In other words, we have:
MAX_NUM_VIDEO_DELAY_FRAMES = 3
MAX_NUM_DELAY_FRAMES = 2
The MAX_NUM_DELAY_FRAMES macro is used only only when allocating
memory. On all other parts, including looping over such array,
MAX_NUM_VIDEO_DELAY_FRAMES is used instead, like:
void sh_css_binary_args_reset(struct sh_css_binary_args *args)
{
unsigned int i;
...
for (i = 0; i < MAX_NUM_VIDEO_DELAY_FRAMES; i++)
args->delay_frames[i] = NULL;
Which will cause buffer overflows, with may override the next array
(tnr_frames[]).
In practice, this may not be causing real issues, as the code
checks for num_delay_frames on some parts (but not everywhere).
Some arguments for tnf and ref settings are meant to be const, but
they're defined without such annotation. Due to that, there's an
ugly cast at sh_css_sp.c.
media: atomisp: get rid of some old broken debug code
It sounds that someone once changed the debug level at compile
time for some testing, but forgot to remove the legacy code after
finishing debuging it.
media: atomisp: don't flood dmesg with -EAGAIN return codes
Using DQBUF on non-blocking mode will return -EAGAIN
if nothing arrives. Printing it has no value, even for debug
purposes. So, only display real return codes.
Flooding the dmesg with lots of messages per second. The same
pattern happens for all other IRQs.
Change the logic for printing just one message per IRQ and
rate-limit those, as, for debugging purposes, it is usually
interesting to know that IRQs are being received, but not
displaying every single one.
Xiaoguang Wang [Thu, 11 Jun 2020 15:39:36 +0000 (23:39 +0800)]
io_uring: fix io_kiocb.flags modification race in IOPOLL mode
While testing io_uring in arm, we found sometimes io_sq_thread() keeps
polling io requests even though there are not inflight io requests in
block layer. After some investigations, found a possible race about
io_kiocb.flags, see below race codes:
1) in the end of io_write() or io_read()
req->flags &= ~REQ_F_NEED_CLEANUP;
kfree(iovec);
return ret;
2) in io_complete_rw_iopoll()
if (res != -EAGAIN)
req->flags |= REQ_F_IOPOLL_COMPLETED;
In IOPOLL mode, io requests still maybe completed by interrupt, then
above codes are not safe, concurrent modifications to req->flags, which
is not protected by lock or is not atomic modifications. I also had
disassemble io_complete_rw_iopoll() in arm:
req->flags |= REQ_F_IOPOLL_COMPLETED;
0xffff000008387b18 <+76>: ldr w0, [x19,#104]
0xffff000008387b1c <+80>: orr w0, w0, #0x1000
0xffff000008387b20 <+84>: str w0, [x19,#104]
Seems that the "req->flags |= REQ_F_IOPOLL_COMPLETED;" is load and
modification, two instructions, which obviously is not atomic.
To fix this issue, add a new iopoll_completed in io_kiocb to indicate
whether io request is completed.
KVM: nVMX: Consult only the "basic" exit reason when routing nested exit
Consult only the basic exit reason, i.e. bits 15:0 of vmcs.EXIT_REASON,
when determining whether a nested VM-Exit should be reflected into L1 or
handled by KVM in L0.
For better or worse, the switch statement in nested_vmx_exit_reflected()
currently defaults to "true", i.e. reflects any nested VM-Exit without
dedicated logic. Because the case statements only contain the basic
exit reason, any VM-Exit with modifier bits set will be reflected to L1,
even if KVM intended to handle it in L0.
Practically speaking, this only affects EXIT_REASON_MCE_DURING_VMENTRY,
i.e. a #MC that occurs on nested VM-Enter would be incorrectly routed to
L1, as "failed VM-Entry" is the only modifier that KVM can currently
encounter. The SMM modifiers will never be generated as KVM doesn't
support/employ a SMI Transfer Monitor. Ditto for "exit from enclave",
as KVM doesn't yet support virtualizing SGX, i.e. it's impossible to
enter an enclave in a KVM guest (L1 or L2).
