbf442c14c0
[4] https://lore.kernel.org/lkml/20200731202738.2577854-1-nivedita@alum.mit.edu/
Reported-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Nick Desaulniers <ndesaulniers@google.com>
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200812004308.1448603-1-nivedita@alum.mit.edu
- Multiple stolen time fixes, with a new capability to match x86
- Fix for hugetlbfs mappings when PUD and PMD are the same level
- Fix for hugetlbfs mappings when PTE mappings are enforced
(dirty logging, for example)
- Fix tracing output of 64bit values
x86:
- nSVM state restore fixes
- Async page fault fixes
- Lots of small fixes everywhere
-----BEGIN PGP SIGNATURE-----
iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAl9dM5kUHHBib256aW5p
QHJlZGhhdC5jb20ACgkQv/vSX3jHroM+Iwf+LbISO7ccpPMK1kKtOeug/jZv+xQA
sVaBGRzYo+k2e0XtV8E8IV4N30FBtYSwXsbBKkMAoy2FpmMebgDWDQ7xspb6RJMS
/y8t1iqPwdOaLIkUkgc7UihSTlZm05Es3f3q6uZ9+oaM4Fe+V7xWzTUX4Oy89JO7
KcQsTD7pMqS4bfZGADK781ITR/WPgCi0aYx5s6dcwcZAQXhb1K1UKEjB8OGKnjUh
jliReJtxRA16rjF+S5aJ7L07Ce/ksrfwkI4NXJ4GxW+lyOfVNdSBJUBaZt1m7G2M
1We5+i5EjKCjuxmgtUUUfVdazpj1yl+gBGT7KKkLte9T9WZdXyDnixAbvg==
=OFb3
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm fixes from Paolo Bonzini:
"A bit on the bigger side, mostly due to me being on vacation, then
busy, then on parental leave, but there's nothing worrisome.
ARM:
- Multiple stolen time fixes, with a new capability to match x86
- Fix for hugetlbfs mappings when PUD and PMD are the same level
- Fix for hugetlbfs mappings when PTE mappings are enforced (dirty
logging, for example)
- Fix tracing output of 64bit values
x86:
- nSVM state restore fixes
- Async page fault fixes
- Lots of small fixes everywhere"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (25 commits)
KVM: emulator: more strict rsm checks.
KVM: nSVM: more strict SMM checks when returning to nested guest
SVM: nSVM: setup nested msr permission bitmap on nested state load
SVM: nSVM: correctly restore GIF on vmexit from nesting after migration
x86/kvm: don't forget to ACK async PF IRQ
x86/kvm: properly use DEFINE_IDTENTRY_SYSVEC() macro
KVM: VMX: Don't freeze guest when event delivery causes an APIC-access exit
KVM: SVM: avoid emulation with stale next_rip
KVM: x86: always allow writing '0' to MSR_KVM_ASYNC_PF_EN
KVM: SVM: Periodically schedule when unregistering regions on destroy
KVM: MIPS: Change the definition of kvm type
kvm x86/mmu: use KVM_REQ_MMU_SYNC to sync when needed
KVM: nVMX: Fix the update value of nested load IA32_PERF_GLOBAL_CTRL control
KVM: fix memory leak in kvm_io_bus_unregister_dev()
KVM: Check the allocation of pv cpu mask
KVM: nVMX: Update VMCS02 when L2 PAE PDPTE updates detected
KVM: arm64: Update page shift if stage 2 block mapping not supported
KVM: arm64: Fix address truncation in traces
KVM: arm64: Do not try to map PUDs when they are folded into PMD
arm64/x86: KVM: Introduce steal-time cap
...
Don't ignore return values in rsm_load_state_64/32 to avoid
loading invalid state from SMM state area if it was tampered with
by the guest.
This is primarly intended to avoid letting guest set bits in EFER
(like EFER.SVME when nesting is disabled) by manipulating SMM save area.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20200827171145.374620-8-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
* check that guest is 64 bit guest, otherwise the SVM related fields
in the smm state area are not defined
* If the SMM area indicates that SMM interrupted a running guest,
check that EFER.SVME which is also saved in this area is set, otherwise
the guest might have tampered with SMM save area, and so indicate
emulation failure which should triple fault the guest.
* Check that that guest CPUID supports SVM (due to the same issue as above)
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20200827162720.278690-4-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This code was missing and was forcing the L2 run with L1's msr
permission bitmap
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20200827162720.278690-3-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Currently code in svm_set_nested_state copies the current vmcb control
area to L1 control area (hsave->control), under assumption that
it mostly reflects the defaults that kvm choose, and later qemu
overrides these defaults with L2 state using standard KVM interfaces,
like KVM_SET_REGS.
