Janosch Frank [Fri, 13 Jul 2018 10:28:31 +0000 (11:28 +0100)]
KVM: s390: Add huge page enablement control
General KVM huge page support on s390 has to be enabled via the
kvm.hpage module parameter. Either nested or hpage can be enabled, as
we currently do not support vSIE for huge backed guests. Once the vSIE
support is added we will either drop the parameter or enable it as
default.
For a guest the feature has to be enabled through the new
KVM_CAP_S390_HPAGE_1M capability and the hpage module
parameter. Enabling it means that cmm can't be enabled for the vm and
disables pfmf and storage key interpretation.
This is due to the fact that in some cases, in upcoming patches, we
have to split huge pages in the guest mapping to be able to set more
granular memory protection on 4k pages. These split pages have fake
page tables that are not visible to the Linux memory management which
subsequently will not manage its PGSTEs, while the SIE will. Disabling
these features lets us manage PGSTE data in a consistent matter and
solve that problem.
Signed-off-by: Janosch Frank <frankja@linux.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com>
Janosch Frank [Fri, 13 Jul 2018 10:28:37 +0000 (11:28 +0100)]
s390/mm: Add huge page gmap linking support
Let's allow huge pmd linking when enabled through the
KVM_CAP_S390_HPAGE_1M capability. Also we can now restrict gmap
invalidation and notification to the cases where the capability has
been activated and save some cycles when that's not the case.
Signed-off-by: Janosch Frank <frankja@linux.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com>
s390/mm: hugetlb pages within a gmap can not be freed
Guests backed by huge pages could theoretically free unused pages via
the diagnose 10 instruction. We currently don't allow that, so we
don't have to refault it once it's needed again.
Signed-off-by: Dominik Dingel <dingel@linux.vnet.ibm.com> Reviewed-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com> Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Janosch Frank [Wed, 18 Jul 2018 12:40:22 +0000 (13:40 +0100)]
KVM: s390: Add skey emulation fault handling
When doing skey emulation for huge guests, we now need to fault in
pmds, as we don't have PGSTES anymore to store them when we do not
have valid table entries.
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Janosch Frank [Fri, 13 Jul 2018 10:28:28 +0000 (11:28 +0100)]
s390/mm: Add huge pmd storage key handling
Storage keys for guests with huge page mappings have to be managed in
hardware. There are no PGSTEs for PMDs that we could use to retain the
guests's logical view of the key.
Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com>
s390/mm: Clear huge page storage keys on enable_skey
When a guest starts using storage keys, we trap and set a default one
for its whole valid address space. With this patch we are now able to
do that for large pages.
To speed up the storage key insertion, we use
__storage_key_init_range, which in-turn will use sske_frame to set
multiple storage keys with one instruction. As it has been previously
used for debuging we have to get rid of the default key check and make
it quiescing.
Signed-off-by: Dominik Dingel <dingel@linux.vnet.ibm.com> Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com>
[replaced page_set_storage_key loop with __storage_key_init_range] Reviewed-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com>
Janosch Frank [Fri, 13 Jul 2018 10:28:22 +0000 (11:28 +0100)]
s390/mm: Add gmap pmd invalidation and clearing
If the host invalidates a pmd, we also have to invalidate the
corresponding gmap pmds, as well as flush them from the TLB. This is
necessary, as we don't share the pmd tables between host and guest as
we do with ptes.
The clearing part of these three new functions sets a guest pmd entry
to _SEGMENT_ENTRY_EMPTY, so the guest will fault on it and we will
re-link it.
Flushing the gmap is not necessary in the host's lazy local and csp
cases. Both purge the TLB completely.
Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com> Reviewed-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Acked-by: David Hildenbrand <david@redhat.com>
Janosch Frank [Fri, 13 Jul 2018 10:28:21 +0000 (11:28 +0100)]
s390/mm: Add gmap pmd notification bit setting
Like for ptes, we also need invalidation notification for pmds, to
make sure the guest lowcore pages are always accessible and later
addition of shadowed pmds.
With PMDs we do not have PGSTEs or some other bits we could use in the
host PMD. Instead we pick one of the free bits in the gmap PMD. Every
time a host pmd will be invalidated, we will check if the respective
gmap PMD has the bit set and in that case fire up the notifier.