Qing Zhang [Thu, 11 Jun 2020 12:56:52 +0000 (20:56 +0800)]
spi: tools: Add macro definitions to fix build errors
Add SPI_TX_OCTAL and SPI_RX_OCTAL to fix the following build errors:
CC spidev_test.o
spidev_test.c: In function ‘transfer’:
spidev_test.c:131:13: error: ‘SPI_TX_OCTAL’ undeclared (first use in this function)
if (mode & SPI_TX_OCTAL)
^
spidev_test.c:131:13: note: each undeclared identifier is reported only once for each function it appears in
spidev_test.c:137:13: error: ‘SPI_RX_OCTAL’ undeclared (first use in this function)
if (mode & SPI_RX_OCTAL)
^
spidev_test.c: In function ‘parse_opts’:
spidev_test.c:290:12: error: ‘SPI_TX_OCTAL’ undeclared (first use in this function)
mode |= SPI_TX_OCTAL;
^
spidev_test.c:308:12: error: ‘SPI_RX_OCTAL’ undeclared (first use in this function)
mode |= SPI_RX_OCTAL;
^
LD spidev_test-in.o
ld: cannot find spidev_test.o: No such file or directory
Additionally, maybe SPI_CS_WORD and SPI_3WIRE_HIZ will be used in the future,
so add them too.
Qing Zhang [Thu, 11 Jun 2020 12:56:51 +0000 (20:56 +0800)]
spi: tools: Make default_tx/rx and input_tx static
Fix the following sparse warning:
./spidev_test.c:50:9: warning: symbol 'default_tx' was not declared. Should it be static?
./spidev_test.c:59:9: warning: symbol 'default_rx' was not declared. Should it be static?
./spidev_test.c:60:6: warning: symbol 'input_tx' was not declared. Should it be static?
Colin Ian King [Thu, 11 Jun 2020 14:35:27 +0000 (15:35 +0100)]
umem: remove redundant initialization of variable ret
The variable ret is being initialized with a value that is never read
and it is being updated later with a new value. The initialization is
redundant and can be removed.
Colin Ian King [Thu, 11 Jun 2020 14:30:14 +0000 (15:30 +0100)]
pktcdvd: remove redundant initialization of variable ret
The variable ret is being initialized with a value that is never read
and it is being updated later with a new value. The initialization is
redundant and can be removed.
In function nvmet_async_event_process() we only process AENs iff
there is an open slot on the ctrl->async_event_cmds[] && aen
event list posted by the target is not empty. This keeps host
posted AEN outstanding if target generated AEN list is empty.
We do cleanup the target generated entries from the aen list in
nvmet_ctrl_free()-> nvmet_async_events_free() but we don't
process AEN posted by the host. This leads to following problem :-
When processing admin sq at the time of nvmet_sq_destroy() holds
an extra percpu reference(atomic value = 1), so in the following code
path after switching to atomic rcu, release function (nvmet_sq_free())
is not getting called which blocks the sq->free_done in
nvmet_sq_destroy() :-
void nvmet_sq_destroy(struct nvmet_sq *sq)
...
if (ctrl && ctrl->sqs && ctrl->sqs[0] == sq) {
nvmet_async_events_process(ctrl, status);
percpu_ref_put(&sq->ref);
}
percpu_ref_kill_and_confirm(&sq->ref, nvmet_confirm_sq);
wait_for_completion(&sq->confirm_done);
wait_for_completion(&sq->free_done); <-- Hang here
Which breaks the further disconnect sequence. This problem seems to be
introduced after commit 64f5e9cdd711b ("nvmet: fix memory leak when
removing namespaces and controllers concurrently").
This patch processes ctrl->async_event_cmds[] in the admin sq destroy()
context irrespetive of aen_list. Also we get rid of the controller's
aen_list processing in the nvmet_sq_destroy() context and just ignore
ctrl->aen_list.
This results in nvmet_async_events_process() being called from workqueue
context so we adjust the code accordingly.
nvme-pci: use simple suspend when a HMB is enabled
While the NVMe specification allows the device to access the host memory
buffer in host DRAM from all power states, hosts will fail access to
DRAM during S3 and similar power states.