However nested GIF (which is AMD specific thing) is by default is true,
and it is copied to hsave area as such.
This alone is not a big deal since on VMexit, GIF is always set to false,
regardless of what it was on VM entry. However in nested_svm_vmexit we
were first were setting GIF to false, but then we overwrite the control
fields with value from the hsave area. (including the nested GIF field
itself if GIF virtualization is enabled).
Now on normal vm entry this is not a problem, since GIF is usually false
prior to normal vm entry, and this is the value that copied to hsave,
and then restored, but this is not always the case when the nested state
is loaded as explained above.
To fix this issue, move svm_set_gif after we restore the L1 control
state in nested_svm_vmexit, so that even with wrong GIF in the
saved L1 control area, we still clear GIF as the spec says.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20200827162720.278690-2-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Merge commit 26d05b368a ("Merge branch 'kvm-async-pf-int' into HEAD")
tried to adapt the new interrupt based async PF mechanism to the newly
introduced IDTENTRY magic but unfortunately it missed the fact that
DEFINE_IDTENTRY_SYSVEC() doesn't call ack_APIC_irq() on its own and
all DEFINE_IDTENTRY_SYSVEC() users have to call it manually.
As the result all multi-CPU KVM guest hang on boot when
KVM_FEATURE_ASYNC_PF_INT is present. The breakage went unnoticed because no
KVM userspace (e.g. QEMU) currently set it (and thus async PF mechanism
is currently disabled) but we're about to change that.
Fixes: 26d05b368a ("Merge branch 'kvm-async-pf-int' into HEAD")
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Message-Id: <20200908135350.355053-3-vkuznets@redhat.com>
Tested-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
According to SDM 27.2.4, Event delivery causes an APIC-access VM exit.
Don't report internal error and freeze guest when event delivery causes
an APIC-access exit, it is handleable and the event will be re-injected
during the next vmentry.
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
Message-Id: <1597827327-25055-2-git-send-email-wanpengli@tencent.com>
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
svm->next_rip is reset in svm_vcpu_run() only after calling
svm_exit_handlers_fastpath(), which will cause SVM's
skip_emulated_instruction() to write a stale RIP.
We can move svm_exit_handlers_fastpath towards the end of
svm_vcpu_run(). To align VMX with SVM, keep svm_complete_interrupts()
close as well.
Suggested-by: Sean Christopherson <sean.j.christopherson@intel.com>
Cc: Paul K. <kronenpj@kronenpj.dyndns.org>
Reviewed-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
[Also move vmcb_mark_all_clean before any possible write to the VMCB.
- Paolo]
Even without in-kernel LAPIC we should allow writing '0' to
MSR_KVM_ASYNC_PF_EN as we're not enabling the mechanism. In
particular, QEMU with 'kernel-irqchip=off' fails to start
a guest with
qemu-system-x86_64: error: failed to set MSR 0x4b564d02 to 0x0
Fixes: 9d3c447c72 ("KVM: X86: Fix async pf caused null-ptr-deref")
Reported-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Message-Id: <20200911093147.484565-1-vkuznets@redhat.com>
[Actually commit the version proposed by Sean Christopherson. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
There may be many encrypted regions that need to be unregistered when a
SEV VM is destroyed. This can lead to soft lockups. For example, on a
host running 4.15:
watchdog: BUG: soft lockup - CPU#206 stuck for 11s! [t_virtual_machi:194348]
CPU: 206 PID: 194348 Comm: t_virtual_machi
RIP: 0010:free_unref_page_list+0x105/0x170
...
Call Trace:
[<0>] release_pages+0x159/0x3d0
[<0>] sev_unpin_memory+0x2c/0x50 [kvm_amd]
[<0>] __unregister_enc_region_locked+0x2f/0x70 [kvm_amd]
[<0>] svm_vm_destroy+0xa9/0x200 [kvm_amd]
[<0>] kvm_arch_destroy_vm+0x47/0x200
[<0>] kvm_put_kvm+0x1a8/0x2f0
[<0>] kvm_vm_release+0x25/0x30
[<0>] do_exit+0x335/0xc10
[<0>] do_group_exit+0x3f/0xa0
[<0>] get_signal+0x1bc/0x670
[<0>] do_signal+0x31/0x130
Although the CLFLUSH is no longer issued on every encrypted region to be
unregistered, there are no other changes that can prevent soft lockups for
very large SEV VMs in the latest kernel.
Periodically schedule if necessary. This still holds kvm->lock across the
resched, but since this only happens when the VM is destroyed this is
assumed to be acceptable.
Signed-off-by: David Rientjes <rientjes@google.com>
Message-Id: <alpine.DEB.2.23.453.2008251255240.2987727@chino.kir.corp.google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When kvm_mmu_get_page() gets a page with unsynced children, the spt
pagetable is unsynchronized with the guest pagetable. But the
guest might not issue a "flush" operation on it when the pagetable
entry is changed from zero or other cases. The hypervisor has the
responsibility to synchronize the pagetables.