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Janosch Frank [Fri, 13 Jul 2018 10:28:20 +0000 (11:28 +0100)]
s390/mm: Add gmap pmd linking
Let's allow pmds to be linked into gmap for the upcoming s390 KVM huge
page support.
Before this patch we copied the full userspace pmd entry. This is not
correct, as it contains SW defined bits that might be interpreted
differently in the GMAP context. Now we only copy over all hardware
relevant information leaving out the software bits.
Signed-off-by: Janosch Frank <frankja@linux.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com>
Janosch Frank [Fri, 13 Jul 2018 10:28:18 +0000 (11:28 +0100)]
s390/mm: Abstract gmap notify bit setting
Currently we use the software PGSTE bits PGSTE_IN_BIT and
PGSTE_VSIE_BIT to notify before an invalidation occurs on a prefix
page or a VSIE page respectively. Both bits are pgste specific, but
are used when protecting a memory range.
Let's introduce abstract GMAP_NOTIFY_* bits that will be realized into
the respective bits when gmap DAT table entries are protected.
Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com> Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com>
Janosch Frank [Fri, 13 Jul 2018 10:28:16 +0000 (11:28 +0100)]
s390/mm: Make gmap_protect_range more modular
This patch reworks the gmap_protect_range logic and extracts the pte
handling into an own function. Also we do now walk to the pmd and make
it accessible in the function for later use. This way we can add huge
page handling logic more easily.
Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com> Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> Reviewed-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Now that the early boot rework is upstream we can enable the gcc plugins
again. See git commit 72f108b308707f21499e0ac05bf7370360cf06d8
"s390: disable gcc plugins" for reference.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Thomas Richter [Tue, 3 Jul 2018 09:21:06 +0000 (11:21 +0200)]
s390 cio: Rewrite trace point class s390_class_schib
Tools like 'perf stat' parse the trace point format files defined
in /sys/kernel/debug/tracing/events/s390/.../format to handle
the print fmt: statement. The kernel provides a library in
directory linux/tools/lib/traceevent/* for this reason.
This library can not handle structures or unions defined in
the TRACE_EVENT/TP_STRUCT__entry macros with __field_struct macro.
There is no possibility to extract a structure member
(which might be a bit field) since there is no packing
information nor bit field offset by parsing the printf fmt line.
Therefore rewrite the TRACE_EVENT macro and add the
__field macro for the necessary members.
Keep the __fieldstruct macro to extract the complete
structure when dumps are analysed.
Note that the same information is displayed, this is no
interface change.
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Acked-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Thomas Richter [Mon, 9 Jul 2018 08:30:34 +0000 (10:30 +0200)]
s390 cio: Rewrite trace point in s390_cio_tsch
Tools like 'perf stat' parse the trace point format files defined
in /sys/kernel/debug/tracing/events/s390/.../format to handle
the print fmt: statement. The kernel provides a library in
directory linux/tools/lib/traceevent/* for this reason.
This library can not handle structures or unions defined in
the TRACE_EVENT/TP_STRUCT__entry macros with __field_struct macro.
There is no possibility to extract a structure member
(which might be a bit field) since there is no packing
information nor bit field offset by parsing the printf fmt line.
Therefore rewrite the TRACE_EVENT macro and add the
__field macro for the necessary members.
Keep the __fieldstruct macro to extract the complete
structure when dumps are analysed.
Note that the same information is displayed, this is no
interface change.
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Acked-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Thomas Richter [Tue, 3 Jul 2018 08:38:16 +0000 (10:38 +0200)]
s390 cio: Rewrite trace point in s390_cio_adapter_int
Tools like 'perf stat' parse the trace point format files defined
in /sys/kernel/debug/tracing/events/s390/.../format to handle
the print fmt: statement. The kernel provides a library in
directory linux/tools/lib/traceevent/* for this reason.
This library can not handle structures or unions defined in
the TRACE_EVENT/TP_STRUCT__entry macros with __field_struct macro.
There is no possibility to extract a structure member
(which might be a bit field) since there is no packing
information nor bit field offset by parsing the printf fmt line.
Therefore rewrite the TRACE_EVENT macro and add the
__field macro for the necessary members.
Keep the __fieldstruct macro to extract the complete
structure when dumps are analysed.
Note that the same information is displayed, this is no
interface change.