Daniel Wagner [Fri, 29 May 2020 11:37:40 +0000 (13:37 +0200)]
nvme-fc: don't call nvme_cleanup_cmd() for AENs
Asynchronous event notifications do not have an associated request.
When fcp_io() fails we unconditionally call nvme_cleanup_cmd() which
leads to a crash.
Niklas Cassel [Sun, 7 Jun 2020 11:45:20 +0000 (13:45 +0200)]
nvme: do not call del_gendisk() on a disk that was never added
device_add_disk() is negated by del_gendisk().
alloc_disk_node() is negated by put_disk().
In nvme_alloc_ns(), device_add_disk() is one of the last things being
called in the success case, and only void functions are being called
after this. Therefore this call should not be negated in the error path.
The superfluous call to del_gendisk() leads to the following prints:
[ 7.839975] kobject: '(null)' (000000001ff73734): is not initialized, yet kobject_put() is being called.
[ 7.840865] WARNING: CPU: 2 PID: 361 at lib/kobject.c:736 kobject_put+0x70/0x120
YiFei Zhu [Wed, 10 Jun 2020 18:41:40 +0000 (13:41 -0500)]
selftests/bpf: Add cgroup_skb/egress test for load_bytes_relative
When cgroup_skb/egress triggers the MAC header is not set. Added a
test that asserts reading MAC header is a -EFAULT but NET header
succeeds. The test result from within the eBPF program is stored in
an 1-element array map that the userspace then reads and asserts on.
Another assertion is added that reading from a large offset, past
the end of packet, returns -EFAULT.
YiFei Zhu [Wed, 10 Jun 2020 18:41:39 +0000 (13:41 -0500)]
net/filter: Permit reading NET in load_bytes_relative when MAC not set
Added a check in the switch case on start_header that checks for
the existence of the header, and in the case that MAC is not set
and the caller requests for MAC, -EFAULT. If the caller requests
for NET then MAC's existence is completely ignored.
There is no function to check NET header's existence and as far
as cgroup_skb/egress is concerned it should always be set.
Removed for ptr >= the start of header, considering offset is
bounded unsigned and should always be true. len <= end - mac is
redundant to ptr + len <= end.
Tony Luck [Wed, 27 May 2020 18:28:08 +0000 (11:28 -0700)]
x86/mce/dev-mcelog: Fix -Wstringop-truncation warning about strncpy()
The kbuild test robot reported this warning:
arch/x86/kernel/cpu/mce/dev-mcelog.c: In function 'dev_mcelog_init_device':
arch/x86/kernel/cpu/mce/dev-mcelog.c:346:2: warning: 'strncpy' output \
truncated before terminating nul copying 12 bytes from a string of the \
same length [-Wstringop-truncation]
This is accurate, but I don't care that the trailing NUL character isn't
copied. The string being copied is just a magic number signature so that
crash dump tools can be sure they are decoding the right blob of memory.
Tony Luck [Wed, 20 May 2020 16:35:46 +0000 (09:35 -0700)]
x86/{mce,mm}: Unmap the entire page if the whole page is affected and poisoned
An interesting thing happened when a guest Linux instance took a machine
check. The VMM unmapped the bad page from guest physical space and
passed the machine check to the guest.
Linux took all the normal actions to offline the page from the process
that was using it. But then guest Linux crashed because it said there
was a second machine check inside the kernel with this stack trace:
This was odd, because a CLFLUSH instruction shouldn't raise a machine
check (it isn't consuming the data). Further investigation showed that
the VMM had passed in another machine check because is appeared that the
guest was accessing the bad page.
Fix is to check the scope of the poison by checking the MCi_MISC register.
If the entire page is affected, then unmap the page. If only part of the
page is affected, then mark the page as uncacheable.
This assumes that VMMs will do the logical thing and pass in the "whole
page scope" via the MCi_MISC register (since they unmapped the entire
page).