KVM behaved as above for many years, But commit 8c8560b833
("KVM: x86/mmu: Use KVM_REQ_TLB_FLUSH_CURRENT for MMU specific flushes")
inadvertently included a line of code to change it without giving any
reason in the changelog. It is clear that the commit's intention was to
change KVM_REQ_TLB_FLUSH -> KVM_REQ_TLB_FLUSH_CURRENT, so we don't
needlessly flush other contexts; however, one of the hunks changed
a nearby KVM_REQ_MMU_SYNC instead. This patch changes it back.
Link: https://lore.kernel.org/lkml/20200320212833.3507-26-sean.j.christopherson@intel.com/
Cc: Sean Christopherson <sean.j.christopherson@intel.com>
Cc: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Lai Jiangshan <laijs@linux.alibaba.com>
Message-Id: <20200902135421.31158-1-jiangshanlai@gmail.com>
fixes: 8c8560b833 ("KVM: x86/mmu: Use KVM_REQ_TLB_FLUSH_CURRENT for MMU specific flushes")
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
A minor fix for the update of VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL field
in exit_ctls_high.
Fixes: 03a8871add ("KVM: nVMX: Expose load IA32_PERF_GLOBAL_CTRL
VM-{Entry,Exit} control")
Signed-off-by: Chenyi Qiang <chenyi.qiang@intel.com>
Reviewed-by: Xiaoyao Li <xiaoyao.li@intel.com>
Message-Id: <20200828085622.8365-5-chenyi.qiang@intel.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
check the allocation of per-cpu __pv_cpu_mask. Initialize ops only when
successful.
Signed-off-by: Haiwei Li <lihaiwei@tencent.com>
Message-Id: <d59f05df-e6d3-3d31-a036-cc25a2b2f33f@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When L2 uses PAE, L0 intercepts of L2 writes to CR0/CR3/CR4 call
load_pdptrs to read the possibly updated PDPTEs from the guest
physical address referenced by CR3. It loads them into
vcpu->arch.walk_mmu->pdptrs and sets VCPU_EXREG_PDPTR in
vcpu->arch.regs_dirty.
At the subsequent assumed reentry into L2, the mmu will call
vmx_load_mmu_pgd which calls ept_load_pdptrs. ept_load_pdptrs sees
VCPU_EXREG_PDPTR set in vcpu->arch.regs_dirty and loads
VMCS02.GUEST_PDPTRn from vcpu->arch.walk_mmu->pdptrs[]. This all works
if the L2 CRn write intercept always resumes L2.
The resume path calls vmx_check_nested_events which checks for
exceptions, MTF, and expired VMX preemption timers. If
vmx_check_nested_events finds any of these conditions pending it will
reflect the corresponding exit into L1. Live migration at this point
would also cause a missed immediate reentry into L2.
After L1 exits, vmx_vcpu_run calls vmx_register_cache_reset which
clears VCPU_EXREG_PDPTR in vcpu->arch.regs_dirty. When L2 next
resumes, ept_load_pdptrs finds VCPU_EXREG_PDPTR clear in
vcpu->arch.regs_dirty and does not load VMCS02.GUEST_PDPTRn from
vcpu->arch.walk_mmu->pdptrs[]. prepare_vmcs02 will then load
VMCS02.GUEST_PDPTRn from vmcs12->pdptr0/1/2/3 which contain the stale
values stored at last L2 exit. A repro of this bug showed L2 entering
triple fault immediately due to the bad VMCS02.GUEST_PDPTRn values.
When L2 is in PAE paging mode add a call to ept_load_pdptrs before
leaving L2. This will update VMCS02.GUEST_PDPTRn if they are dirty in
vcpu->arch.walk_mmu->pdptrs[].
Tested:
kvm-unit-tests with new directed test: vmx_mtf_pdpte_test.
Verified that test fails without the fix.
Also ran Google internal VMM with an Ubuntu 16.04 4.4.0-83 guest running a
custom hypervisor with a 32-bit Windows XP L2 guest using PAE. Prior to fix
would repro readily. Ran 14 simultaneous L2s for 140 iterations with no
failures.