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Acked-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Thomas Richter [Wed, 11 Jul 2018 08:27:12 +0000 (10:27 +0200)]
s390 cio: Rewrite trace point in s390_cio_stcrw
Tools like 'perf stat' parse the trace point format files defined
in /sys/kernel/debug/tracing/events/s390/.../format to handle
the print fmt: statement. The kernel provides a library in
directory linux/tools/lib/traceevent/* for this reason.
This library can not handle structures or unions defined in
the TRACE_EVENT/TP_STRUCT__entry macros with __field_struct macro.
There is no possibility to extract a structure member
(which might be a bit field) since there is no packing
information nor bit field offset by parsing the printf fmt line.
Therefore rewrite the TRACE_EVENT macro and add the
the __field macro for the missing members.
Keep the __fieldstruct macro to extract the complete
structure when dumps are analysed.
Note that the same information is displayed, this is no
interface change.
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Acked-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Thomas Richter [Tue, 3 Jul 2018 08:15:10 +0000 (10:15 +0200)]
s390 cio: Rewrite trace point in s390_cio_tpi
Tools like 'perf stat' parse the trace point format files defined
in /sys/kernel/debug/tracing/events/s390/.../format to handle
the print fmt: statement. The kernel provides a library in
directory linux/tools/lib/traceevent/* for this reason.
This library can not handle structures or unions defined in
the TRACE_EVENT/TP_STRUCT__entry macros with __field_struct macro.
There is no possibility to extract a structure member
(which might be a bit field) since there is no packing
information nor bit field offset by parsing the printf fmt line.
Therefore rewrite the TRACE_EVENT macro and add the
__field macro for the members adapter_IO, isc and type
of struct tpi_info.
Note that the same information is displayed, this is no
interface change.
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Acked-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Thomas Richter [Tue, 3 Jul 2018 07:41:12 +0000 (09:41 +0200)]
s390 cio: Rewrite trace point in s390_cio_interrupt
Tools like 'perf stat' parse the trace point format files defined
in /sys/kernel/debug/tracing/events/s390/.../format to handle
the print fmt: statement. The kernel provides a library in
directory linux/tools/lib/traceevent/* for this reason.
This library can not handle structures or unions defined in
the TRACE_EVENT/TP_STRUCT__entry macros with __field_struct macro.
There is no possibility to extract a structure member
(which might be a bit field) since there is no packing
information nor bit field offset by parsing the printf fmt line.
Therefore rewrite the TRACE_EVENT macro and add the
__field macro for the necessary fields.
Keep the __fieldstruct macro to extract the complete
structure when dumps are analysed.
Note that the same information is displayed, this is no
interface change.
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Acked-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Detect and report the etoken facility. With spectre_v2=auto or
CONFIG_EXPOLINE_AUTO=y automatically disable expolines and use
the full branch prediction mode for the kernel.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Sebastian Ott [Mon, 25 Jun 2018 12:25:59 +0000 (14:25 +0200)]
s390/chsc: fix packed-not-aligned warnings
Remove attribute packed where possible failing this add proper alignment
information to fix warnings like the one below:
drivers/s390/cio/chsc.c: In function 'chsc_siosl':
drivers/s390/cio/chsc.c:1287:2: warning: alignment 1 of 'struct <anonymous>' is less than 4 [-Wpacked-not-aligned]
} __attribute__ ((packed)) *siosl_area;
Note: this patch should be a nop since non of these structs use auto
storage but allocated pages. However there are changes to the generated
code because of additional padding at the end of some of the structs due
to alignment when memset(foo, 0, sizeof(*foo)) is used.
Signed-off-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Sebastian Ott [Tue, 26 Jun 2018 13:09:32 +0000 (15:09 +0200)]
s390/css: reduce stsch calls
Both css_evaluate_new_subchannel and cio_validate_subchannel used
stsch and css_sch_is_valid to check for a valid device.
Reduce stsch calls during subchannel evaluation by re-using schib
data. Also the type/devno valid information is only checked once.
Signed-off-by: Sebastian Ott <sebott@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Sebastian Ott [Mon, 25 Jun 2018 09:23:26 +0000 (11:23 +0200)]
s390/css: validate subchannel prior to allocation
In css_alloc_subchannel we allocate the subchannel and do a
validation of the subchannel (to decide if we should look for
devices via this subchannel). On a typical LPAR we find lots
of subchannels to be invalid (because there is no device
attached or the device is blacklisted) leading to lots of
useless kmalloc and kfree calls.