Thomas Gleixner [Thu, 11 Jun 2020 13:17:57 +0000 (15:17 +0200)]
Merge branch 'x86/entry' into ras/core
to fixup conflicts in arch/x86/kernel/cpu/mce/core.c so MCE specific follow
up patches can be applied without creating a horrible merge conflict
afterwards.
The entry rework moved interrupt entry code from the irqentry to the
noinstr section which made the irqentry section empty.
This breaks boundary checks which rely on the __irqentry_text_start/end
markers to find out whether a function in a stack trace is
interrupt/exception entry code. This affects the function graph tracer and
filter_irq_stacks().
As the IDT entry points are all sequentialy emitted this is rather simple
to unbreak by injecting __irqentry_text_start/end as global labels.
To make this work correctly:
- Remove the IRQENTRY_TEXT section from the x86 linker script
- Define __irqentry so it breaks the build if it's used
- Adjust the entry mirroring in PTI
- Remove the redundant kprobes and unwinder bound checks
Thomas Gleixner [Thu, 28 May 2020 14:53:16 +0000 (16:53 +0200)]
x86/idt: Mark init only functions __init
Since 8175cfbbbfcb ("x86/idt: Remove update_intr_gate()") set_intr_gate()
and idt_setup_from_table() are only called from __init functions. Mark them
as well.
Peter Zijlstra [Fri, 29 May 2020 21:27:40 +0000 (23:27 +0200)]
x86/entry: Rename trace_hardirqs_off_prepare()
The typical pattern for trace_hardirqs_off_prepare() is:
ENTRY
lockdep_hardirqs_off(); // because hardware
... do entry magic
instrumentation_begin();
trace_hardirqs_off_prepare();
... do actual work
trace_hardirqs_on_prepare();
lockdep_hardirqs_on_prepare();
instrumentation_end();
... do exit magic
lockdep_hardirqs_on();
which shows that it's named wrong, rename it to
trace_hardirqs_off_finish(), as it concludes the hardirq_off transition.
Also, given that the above is the only correct order, make the traditional
all-in-one trace_hardirqs_off() follow suit.
Peter Zijlstra [Fri, 29 May 2020 21:27:38 +0000 (23:27 +0200)]
x86/entry: Remove DBn stacks
Both #DB itself, as all other IST users (NMI, #MC) now clear DR7 on
entry. Combined with not allowing breakpoints on entry/noinstr/NOKPROBE
text and no single step (EFLAGS.TF) inside the #DB handler should guarantee
no nested #DB.
Peter Zijlstra [Fri, 29 May 2020 21:27:36 +0000 (23:27 +0200)]
x86/entry: Optimize local_db_save() for virt
Because DRn access is 'difficult' with virt; but the DR7 read is cheaper
than a cacheline miss on native, add a virt specific fast path to
local_db_save(), such that when breakpoints are not in use to avoid
touching DRn entirely.
Peter Zijlstra [Fri, 29 May 2020 21:27:34 +0000 (23:27 +0200)]
x86/entry, nmi: Disable #DB
Instead of playing stupid games with IST stacks, fully disallow #DB
during NMIs. There is absolutely no reason to allow them, and killing
this saves a heap of trouble.
#DB is already forbidden on noinstr and CEA, so there can't be a #DB before
this. Disabling it right after nmi_enter() ensures that the full NMI code
is protected.
Vitaly Kuznetsov [Wed, 20 May 2020 16:16:00 +0000 (18:16 +0200)]
xen: Move xen_setup_callback_vector() definition to include/xen/hvm.h
Kbuild test robot reports the following problem on ARM:
for 'xen_setup_callback_vector' [-Wmissing-prototypes]
1664 | void xen_setup_callback_vector(void) {}
| ^~~~~~~~~~~~~~~~~~~~~~~~~
The problem is that xen_setup_callback_vector is a x86 only thing, its
definition is present in arch/x86/xen/xen-ops.h but not on ARM. In
events_base.c there is a stub for !CONFIG_XEN_PVHVM but it is not declared
as 'static'.
On x86 the situation is hardly better: drivers/xen/events/events_base.c
doesn't include 'xen-ops.h' from arch/x86/xen/, it includes its namesake
from include/xen/ which also results in a 'no previous prototype' warning.