Signed-off-by: Peter Shier <pshier@google.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
Message-Id: <20200820230545.2411347-1-pshier@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
- Multiple stolen time fixes, with a new capability to match x86
- Fix for hugetlbfs mappings when PUD and PMD are the same level
- Fix for hugetlbfs mappings when PTE mappings are enforced
(dirty logging, for example)
- Fix tracing output of 64bit values
-----BEGIN PGP SIGNATURE-----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=VBrL
-----END PGP SIGNATURE-----
Merge tag 'kvmarm-fixes-5.9-1' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD
KVM/arm64 fixes for Linux 5.9, take #1
- Multiple stolen time fixes, with a new capability to match x86
- Fix for hugetlbfs mappings when PUD and PMD are the same level
- Fix for hugetlbfs mappings when PTE mappings are enforced
(dirty logging, for example)
- Fix tracing output of 64bit values
A recent refresh of the defconfigs got rid of the following
(unset) config:
# CONFIG_64BIT is not set
Innocuous as it seems, when the config file is saved again the
behavior is changed so that CONFIG_64BIT=y.
Currently,
$ make i386_defconfig
$ grep CONFIG_64BIT .config
CONFIG_64BIT=y
whereas previously (and with this patch):
$ make i386_defconfig
$ grep CONFIG_64BIT .config
# CONFIG_64BIT is not set
( This was found with weird compiler errors on OpenEmbedded
builds, as the compiler was unable to cope with 64-bits data
types. )
Fixes: 1d0e12fd3a ("x86/defconfigs: Refresh defconfig files")
Reported-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>
Reported-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
We got slightly different patches removing a double word
in a comment in net/ipv4/raw.c - picked the version from net.
Simple conflict in drivers/net/ethernet/ibm/ibmvnic.c. Use cached
values instead of VNIC login response buffer (following what
commit 507ebe6444 ("ibmvnic: Fix use-after-free of VNIC login
response buffer") did).
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Andy reported that the syscall treacing for 32bit fast syscall fails:
# ./tools/testing/selftests/x86/ptrace_syscall_32
...
[RUN] SYSEMU
[FAIL] Initial args are wrong (nr=224, args=10 11 12 13 14 4289172732)
...
[RUN] SYSCALL
[FAIL] Initial args are wrong (nr=29, args=0 0 0 0 0 4289172732)
The eason is that the conversion to generic entry code moved the retrieval
of the sixth argument (EBP) after the point where the syscall entry work
runs, i.e. ptrace, seccomp, audit...
Unbreak it by providing a split up version of syscall_enter_from_user_mode().
- syscall_enter_from_user_mode_prepare() establishes state and enables
interrupts
- syscall_enter_from_user_mode_work() runs the entry work
Replace the call to syscall_enter_from_user_mode() in the 32bit fast
syscall C-entry with the split functions and stick the EBP retrieval
between them.
Fixes: 27d6b4d14f ("x86/entry: Use generic syscall entry function")
Reported-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/87k0xdjbtt.fsf@nanos.tec.linutronix.de
Trying to clear DR7 around a #DB from usermode malfunctions if the tasks
schedules when delivering SIGTRAP.
Rather than trying to define a special no-recursion region, just allow a
single level of recursion. The same mechanism is used for NMI, and it
hasn't caused any problems yet.
Fixes: 9f58fdde95 ("x86/db: Split out dr6/7 handling")
Reported-by: Kyle Huey <me@kylehuey.com>
Debugged-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Daniel Thompson <daniel.thompson@linaro.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/8b9bd05f187231df008d48cf818a6a311cbd5c98.1597882384.git.luto@kernel.org
Link: https://lore.kernel.org/r/20200902133200.726584153@infradead.org
The WARN added in commit 3c73b81a91 ("x86/entry, selftests: Further
improve user entry sanity checks") unconditionally triggers on a IVB
machine because it does not support SMAP.
For !SMAP hardware the CLAC/STAC instructions are patched out and thus if
userspace sets AC, it is still have set after entry.
Fixes: 3c73b81a91 ("x86/entry, selftests: Further improve user entry sanity checks")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Daniel Thompson <daniel.thompson@linaro.org>
Acked-by: Andy Lutomirski <luto@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200902133200.666781610@infradead.org
On i386, the order of parameters passed on regs is eax,edx,and ecx
(as per regparm(3) calling conventions).
Change the mapping in regs_get_kernel_argument(), so that arg1=ax
arg2=dx, and arg3=cx.
Running the selftests testcase kprobes_args_use.tc shows the result
as passed.
Fixes: 3c88ee194c ("x86: ptrace: Add function argument access API")
Signed-off-by: Vamshi K Sthambamkadi <vamshi.k.sthambamkadi@gmail.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Masami Hiramatsu <mhiramat@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@vger.kernel.org>
Link: https://lkml.kernel.org/r/20200828113242.GA1424@cosmos
One can not simply remove vmalloc faulting on x86-32. Upstream
commit: 7f0a002b5a ("x86/mm: remove vmalloc faulting")
removed it on x86 alltogether because previously the
arch_sync_kernel_mappings() interface was introduced. This interface
added synchronization of vmalloc/ioremap page-table updates to all
page-tables in the system at creation time and was thought to make
vmalloc faulting obsolete.
