When a queue is switched to soft offline during heavy load and later
switched to soft online again and now used, it may be that the caller
is blocked forever in the ioctl call.

The failure occurs because there is a pending reply after the queue(s)
have been switched to offline. This orphaned reply is received when
the queue is switched to online and is accidentally counted for the
outstanding replies. So when there was a valid outstanding reply and
this orphaned reply is received it counts as the outstanding one thus
dropping the outstanding counter to 0. Voila, with this counter the
receive function is not called any more and the real outstanding reply
is never received (until another request comes in...) and the ioctl
blocks.

The fix is simple. However, instead of readjusting the counter when an
orphaned reply is detected, I check the queue status for not empty and
compare this to the outstanding counter. So if the queue is not empty
then the counter must not drop to 0 but at least have a value of 1.
Signed-off-by: NHarald Freudenberger <freude@linux.ibm.com>
Cc: stable@vger.kernel.org
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

pci and string functions changes on features depend on changes from the
fixes branch.

* fixes:
  s390: add Alexander Gordeev as reviewer
  s390: fix strrchr() implementation
  vfio-ccw: step down as maintainer
  KVM: s390: remove myself as reviewer
  s390/pci: fix zpci_zdev_put() on reserve
  bpf, s390: Fix potential memory leak about jit_data
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Add s390 support for ftrace direct call samples, which also enables
ftrace direct call selftests within ftrace selftests.
Acked-by: NIlya Leoshkevich <iii@linux.ibm.com>
Reviewed-by: NSven Schnelle <svens@linux.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Acked-by: NSteven Rostedt (VMware) <rostedt@goodmis.org>
Link: https://lore.kernel.org/r/20211012133802.2460757-5-hca@linux.ibm.comSigned-off-by: NVasily Gorbik <gor@linux.ibm.com>

Add HAVE_SAMPLE_FTRACE_DIRECT config option which can be selected by
architectures which have support for ftrace direct call samples.
Acked-by: NIlya Leoshkevich <iii@linux.ibm.com>
Reviewed-by: NSven Schnelle <svens@linux.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Acked-by: NSteven Rostedt (VMware) <rostedt@goodmis.org>
Link: https://lore.kernel.org/r/20211012133802.2460757-4-hca@linux.ibm.comSigned-off-by: NVasily Gorbik <gor@linux.ibm.com>

Make STACK_FRAME_OVERHEAD available via asm-offsets.h. This allows to
add s390 specific asm code to e.g. ftrace samples, without requiring
to add random header files, which might cause all sort of problems on
other architectures. asm-offsets.h can be assumed to be non-problematic.
Acked-by: NIlya Leoshkevich <iii@linux.ibm.com>
Reviewed-by: NSven Schnelle <svens@linux.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Acked-by: NSteven Rostedt (VMware) <rostedt@goodmis.org>
Link: https://lore.kernel.org/r/20211012133802.2460757-3-hca@linux.ibm.comSigned-off-by: NVasily Gorbik <gor@linux.ibm.com>

This is the s390 variant of commit 562955fe ("ftrace/x86: Add
register_ftrace_direct() for custom trampolines").
Acked-by: NIlya Leoshkevich <iii@linux.ibm.com>
Reviewed-by: NSven Schnelle <svens@linux.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Acked-by: NSteven Rostedt (VMware) <rostedt@goodmis.org>
Link: https://lore.kernel.org/r/20211012133802.2460757-2-hca@linux.ibm.comSigned-off-by: NVasily Gorbik <gor@linux.ibm.com>

Alexander Gordeev will help reviewing s390 code.
Acked-by: NAlexander Gordeev <agordeev@linux.ibm.com>
Acked-by: NVasily Gorbik <gor@linux.ibm.com>
Acked-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Fix two problems found in the strrchr() implementation for s390
architectures: evaluate empty strings (return the string address instead of
NULL, if '\0' is passed as second argument); evaluate the first character
of non-empty strings (the current implementation stops at the second).

Fixes: 1da177e4 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reported-by: Heiko Carstens <hca@linux.ibm.com> (incorrect behavior with empty strings)
Signed-off-by: NRoberto Sassu <roberto.sassu@huawei.com>
Link: https://lore.kernel.org/r/20211005120836.60630-1-roberto.sassu@huawei.comSigned-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

I currently don't have time to act as vfio-ccw maintainer
anymore, but I trust that I leave it in capable hands.
Signed-off-by: NCornelia Huck <cohuck@redhat.com>
Acked-by: NEric Farman <farman@linux.ibm.com>
Acked-by: NMatthew Rosato <mjrosato@linux.ibm.com>
Acked-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Link: https://lore.kernel.org/r/20211006160120.217636-3-cohuck@redhat.comSigned-off-by: NVasily Gorbik <gor@linux.ibm.com>

I currently don't have time anymore to review KVM/s390
code.
Signed-off-by: NCornelia Huck <cohuck@redhat.com>
Acked-by: NJanosch Frank <frankja@linux.ibm.com>
Acked-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Link: https://lore.kernel.org/r/20211006160120.217636-2-cohuck@redhat.comSigned-off-by: NVasily Gorbik <gor@linux.ibm.com>

Use string_upper() from string helper module instead of open coded variant.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20211001130201.72545-1-andriy.shevchenko@linux.intel.com
[hca@linux.ibm.com: removed hunk which converts extmem.c]
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Add regs_get_kernel_argument() which returns Nth argument of a
function call.