Currently, xen_setup_callback_vector() has two call sites: one in
drivers/xen/events_base.c and another in arch/x86/xen/suspend_hvm.c. The
former is placed under #ifdef CONFIG_X86 and the later is only compiled
in when CONFIG_XEN_PVHVM.
Resolve the issue by moving xen_setup_callback_vector() declaration to
arch neutral 'include/xen/hvm.h' as the implementation lives in arch
neutral drivers/xen/events/events_base.c.
Thomas Gleixner [Thu, 21 May 2020 20:05:45 +0000 (22:05 +0200)]
x86/entry: Convert reschedule interrupt to IDTENTRY_SYSVEC_SIMPLE
The scheduler IPI does not need the full interrupt entry handling logic
when the entry is from kernel mode. Use IDTENTRY_SYSVEC_SIMPLE and spare
all the overhead.
Thomas Gleixner [Thu, 21 May 2020 20:05:44 +0000 (22:05 +0200)]
x86/entry: Convert XEN hypercall vector to IDTENTRY_SYSVEC
Convert the last oldstyle defined vector to IDTENTRY_SYSVEC:
- Implement the C entry point with DEFINE_IDTENTRY_SYSVEC
- Emit the ASM stub with DECLARE_IDTENTRY_SYSVEC
- Remove the ASM idtentries in 64-bit
- Remove the BUILD_INTERRUPT entries in 32-bit
- Remove the old prototypes
Fixup the related XEN code by providing the primary C entry point in x86 to
avoid cluttering the generic code with X86'isms.
Thomas Gleixner [Thu, 21 May 2020 20:05:43 +0000 (22:05 +0200)]
x86/entry: Convert various hypervisor vectors to IDTENTRY_SYSVEC
Convert various hypervisor vectors to IDTENTRY_SYSVEC:
- Implement the C entry point with DEFINE_IDTENTRY_SYSVEC
- Emit the ASM stub with DECLARE_IDTENTRY_SYSVEC
- Remove the ASM idtentries in 64-bit
- Remove the BUILD_INTERRUPT entries in 32-bit
- Remove the old prototypes
Thomas Gleixner [Thu, 21 May 2020 20:05:41 +0000 (22:05 +0200)]
x86/entry: Convert various system vectors
Convert various system vectors to IDTENTRY_SYSVEC:
- Implement the C entry point with DEFINE_IDTENTRY_SYSVEC
- Emit the ASM stub with DECLARE_IDTENTRY_SYSVEC
- Remove the ASM idtentries in 64-bit
- Remove the BUILD_INTERRUPT entries in 32-bit
- Remove the old prototypes
Thomas Gleixner [Thu, 21 May 2020 20:05:40 +0000 (22:05 +0200)]
x86/entry: Convert SMP system vectors to IDTENTRY_SYSVEC
Convert SMP system vectors to IDTENTRY_SYSVEC:
- Implement the C entry point with DEFINE_IDTENTRY_SYSVEC
- Emit the ASM stub with DECLARE_IDTENTRY_SYSVEC
- Remove the ASM idtentries in 64-bit
- Remove the BUILD_INTERRUPT entries in 32-bit
- Remove the old prototypes
Thomas Gleixner [Thu, 21 May 2020 20:05:39 +0000 (22:05 +0200)]
x86/entry: Convert APIC interrupts to IDTENTRY_SYSVEC
Convert APIC interrupts to IDTENTRY_SYSVEC:
- Implement the C entry point with DEFINE_IDTENTRY_SYSVEC
- Emit the ASM stub with DECLARE_IDTENTRY_SYSVEC
- Remove the ASM idtentries in 64-bit
- Remove the BUILD_INTERRUPT entries in 32-bit
- Remove the old prototypes
Thomas Gleixner [Thu, 21 May 2020 20:05:38 +0000 (22:05 +0200)]
x86/entry: Provide IDTENTRY_SYSVEC
Provide IDTENTRY variants for system vectors to consolidate the different
mechanisms to emit the ASM stubs for 32- and 64-bit.