But that assumption was incredibly naive.
It turned out that there is a race window between the time the vmalloc
or ioremap code establishes a mapping and the time it synchronizes
this change to other page-tables in the system.
During this race window another CPU or thread can establish a vmalloc
mapping which uses the same intermediate page-table entries (e.g. PMD
or PUD) and does no synchronization in the end, because it found all
necessary mappings already present in the kernel reference page-table.
But when these intermediate page-table entries are not yet
synchronized, the other CPU or thread will continue with a vmalloc
address that is not yet mapped in the page-table it currently uses,
causing an unhandled page fault and oops like below:
BUG: unable to handle page fault for address: fe80c000
#PF: supervisor write access in kernel mode
#PF: error_code(0x0002) - not-present page
*pde = 33183067 *pte = a8648163
Oops: 0002 [#1] SMP
CPU: 1 PID: 13514 Comm: cve-2017-17053 Tainted: G
...
Call Trace:
ldt_dup_context+0x66/0x80
dup_mm+0x2b3/0x480
copy_process+0x133b/0x15c0
_do_fork+0x94/0x3e0
__ia32_sys_clone+0x67/0x80
__do_fast_syscall_32+0x3f/0x70
do_fast_syscall_32+0x29/0x60
do_SYSENTER_32+0x15/0x20
entry_SYSENTER_32+0x9f/0xf2
EIP: 0xb7eef549
So the arch_sync_kernel_mappings() interface is racy, but removing it
would mean to re-introduce the vmalloc_sync_all() interface, which is
even more awful. Keep arch_sync_kernel_mappings() in place and catch
the race condition in the page-fault handler instead.
Do a partial revert of above commit to get vmalloc faulting on x86-32
back in place.
Fixes: 7f0a002b5a ("x86/mm: remove vmalloc faulting")
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200902155904.17544-1-joro@8bytes.org
When CONFIG_RETPOLINE is disabled, Clang uses a jump table for the
switch statement in cmdline_find_option (jump tables are disabled when
CONFIG_RETPOLINE is enabled). This function is called very early in boot
from sme_enable() if CONFIG_AMD_MEM_ENCRYPT is enabled. At this time,
the kernel is still executing out of the identity mapping, but the jump
table will contain virtual addresses.
Fix this by disabling jump tables for cmdline.c when AMD_MEM_ENCRYPT is
enabled.
Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200903023056.3914690-1-nivedita@alum.mit.edu
Daniel Borkmann says:
====================
pull-request: bpf-next 2020-09-01
The following pull-request contains BPF updates for your *net-next* tree.
There are two small conflicts when pulling, resolve as follows:
1) Merge conflict in tools/lib/bpf/libbpf.c between 88a8212028 ("libbpf: Factor
out common ELF operations and improve logging") in bpf-next and 1e891e513e
("libbpf: Fix map index used in error message") in net-next. Resolve by taking
the hunk in bpf-next:
[...]
scn = elf_sec_by_idx(obj, obj->efile.btf_maps_shndx);
data = elf_sec_data(obj, scn);
if (!scn || !data) {
pr_warn("elf: failed to get %s map definitions for %s\n",
MAPS_ELF_SEC, obj->path);
return -EINVAL;
}
[...]
2) Merge conflict in drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c between
9647c57b11 ("xsk: i40e: ice: ixgbe: mlx5: Test for dma_need_sync earlier for
better performance") in bpf-next and e20f0dbf20 ("net/mlx5e: RX, Add a prefetch
command for small L1_CACHE_BYTES") in net-next. Resolve the two locations by retaining
net_prefetch() and taking xsk_buff_dma_sync_for_cpu() from bpf-next. Should look like:
[...]
xdp_set_data_meta_invalid(xdp);
xsk_buff_dma_sync_for_cpu(xdp, rq->xsk_pool);
net_prefetch(xdp->data);
[...]
We've added 133 non-merge commits during the last 14 day(s) which contain
a total of 246 files changed, 13832 insertions(+), 3105 deletions(-).
The main changes are:
1) Initial support for sleepable BPF programs along with bpf_copy_from_user() helper
for tracing to reliably access user memory, from Alexei Starovoitov.
2) Add BPF infra for writing and parsing TCP header options, from Martin KaFai Lau.
3) bpf_d_path() helper for returning full path for given 'struct path', from Jiri Olsa.
4) AF_XDP support for shared umems between devices and queues, from Magnus Karlsson.
5) Initial prep work for full BPF-to-BPF call support in libbpf, from Andrii Nakryiko.
6) Generalize bpf_sk_storage map & add local storage for inodes, from KP Singh.