This enables ftrace kprobe events to access kernel function arguments
via $argN syntax.

This is the s390 variant of commit a823c35f ("arm64: ptrace: Add
function argument access API").
Acked-by: NIlya Leoshkevich <iii@linux.ibm.com>
Reviewed-by: NSteffen Maier <maier@linux.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

This is the s390 variant of commit 9855c462 ("selftests/ftrace:
Add ppc support for kprobe args tests").
Acked-by: NIlya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Reported-by: NSteffen Maier <maier@linux.ibm.com>
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

ftrace_regs_caller is an alias to ftrace_caller - making ftrace_caller
quite heavyweight. Split the function and provide an ftrace_caller
implementation which comes with fewer instructions. Especially getting
rid of 'stosm' on each function entry should help here, e.g. to
have less performance impact on live patched functions.
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Add ftrace_instruction_pointer_set() helper function to match x86.
See commit 2860cd8a ("livepatch: Use the default ftrace_ops
instead of REGS when ARGS is available").
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Add HAVE_DYNAMIC_FTRACE_WITH_ARGS support similar to commit 02a474ca
("ftrace/x86: Allow for arguments to be passed in to ftrace_regs by default").

s390's ftrace implementation always provides all registers with
pt_regs, therefore this is trivial.
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Add missing __init_or_module to arch_jump_label_transform_static().
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Trivial patch just to get rid of the leading underscores.
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Specify HAVE_JUMP_LABEL_BATCH in header file. This allows to make use
of the arch_jump_label_transform_queue()/arch_jump_label_transform_apply()
mechanism.

However unlike on x86, which currently is the only user of this
mechanism, the to be patched instructions are still directly
modified. The only difference to before is that serialization is only
done after all instructions have been modified. This way the number of
serialization/synchronization events is reduced.
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

CPUs must be serialized also when ftrace_graph_caller gets patched.
This is missing since ftrace function graph support was added on s390.
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Introduce a text_poke_sync() similar to what x86 has. This can be
used to execute a serializing instruction on all CPUs (including
the current one).

Note: according to the Principles of Operation an IPI (= interrupt)
will already serialize a CPU, however it is better to be explicit. In
addition on_each_cpu() makes sure that also the current CPU get
serialized - just to make sure that possible preemption can prevent
some theoretical case where a CPU will not be serialized.

Therefore text_poke_sync() has to be used whenever code got modified,
just to avoid to rely on implicit serialization.

Also introduce text_poke_sync_lock() which will also disable CPU
hotplug, to prevent that any CPU is just going online with a
prefetched old version of a modified instruction.
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Factor out bcr_serialize() inline assembly function which describes
what the bcr instruction is used for.
Use bcr_serialize() like before in mb(), but also in upcoming changes.
Signed-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Modified stack unwinder self tests to use kunit framework. The
functionality stayed the same but the output format is now in tap13
format.
Reviewed-by: NVasily Gorbik <gor@linux.ibm.com>
Signed-off-by: NMete Durlu <meted@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Fix kernel-doc warning due to incorrect parameter name in
kernel-doc function notation:

../arch/s390/include/asm/debug.h:484: warning: Function parameter or member 'pages' not described in 'DEFINE_STATIC_DEBUG_INFO'
../arch/s390/include/asm/debug.h:484: warning: Excess function parameter 'pages_per_area' description in 'DEFINE_STATIC_DEBUG_INFO'

Fixes: d72541f9 ("s390/debug: add early tracing support")
Signed-off-by: NRandy Dunlap <rdunlap@infradead.org>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: linux-s390@vger.kernel.org
Cc: Peter Oberparleiter <oberpar@linux.ibm.com>
Link: https://lore.kernel.org/r/20211005051657.16714-1-rdunlap@infradead.orgSigned-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

We never checked for errors on add_disk() as this function
returned void. Now that this is fixed, use the shiny new
error handling.
Acked-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NLuis Chamberlain <mcgrof@kernel.org>
Link: https://lore.kernel.org/r/20210927220232.1071926-7-mcgrof@kernel.orgSigned-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

We never checked for errors on add_disk() as this function
returned void. Now that this is fixed, use the shiny new
error handling.
Signed-off-by: NLuis Chamberlain <mcgrof@kernel.org>
Signed-off-by: NGerald Schaefer <gerald.schaefer@linux.ibm.com>
Link: https://lore.kernel.org/r/20210927220232.1071926-6-mcgrof@kernel.orgSigned-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

We never checked for errors on add_disk() as this function
returned void. Now that this is fixed, use the shiny new
error handling.