On 64-bit this also moves the stack switching from ASM to C code. 32-bit will
excute the system vectors w/o stack switching as before.
The simple variant is meant for "empty" system vectors like scheduler IPI
and KVM posted interrupt vectors. These do not need the full glory of irq
enter/exit handling with softirq processing and more.
Thomas Gleixner [Thu, 21 May 2020 20:05:37 +0000 (22:05 +0200)]
x86/entry: Use idtentry for interrupts
Replace the extra interrupt handling code and reuse the existing idtentry
machinery. This moves the irq stack switching on 64-bit from ASM to C code;
32-bit already does the stack switching in C.
This requires to remove HAVE_IRQ_EXIT_ON_IRQ_STACK as the stack switch is
not longer in the low level entry code.
Thomas Gleixner [Thu, 21 May 2020 20:05:36 +0000 (22:05 +0200)]
x86/entry: Add IRQENTRY_IRQ macro
Provide a seperate IDTENTRY macro for device interrupts. Similar to
IDTENTRY_ERRORCODE with the addition of invoking irq_enter/exit_rcu() and
providing the errorcode as a 'u8' argument to the C function, which
truncates the sign extended vector number.
Thomas Gleixner [Thu, 21 May 2020 20:05:35 +0000 (22:05 +0200)]
x86/irq: Rework handle_irq() for 64-bit
To consolidate the interrupt entry/exit code vs. the other exceptions
make handle_irq() an inline and handle both 64-bit and 32-bit mode.
Preparatory change to move irq stack switching for 64-bit to C which allows
to consolidate the entry exit handling by reusing the idtentry machinery
both in ASM and C.
Thomas Gleixner [Thu, 21 May 2020 20:05:34 +0000 (22:05 +0200)]
x86/irq: Convey vector as argument and not in ptregs
Device interrupts which go through do_IRQ() or the spurious interrupt
handler have their separate entry code on 64 bit for no good reason.
Both 32 and 64 bit transport the vector number through ORIG_[RE]AX in
pt_regs. Further the vector number is forced to fit into an u8 and is
complemented and offset by 0x80 so it's in the signed character
range. Otherwise GAS would expand the pushq to a 5 byte instruction for any
vector > 0x7F.
Treat the vector number like an error code and hand it to the C function as
argument. This allows to get rid of the extra entry code in a later step.
Simplify the error code push magic by implementing the pushq imm8 via a
'.byte 0x6a, vector' sequence so GAS is not able to screw it up. As the
pushq imm8 is sign extending the resulting error code needs to be truncated
to 8 bits in C code.
Thomas Gleixner [Thu, 21 May 2020 20:05:30 +0000 (22:05 +0200)]
x86/entry: Change exit path of xen_failsafe_callback
xen_failsafe_callback() is invoked from XEN for two cases:
1. Fault while reloading DS, ES, FS or GS
2. Fault while executing IRET
#1 retries the IRET after XEN has fixed up the segments.
#2 injects a #GP which kills the task
For #1 there is no reason to go through the full exception return path
because the tasks TIF state is still the same. So just going straight to
the IRET path is good enough.
Thomas Gleixner [Thu, 21 May 2020 20:05:28 +0000 (22:05 +0200)]
x86/entry: Switch page fault exception to IDTENTRY_RAW
Convert page fault exceptions to IDTENTRY_RAW:
- Implement the C entry point with DEFINE_IDTENTRY_RAW
- Add the CR2 read into the exception handler
- Add the idtentry_enter/exit_cond_rcu() invocations in
in the regular page fault handler and in the async PF
part.
- Emit the ASM stub with DECLARE_IDTENTRY_RAW
- Remove the ASM idtentry in 64-bit
- Remove the CR2 read from 64-bit
- Remove the open coded ASM entry code in 32-bit
- Fix up the XEN/PV code
- Remove the old prototypes
Thomas Gleixner [Thu, 21 May 2020 20:05:27 +0000 (22:05 +0200)]
x86/entry/64: Simplify idtentry_body
All C functions which do not have an error code have been converted to the
new IDTENTRY interface which does not expect an error code in the
arguments. Spare the XORL.