7) Implement sockmap/hash updates from BPF context, from Lorenz Bauer.
8) BPF xor verification for scalar types & add BPF link iterator, from Yonghong Song.
9) Use target's prog type for BPF_PROG_TYPE_EXT prog verification, from Udip Pant.
10) Rework BPF tracing samples to use libbpf loader, from Daniel T. Lee.
11) Fix xdpsock sample to really cycle through all buffers, from Weqaar Janjua.
12) Improve type safety for tun/veth XDP frame handling, from Maciej Żenczykowski.
13) Various smaller cleanups and improvements all over the place.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
- Move disabling of the local APIC after invoking fixup_irqs() to ensure
that interrupts which are incoming are noted in the IRR and not ignored.
- Unbreak affinity setting. The rework of the entry code reused the
regular exception entry code for device interrupts. The vector number is
pushed into the errorcode slot on the stack which is then lifted into an
argument and set to -1 because that's regs->orig_ax which is used in
quite some places to check whether the entry came from a syscall. But it
was overlooked that orig_ax is used in the affinity cleanup code to
validate whether the interrupt has arrived on the new target. It turned
out that this vector check is pointless because interrupts are never
moved from one vector to another on the same CPU. That check is a
historical leftover from the time where x86 supported multi-CPU
affinities, but not longer needed with the now strict single CPU
affinity. Famous last words ...
- Add a missing check for an empty cpumask into the matrix allocator. The
affinity change added a warning to catch the case where an interrupt is
moved on the same CPU to a different vector. This triggers because a
condition with an empty cpumask returns an assignment from the allocator
as the allocator uses for_each_cpu() without checking the cpumask for
being empty. The historical inconsistent for_each_cpu() behaviour of
ignoring the cpumask and unconditionally claiming that CPU0 is in the
mask striked again. Sigh.
-----BEGIN PGP SIGNATURE-----
iQJHBAABCgAxFiEEQp8+kY+LLUocC4bMphj1TA10mKEFAl9L6WYTHHRnbHhAbGlu
dXRyb25peC5kZQAKCRCmGPVMDXSYoRV5D/9dRq/4pn5g1esnzm4GhIr2To3Qp6cl
s7VswTdN8FmWBqVz79ZVYqj663UpL3pPY1np01ctrxRQLeDVfWcI2BMR5irnny8h
otORhFysuDUl+yfuomWVbzfQQNJ+VeQVeWKD3cIhD1I3sXqDX5Wpa8n086hYKQXx
eutVC3+JdzJZFm68xarlLW7h2f1au1eZZFgVnyY+J5KO9Dwm63a4RITdDVk7KV4t
uKEDza5P9SY+kE9LAGNq8BAEObf9FeMXw0mRM7atRKVsJQQGVk6bgiuaRr01w1+W
hQCPx/3g6PHFnGgx/KQgHf1jgrZFhXOyIDo6ZeFy+SJGIZRB3n8o5Kjns2l8Pa+K
2qy1TRoZIsGkwGCi/BM6viLzBikbh/gnGYy/8KTEJdKs8P3ZKHUZVSAB1dpapOWX
4n+rKoVPnvxgRSeZZo+tgLkvUdh+/9Huyr9vHiYjtbbB8tFvjlkOmrZ6sirHByDy
jg6TjOJVb1CC/PoW4M7JNfmeKvHQnTACwH6djdVGDLPJspuUsYkPI0Uk0CX21SA3
45Tuylvl9jT6+vq95Av2RbAiipmSpZ/O1NHV8Paf466SKmhUgG3lv5PHh3xTm1U2
Be/RbJ75x4Muuw42ttU1LcpcLPcOZRQNEREoNd5UysgYYgWRekBvU+ZRQNW4g2nw
3JDgJgm0iBUN9w==
=zIi4
-----END PGP SIGNATURE-----
Merge tag 'x86-urgent-2020-08-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Thomas Gleixner:
"Three interrupt related fixes for X86:
- Move disabling of the local APIC after invoking fixup_irqs() to
ensure that interrupts which are incoming are noted in the IRR and
not ignored.
- Unbreak affinity setting.
The rework of the entry code reused the regular exception entry
code for device interrupts. The vector number is pushed into the
errorcode slot on the stack which is then lifted into an argument
and set to -1 because that's regs->orig_ax which is used in quite
some places to check whether the entry came from a syscall.
But it was overlooked that orig_ax is used in the affinity cleanup
code to validate whether the interrupt has arrived on the new
target. It turned out that this vector check is pointless because
interrupts are never moved from one vector to another on the same
CPU. That check is a historical leftover from the time where x86
supported multi-CPU affinities, but not longer needed with the now
strict single CPU affinity. Famous last words ...
- Add a missing check for an empty cpumask into the matrix allocator.