Be sure to call dasd_gendisk_free() on error.
Signed-off-by: NLuis Chamberlain <mcgrof@kernel.org>
Link: https://lore.kernel.org/r/20210927220232.1071926-5-mcgrof@kernel.orgSigned-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

The memory for amode31 section is allocated from the decompressed
kernel. Instead, allocate that memory from the decompressor. This
is a prerequisite to allow initialization of the virtual memory
before the decompressed kernel takes over.
Reviewed-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NAlexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Partially revert commit 4555b9f34296 ("s390/boot: move
dma sections from decompressor to decompressed kernel").
This is a prerequisite to allow initialization of virtual
memory in decompressor and avoid overwriting of ASCEs in
the decompressed kernel otherwise.

Since the control registers 2, 5 and 15 are reinitialized
in the decompressed kernel again, this change does not
prevent relocating of amode31 section in any way.
Reviewed-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NAlexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

When CONFIG_SCLP_VT220_TTY=y and CONFIG_SCLP_VT220_CONSOLE is not set:

../drivers/s390/char/sclp_vt220.c:771:13: warning: '__sclp_vt220_flush_buffer' defined but not used [-Wunused-function]
  771 | static void __sclp_vt220_flush_buffer(void)

so move this function inside the #ifdef block where it is used.
Signed-off-by: NRandy Dunlap <rdunlap@infradead.org>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Link: https://lore.kernel.org/r/20210927215647.11506-1-rdunlap@infradead.orgSigned-off-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Add a few testcases to make sure that it's not possible to place
a kprobe in the mid of an instruction on s390.
Signed-off-by: NSven Schnelle <svens@linux.ibm.com>
Acked-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

FTRACE_GEN_NOP_ASM(name) can be used to generate assembly functions with
the required information added to allow tracing via kprobes/ftrace. It
adds the nop instruction which will be patched by ftrace later. If the
compiler supports -mnop-mcount it will also add an entry to the
__mcount_loc section.
Signed-off-by: NSven Schnelle <svens@linux.ibm.com>
Acked-by: NHeiko Carstens <hca@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

Fix these warnings that are reported when compiling with W=1:

  drivers/s390/char/sclp_sd.c:132: warning: Function parameter or member 'listener' not described in 'sclp_sd_listener_init'
  drivers/s390/char/sclp_sd.c:408: warning: Function parameter or member 'cookie' not described in 'sclp_sd_file_update_async'
  drivers/s390/char/sclp_sd.c:422: warning: Function parameter or member 'attr' not described in 'reload_store'
  drivers/s390/char/sclp_sd.c:422: warning: Function parameter or member 'buf' not described in 'reload_store'
  drivers/s390/char/sclp_sd.c:422: warning: Function parameter or member 'count' not described in 'reload_store'
  drivers/s390/char/sclp_sd.c:457: warning: Function parameter or member 'file' not described in 'data_read'
  drivers/s390/char/sclp_sd.c:457: warning: Function parameter or member 'attr' not described in 'data_read'
Signed-off-by: NPeter Oberparleiter <oberpar@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

cdev->online for the purge function must not be checked for the
non-IO subchannel type. Make sure that we are deriving the cdev only
from sch-type SUBCHANNEL_TYPE_IO.
Signed-off-by: NVineeth Vijayan <vneethv@linux.ibm.com>
Reviewed-by: NPeter Oberparleiter <oberpar@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

We often need to figure out what operations were performed in response
to an error or availability event. The operations are easily accessible
in s390dbf/pci_msg but the events have to be correlated with these from
either the kernel log or s390dbf/pci_err. Improve this situation by
logging the most important data from error and availability events that
is the FID, PEC and FH together with the operations.
Reviewed-by: NMatthew Rosato <mjrosato@linux.ibm.com>
Signed-off-by: NNiklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>

The cio_ignore list is used to create and maintain the list of devices
which is to be ignored by Linux. During boot-time, this list is adjusted
and accommodate all the devices which are configured on the HMC
interface. Once these devices are accessible, they are then available to
Linux and set online.

cio_ignore purge function should align with this functionality. But
currently, the subchannel associated with the offline-devices are not
unregistered during purge. Add an explicit subchannel-unregister function
in the purge_fn callback.
Signed-off-by: NVineeth Vijayan <vneethv@linux.ibm.com>
Reviewed-by: NPeter Oberparleiter <oberpar@linux.ibm.com>
Signed-off-by: NVasily Gorbik <gor@linux.ibm.com>