Thomas Gleixner [Thu, 21 May 2020 20:05:26 +0000 (22:05 +0200)]
x86/entry: Switch XEN/PV hypercall entry to IDTENTRY
Convert the XEN/PV hypercall to IDTENTRY:
- Emit the ASM stub with DECLARE_IDTENTRY
- Remove the ASM idtentry in 64-bit
- Remove the open coded ASM entry code in 32-bit
- Remove the old prototypes
The handler stubs need to stay in ASM code as they need corner case handling
and adjustment of the stack pointer.
Provide a new C function which invokes the entry/exit handling and calls
into the XEN handler on the interrupt stack if required.
The exit code is slightly different from the regular idtentry_exit() on
non-preemptible kernels. If the hypercall is preemptible and need_resched()
is set then XEN provides a preempt hypercall scheduling function.
Move this functionality into the entry code so it can use the existing
idtentry functionality.
Thomas Gleixner [Thu, 21 May 2020 20:05:23 +0000 (22:05 +0200)]
x86/entry: Provide helpers for executing on the irqstack
Device interrupt handlers and system vector handlers are executed on the
interrupt stack. The stack switch happens in the low level assembly entry
code. This conflicts with the efforts to consolidate the exit code in C to
ensure correctness vs. RCU and tracing.
As there is no way to move #DB away from IST due to the MOV SS issue, the
requirements vs. #DB and NMI for switching to the interrupt stack do not
exist anymore. The only requirement is that interrupts are disabled.
That allows the moving of the stack switching to C code, which simplifies the
entry/exit handling further, because it allows the switching of stacks after
handling the entry and on exit before handling RCU, returning to usermode and
kernel preemption in the same way as for regular exceptions.
The initial attempt of having the stack switching in inline ASM caused too
much headache vs. objtool and the unwinder. After analysing the use cases
it was agreed on that having the stack switch in ASM for the price of an
indirect call is acceptable, as the main users are indirect call heavy
anyway and the few system vectors which are empty shells (scheduler IPI and
KVM posted interrupt vectors) can run from the regular stack.
Provide helper functions to check whether the interrupt stack is already
active and whether stack switching is required.
64-bit only for now, as 32-bit has a variant of that already. Once this is
cleaned up, the two implementations might be consolidated as an additional
cleanup on top.
Thomas Gleixner [Thu, 21 May 2020 20:05:21 +0000 (22:05 +0200)]
genirq: Provide irq_enter/exit_rcu()
irq_enter()/exit() currently include RCU handling. To properly separate the RCU
handling code, provide variants which contain only the non-RCU related
functionality.
Thomas Gleixner [Thu, 21 May 2020 20:05:19 +0000 (22:05 +0200)]
x86/idtentry: Switch to conditional RCU handling
Switch all idtentry_enter/exit() users over to the new conditional RCU
handling scheme and make the user mode entries in #DB, #INT3 and #MCE use
the user mode idtentry functions.
Thomas Gleixner [Thu, 21 May 2020 20:05:18 +0000 (22:05 +0200)]
x86/entry: Provide idtentry_enter/exit_user()
As there are exceptions which already handle entry from user mode and from
kernel mode separately, providing explicit user entry/exit handling callbacks
makes sense and makes the code easier to understand.
Thomas Gleixner [Thu, 21 May 2020 20:05:17 +0000 (22:05 +0200)]
x86/entry: Provide idtentry_entry/exit_cond_rcu()
After a lengthy discussion [1] it turned out that RCU does not need a full
rcu_irq_enter/exit() when RCU is already watching. All it needs if
NOHZ_FULL is active is to check whether the tick needs to be restarted.
This allows to avoid a separate variant for the pagefault handler which
cannot invoke rcu_irq_enter() on a kernel pagefault which might sleep.
The cond_rcu argument is only temporary and will be removed once the
existing users of idtentry_enter/exit() have been cleaned up. After that
the code can be significantly simplified.
projects / linux.git / log