The affinity change added a warning to catch the case where an
interrupt is moved on the same CPU to a different vector. This
triggers because a condition with an empty cpumask returns an
assignment from the allocator as the allocator uses for_each_cpu()
without checking the cpumask for being empty. The historical
inconsistent for_each_cpu() behaviour of ignoring the cpumask and
unconditionally claiming that CPU0 is in the mask struck again.
Sigh.
plus a new entry into the MAINTAINER file for the HPE/UV platform"
* tag 'x86-urgent-2020-08-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq/matrix: Deal with the sillyness of for_each_cpu() on UP
x86/irq: Unbreak interrupt affinity setting
x86/hotplug: Silence APIC only after all interrupts are migrated
MAINTAINERS: Add entry for HPE Superdome Flex (UV) maintainers
- Prevent recursion by using raw_cpu_* operations
- Fixup the interrupt state in the cpu idle code to be consistent
- Push rcu_idle_enter/exit() invocations deeper into the idle path so
that the lock operations are inside the RCU watching sections
- Move trace_cpu_idle() into generic code so it's called before RCU goes
idle.
- Handle raw_local_irq* vs. local_irq* operations correctly
- Move the tracepoints out from under the lockdep recursion handling
which turned out to be fragile and inconsistent.
-----BEGIN PGP SIGNATURE-----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=Hgt6
-----END PGP SIGNATURE-----
Merge tag 'locking-urgent-2020-08-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull locking fixes from Thomas Gleixner:
"A set of fixes for lockdep, tracing and RCU:
- Prevent recursion by using raw_cpu_* operations
- Fixup the interrupt state in the cpu idle code to be consistent
- Push rcu_idle_enter/exit() invocations deeper into the idle path so
that the lock operations are inside the RCU watching sections
- Move trace_cpu_idle() into generic code so it's called before RCU
goes idle.
- Handle raw_local_irq* vs. local_irq* operations correctly
- Move the tracepoints out from under the lockdep recursion handling
which turned out to be fragile and inconsistent"
* tag 'locking-urgent-2020-08-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
lockdep,trace: Expose tracepoints
lockdep: Only trace IRQ edges
mips: Implement arch_irqs_disabled()
arm64: Implement arch_irqs_disabled()
nds32: Implement arch_irqs_disabled()
locking/lockdep: Cleanup
x86/entry: Remove unused THUNKs
cpuidle: Move trace_cpu_idle() into generic code
cpuidle: Make CPUIDLE_FLAG_TLB_FLUSHED generic
sched,idle,rcu: Push rcu_idle deeper into the idle path
cpuidle: Fixup IRQ state
lockdep: Use raw_cpu_*() for per-cpu variables
Introduce sleepable BPF programs that can request such property for themselves
via BPF_F_SLEEPABLE flag at program load time. In such case they will be able
to use helpers like bpf_copy_from_user() that might sleep. At present only
fentry/fexit/fmod_ret and lsm programs can request to be sleepable and only
when they are attached to kernel functions that are known to allow sleeping.
The non-sleepable programs are relying on implicit rcu_read_lock() and
migrate_disable() to protect life time of programs, maps that they use and
per-cpu kernel structures used to pass info between bpf programs and the
kernel. The sleepable programs cannot be enclosed into rcu_read_lock().
migrate_disable() maps to preempt_disable() in non-RT kernels, so the progs
should not be enclosed in migrate_disable() as well. Therefore
rcu_read_lock_trace is used to protect the life time of sleepable progs.
There are many networking and tracing program types. In many cases the
'struct bpf_prog *' pointer itself is rcu protected within some other kernel
data structure and the kernel code is using rcu_dereference() to load that
program pointer and call BPF_PROG_RUN() on it. All these cases are not touched.
Instead sleepable bpf programs are allowed with bpf trampoline only. The
program pointers are hard-coded into generated assembly of bpf trampoline and
synchronize_rcu_tasks_trace() is used to protect the life time of the program.
The same trampoline can hold both sleepable and non-sleepable progs.
When rcu_read_lock_trace is held it means that some sleepable bpf program is
running from bpf trampoline. Those programs can use bpf arrays and preallocated
hash/lru maps. These map types are waiting on programs to complete via
synchronize_rcu_tasks_trace();
Updates to trampoline now has to do synchronize_rcu_tasks_trace() and
synchronize_rcu_tasks() to wait for sleepable progs to finish and for
trampoline assembly to finish.
This is the first step of introducing sleepable progs. Eventually dynamically
allocated hash maps can be allowed and networking program types can become
sleepable too.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Acked-by: KP Singh <kpsingh@google.com>
Link: https://lore.kernel.org/bpf/20200827220114.69225-3-alexei.starovoitov@gmail.com
Several people reported that 5.8 broke the interrupt affinity setting
mechanism.