This patch changes the order to only allocate the subchannels
that have been found valid.
Signed-off-by: Sebastian Ott <sebott@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Sebastian Ott [Tue, 12 Jun 2018 11:56:21 +0000 (13:56 +0200)]
s390/css: start the subchannel evaluation earlier
The css bus code uses 2 initcalls: channel_subsystem_init to
initialize internal data and channel_subsystem_init_sync to
start scanning for devices and wait for it to finish.
The start scanning for devices part is moved to the first
initcall such that more work happens in parallel.
Signed-off-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Sebastian Ott [Wed, 13 Jun 2018 14:26:23 +0000 (16:26 +0200)]
s390/cio: ensure that a chpid is registered only once
Improve locking in chp_new to make sure that we don't register
the same chpid twice. Chpid registration was synchronized via
the machine check handler thread but we also have codepaths to
look for new chpids triggered independent of that thread (during
IPL or resume from hibernate).
Signed-off-by: Sebastian Ott <sebott@linux.ibm.com> Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
When the oom killer kills a userspace process in the page fault handler
while in guest context, the fault handler fails to release the mm_sem
if the FAULT_FLAG_RETRY_NOWAIT option is set. This leads to a deadlock
when tearing down the mm when the process terminates. This bug can only
happen when pfault is enabled, so only KVM clients are affected.
The problem arises in the rare cases in which handle_mm_fault does not
release the mm_sem. This patch fixes the issue by manually releasing
the mm_sem when needed.
s390/cmm: avoid add_timer on concurrently used timer
cmm_set_timer could be called concurrently from cmm_thread, cmm proc
handler, upon cmm smsg receive and timer function itself. To avoid
potential race condition and hitting BUG_ON in add_timer on already
pending timer simply reuse mod_timer which is according to
documentation "the only safe way to modify the timeout" with multiple
unserialized concurrent users. mod_timer can handle both active and
inactive timers which allows to carry out minor code simplification as
well.
Replace strncpy which is used to deliberately avoid string NUL-termination
with memcpy. This allows to get rid of gcc 8 stringop-truncation warning:
inlined from 'query_crypto_facility.constprop' at
drivers/s390/crypto/pkey_api.c:702:2:
./include/linux/string.h:246:9: warning: '__builtin_strncpy' output
truncated before terminating nul copying 8 bytes from a string of the
same length [-Wstringop-truncation]
Reviewed-by: Harald Freudenberger <freude@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
s390/cmm: split and simplify cmm pages proc handler
Split cmm_pages_handler into cmm_pages_handler and
cmm_timed_pages_handler, each handling separate proc entry. And reuse
proc_doulongvec_minmax to simplify proc handlers. Min/max values are
optional and are omitted here.
Vasily Gorbik [Sun, 24 Jun 2018 10:28:46 +0000 (12:28 +0200)]
s390/topology: correct topology mode proc handler
Reuse proc_douintvec_minmax to simplify topology mode proc handler.
This also enforces correct range of 0-1 on proc writes and correctly
handles numbers starting with 0 or spaces.
Since proc_dointvec does not perform value range control,
proc_dointvec_minmax should be used to limit value range, which is
clearly intended here, as the internal representation of the value:
Currently there are some variables in the purgatory (e.g. kernel_entry)
which are defined twice, once in assembler- and once in c-code. The reason
for this is that these variables are set during purgatory load, where
sanity checks on the corresponding Elf_Sym's are made, while they are used
in assembler-code. Thus adding a second definition in c-code is a handy
workaround to guarantee correct Elf_Sym's are created.
When the purgatory is compiled with -fcommon (default for gcc on s390) this
is no problem because both symbols are merged by the linker. However this
is not required by ISO C and when the purgatory is built with -fno-common
the linker fails with errors like
arch/s390/purgatory/purgatory.o:(.bss+0x18): multiple definition of `kernel_entry'
arch/s390/purgatory/head.o:/.../arch/s390/purgatory/head.S:230: first defined here
Thus remove the duplicate definitions and add the required size and type
information to the assembler definition. Also add -fno-common to the
command line options to prevent similar hacks in the future.