The consolidation of the entry code reused the regular exception entry code
for device interrupts and changed the way how the vector number is conveyed
from ptregs->orig_ax to a function argument.
The low level entry uses the hardware error code slot to push the vector
number onto the stack which is retrieved from there into a function
argument and the slot on stack is set to -1.
The reason for setting it to -1 is that the error code slot is at the
position where pt_regs::orig_ax is. A positive value in pt_regs::orig_ax
indicates that the entry came via a syscall. If it's not set to a negative
value then a signal delivery on return to userspace would try to restart a
syscall. But there are other places which rely on pt_regs::orig_ax being a
valid indicator for syscall entry.
But setting pt_regs::orig_ax to -1 has a nasty side effect vs. the
interrupt affinity setting mechanism, which was overlooked when this change
was made.
Moving interrupts on x86 happens in several steps. A new vector on a
different CPU is allocated and the relevant interrupt source is
reprogrammed to that. But that's racy and there might be an interrupt
already in flight to the old vector. So the old vector is preserved until
the first interrupt arrives on the new vector and the new target CPU. Once
that happens the old vector is cleaned up, but this cleanup still depends
on the vector number being stored in pt_regs::orig_ax, which is now -1.
That -1 makes the check for cleanup: pt_regs::orig_ax == new_vector
always false. As a consequence the interrupt is moved once, but then it
cannot be moved anymore because the cleanup of the old vector never
happens.
There would be several ways to convey the vector information to that place
in the guts of the interrupt handling, but on deeper inspection it turned
out that this check is pointless and a leftover from the old affinity model
of X86 which supported multi-CPU affinities. Under this model it was
possible that an interrupt had an old and a new vector on the same CPU, so
the vector match was required.
Under the new model the effective affinity of an interrupt is always a
single CPU from the requested affinity mask. If the affinity mask changes
then either the interrupt stays on the CPU and on the same vector when that
CPU is still in the new affinity mask or it is moved to a different CPU, but
it is never moved to a different vector on the same CPU.
Ergo the cleanup check for the matching vector number is not required and
can be removed which makes the dependency on pt_regs:orig_ax go away.
The remaining check for new_cpu == smp_processsor_id() is completely
sufficient. If it matches then the interrupt was successfully migrated and
the cleanup can proceed.
For paranoia sake add a warning into the vector assignment code to
validate that the assumption of never moving to a different vector on
the same CPU holds.
Fixes: 633260fa14 ("x86/irq: Convey vector as argument and not in ptregs")
Reported-by: Alex bykov <alex.bykov@scylladb.com>
Reported-by: Avi Kivity <avi@scylladb.com>
Reported-by: Alexander Graf <graf@amazon.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Alexander Graf <graf@amazon.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/87wo1ltaxz.fsf@nanos.tec.linutronix.de
There is a race when taking a CPU offline. Current code looks like this:
native_cpu_disable()
{
...
apic_soft_disable();
/*
* Any existing set bits for pending interrupt to
* this CPU are preserved and will be sent via IPI
* to another CPU by fixup_irqs().
*/
cpu_disable_common();
{
....
/*
* Race window happens here. Once local APIC has been
* disabled any new interrupts from the device to
* the old CPU are lost
*/
fixup_irqs(); // Too late to capture anything in IRR.
...
}
}
The fix is to disable the APIC *after* cpu_disable_common().
Testing was done with a USB NIC that provided a source of frequent
interrupts. A script migrated interrupts to a specific CPU and
then took that CPU offline.
Fixes: 60dcaad573 ("x86/hotplug: Silence APIC and NMI when CPU is dead")
Reported-by: Evan Green <evgreen@chromium.org>
Signed-off-by: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Tested-by: Evan Green <evgreen@chromium.org>
Reviewed-by: Evan Green <evgreen@chromium.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/lkml/875zdarr4h.fsf@nanos.tec.linutronix.de/
Link: https://lore.kernel.org/r/1598501530-45821-1-git-send-email-ashok.raj@intel.com
Unused remnants
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Marco Elver <elver@google.com>
Link: https://lkml.kernel.org/r/20200821085348.487040689@infradead.org
Remove trace_cpu_idle() from the arch_cpu_idle() implementations and
put it in the generic code, right before disabling RCU. Gets rid of
more trace_*_rcuidle() users.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Marco Elver <elver@google.com>
Link: https://lkml.kernel.org/r/20200821085348.428433395@infradead.org
This allows moving the leave_mm() call into generic code before
rcu_idle_enter(). Gets rid of more trace_*_rcuidle() users.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Marco Elver <elver@google.com>
Link: https://lkml.kernel.org/r/20200821085348.369441600@infradead.org