Signed-off-by: Philipp Rudo <prudo@linux.ibm.com> Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Philipp Rudo [Tue, 26 Jun 2018 16:24:52 +0000 (18:24 +0200)]
s390/purgatory: Add missing FORCE to Makefile targets
Without FORCE make does not detect changes only made to the command line
options. So object files might not be re-built even when they should be.
Fix this by adding FORCE where it is missing.
Philipp Rudo [Tue, 26 Jun 2018 10:24:30 +0000 (12:24 +0200)]
s390/purgatory: Fix crash with expoline enabled
When the kernel is built with CONFIG_EXPOLINE=y and a compiler with
indirect branch mitigation enabled the purgatory crashes. The reason for
that is that the macros defined for expoline are used in mem.S. These
macros define new sections (.text.__s390x_indirect_*) which are marked
executable. Due to the missing linker script those sections are linked to
address 0, just as the .text section. In combination with the entry point
also being at address 0 this causes the purgatory load code
(kernel/kexec_file.c: kexec_purgatory_setup_sechdrs) to update the entry
point twice. Thus the old kernel jumps to some 'random' address causing the
crash.
To fix this turn off expolines for the purgatory. There is no problem with
this in this case due to the fact that the purgatory only runs once and the
tlb is purged (diag 308) in the end.
In the quest to remove all stack VLA usage from the kernel[1], this moves
the "$#" replacement from being an argument to being inside the function,
which avoids generating VLAs.
Sebastian Ott [Tue, 3 Jul 2018 11:06:45 +0000 (13:06 +0200)]
s390/dasd: fix cast-function-type warnings
Change the tasklets parameter type to fix W=1 warnings when building
with gcc 8 like below:
drivers/s390/block/dasd.c: In function 'dasd_alloc_device':
drivers/s390/block/dasd.c:129:8: warning: cast between incompatible function types
from 'void (*)(struct dasd_device *)' to 'void (*)(long unsigned int)' [-Wcast-function-type]
(void (*)(unsigned long)) dasd_device_tasklet,
^
Signed-off-by: Sebastian Ott <sebott@linux.ibm.com> Acked-by: Jan Höppner <hoeppner@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Thu, 28 Jun 2018 17:21:36 +0000 (19:21 +0200)]
s390/decompressor: avoid packing *.o.chkbss files into startup.a
startup.a build rule packs a list of prerequisites into archive. That
didn't take into account extra prerequisites added by chkbss, so that
zero length *.o.chkbss files were also packed into the archive.
To avoid that filter only real object from prerequisites list.
Correct merging error which replaced startup.a in targets list with
non-existing setup.a. Due to missing startup.a in targets list if_changed
triggered startup.a rebuild unconditionally.
CC arch/s390/kernel/sysinfo.o
arch/s390/kernel/sysinfo.c:275:12: warning: 'sysinfo_show' defined but not used [-Wunused-function]
static int sysinfo_show(struct seq_file *m, void *v)
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Replace strncpy which has been used to copy a substring into buf
without NUL-termination with memcpy to avoid the following gcc 8
warnings:
arch/s390/tools/gen_opcode_table.c: In function ‘add_to_group.isra.4’:
arch/s390/tools/gen_opcode_table.c:260:2: warning: ‘strncpy’
output may be truncated copying 2 bytes from a string of length 19
[-Wstringop-truncation]
strncpy(group->opcode, insn->opcode, 2);
In function ‘print_opcode_table’,
inlined from ‘main’ at arch/s390/tools/gen_opcode_table.c:333:2:
arch/s390/tools/gen_opcode_table.c:286:4: warning: ‘strncpy’
output may be truncated copying 2 bytes from a string of length 19
[-Wstringop-truncation]
strncpy(opcode, insn->opcode, 2);
Since the plain vmlinux ELF file no longer carries all necessary parts
for starting up (like the entry point and decompressor), add a check
which would block boot process and encourage users to use bzImage or
arch/s390/boot/compressed/vmlinux instead.
The check relies on s390 linux entry point ABI definition, which is only
present in bzImage and arch/s390/boot/compressed/vmlinux.
The following gcc warning is issued for strncpy which is used to
deliberately avoid string NUL-termination. Reuse memcpy to avoid the
warning.
inlined from 'monwrite_diag' at drivers/s390/char/monwriter.c:64:2:
./include/linux/string.h:246:9: warning: '__builtin_strncpy' output
truncated before terminating nul copying 7 bytes from a string of the
same length [-Wstringop-truncation]
Vasily Gorbik [Thu, 28 Jun 2018 11:28:37 +0000 (13:28 +0200)]
s390/tape: fix gcc 8 stringop-truncation warning
Replace strncpy which is used to deliberately avoid string NUL-termination
with memcpy. This allows to get rid of gcc 8 stringop-truncation warning:
inlined from 'ext_to_int_kekl' at drivers/s390/char/tape_3590.c:123:2:
./include/linux/string.h:246:9: warning: '__builtin_strncpy'
output may be truncated copying 64 bytes from a string of length 64
[-Wstringop-truncation]
Vasily Gorbik [Thu, 28 Jun 2018 11:12:27 +0000 (13:12 +0200)]
s390: align struct lowcore to double page size
Aligning struct lowcore to double page size allows to get rid of this
gcc warning:
In file included from ./arch/s390/include/asm/setup.h:56,
from ./arch/s390/include/asm/page.h:36,
from ./arch/s390/include/asm/user.h:11,
from ./include/linux/user.h:1,
from ./include/linux/elfcore.h:5,
from ./include/linux/crash_core.h:6,
from ./include/linux/kexec.h:18,
from arch/s390/purgatory/purgatory.c:10:
./arch/s390/include/asm/lowcore.h:189:1: warning: alignment 1 of 'struct
lowcore' is less than 8 [-Wpacked-not-aligned]
} __packed;
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com> Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Thu, 28 Jun 2018 05:39:24 +0000 (07:39 +0200)]
s390/decompressor: correct EXCLUDE_FILE construct
The following linker construct is problematic with linkers of
binutils < 2.28:
EXCLUDE_FILE (*piggy.o) *(.rodata.*)
from 8f1732fc2a11dc of binutils:
"though the linker accepts this without complaint the
EXCLUDE_FILE part is silently ignored and has no effect."
Silent ignoring of EXCLUDE_FILE construct made .rodata.compressed be
part of .rodata, and in case of .rodata.compressed following some
unaligned data, input_len would also become unaligned.
from arch/s390/boot/compressed/vmlinux.map:
.rodata.compressed
0x0000000000012fea 0x4d57e7 arch/s390/boot/compressed/piggy.o
0x0000000000012fea input_len
0x0000000000012fee input_data
input_len is later used here:
arch/s390/boot/compressed/misc.c:113
__decompress(input_data, input_len, NULL, NULL, output, 0, NULL, error);
from what assembler generates invalid (the second operand must be aligned
on a doubleword boundary):
0x00000000000129b4 <+148>: c4 bc 00 00 03 1b lgfrl %r11,0x12fea
hence specification exception is recognized.
To avoid an issue use EXCLUDE_FILE construct which is recognized by
older linkers (since at least binutils-2_11)
*(EXCLUDE_FILE (*piggy.o) .rodata.compressed)
Also ensure that .rodata.compressed is at least doubleword aligned.
Fixes: 89b5202e81df ("s390/decompressor: support uncompressed kernel") Reported-by: Halil Pasic <pasic@linux.ibm.com> Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Wed, 27 Jun 2018 15:32:53 +0000 (17:32 +0200)]
s390: correct _stext offset
Avoid unnecessary rewrite of psw and merge _stext into
startup_continue. This allows to move _stext definition to vmlinux.lds.S,
where _etext is also defined and set _stext to the actual beginning of
.text at 0x100000.
This fixes the problem with setting the last .text page as
not-executable due to vmem_map_init relying on page alinged _stext and
_etext.
Fixes: bd79d6632958 ("s390/decompressor: trim the kernel image up to 1M") Reported-by: Nils Hoppmann <niho@de.ibm.com> Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Wed, 27 Jun 2018 15:10:14 +0000 (17:10 +0200)]
s390: get rid of the first mb of uncompressed image
Instead of generating uncompressed kernel image starting at 0, filling
first mb with zeros (with ".org 0x100000") and then trimming it off
from vmlinux.bin before compression, simply generate a kernel image
starting from 0x100000.
Vasily Gorbik [Wed, 27 Jun 2018 12:58:46 +0000 (14:58 +0200)]
s390: put expoline execute-trampolines into .text section
Currently none of the vmlinux linker script patterns in .text section
match expoline execute-trampolines, and they end up as separate
sections:
.text.__s390_indirect_jump_r1,
.text.__s390x_indirect_jump_r1use_r9,
.text.__s390x_indirect_branch_4_1use_6, ...
Vasily Gorbik [Wed, 27 Jun 2018 12:46:57 +0000 (14:46 +0200)]
s390/decompressor: discard ___kcrctab section
___kcrctab section is not used during the decompressor phase and could be
discarded to save the memory. It is currently generated due to lib/mem.S
usage, which exports few symbols.
Vasily Gorbik [Mon, 18 Jun 2018 15:51:29 +0000 (17:51 +0200)]
s390/perf: fix gcc 8 array-bounds warning
arch/s390/kernel/perf_regs.c:36:19: warning: array subscript 16 is above
array bounds of 'long unsigned int[16]' [-Warray-bounds]
return regs->gprs[idx];
gcc tries to be smart here and since there is a condition:
if (idx >= PERF_REG_S390_R0 && idx <= PERF_REG_S390_R15)
return regs->gprs[idx];
which covers all possible array subscripts, it gives the warning
for the last function return statement:
return regs->gprs[idx];
which in presence of that condition does not really make sense and
should be replaced with "return 0;"
Also move WARN_ON_ONCE((u32)idx >= PERF_REG_S390_MAX) to the end of the
function.
Reviewed-by: Hendrik Brueckner <brueckner@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Mon, 18 Jun 2018 14:50:42 +0000 (16:50 +0200)]
s390: fix gcc 8 stringop-truncation warnings in proc handlers
arch/s390/kernel/topology.c:591:3: warning: 'strncpy' output truncated
before terminating nul copying 2 bytes from a string of the same length
[-Wstringop-truncation]
strncpy(buf, topology_is_enabled() ? "1\n" : "0\n", ARRAY_SIZE(buf));
arch/s390/appldata/appldata_base.c:326:3: warning: 'strncpy' output truncated
before terminating nul copying 2 bytes from a string of the same length
[-Wstringop-truncation]
strncpy(buf, ops->active ? "1\n" : "0\n", ARRAY_SIZE(buf));
arch/s390/appldata/appldata_base.c:217:3: warning: 'strncpy' output truncated
before terminating nul copying 2 bytes from a string of the same length
[-Wstringop-truncation]
strncpy(buf, appldata_timer_active ? "1\n" : "0\n", ARRAY_SIZE(buf));
Vasily Gorbik [Sun, 17 Jun 2018 10:18:56 +0000 (12:18 +0200)]
s390/dasd: fix gcc 8 stringop-truncation warning
drivers/s390/block/dasd_alias.c:711:2: warning: 'strncpy' output truncated
before terminating nul copying 4 bytes from a string of the same length
[-Wstringop-truncation]
strncpy((char *) &cqr->magic, "ECKD", 4);
Replace strncpy to int as a buffer with memcpy.
Acked-by: Stefan Haberland <sth@linux.ibm.com> Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Sat, 16 Jun 2018 22:30:43 +0000 (00:30 +0200)]
s390/extmem: fix gcc 8 stringop-overflow warning
arch/s390/mm/extmem.c: In function '__segment_load':
arch/s390/mm/extmem.c:436:2: warning: 'strncat' specified bound 7 equals
source length [-Wstringop-overflow=]
strncat(seg->res_name, " (DCSS)", 7);
What gcc complains about here is the misuse of strncat function, which
in this case does not limit a number of bytes taken from "src", so it is
in the end the same as strcat(seg->res_name, " (DCSS)");
Keeping in mind that a res_name is 15 bytes, strncat in this case
would overflow the buffer and write 0 into alignment byte between the
fields in the struct. To avoid that increasing res_name size to 16,
and reusing strlcat.
Vasily Gorbik [Mon, 25 Jun 2018 12:30:42 +0000 (14:30 +0200)]
s390/scm_blk: correct numa_node in scm_blk_dev_setup
The numa_node field of the tag_set struct has to be explicitly
initialized, otherwise it stays as 0, which is a valid numa node id and
cause memory allocation failure if node 0 is offline.
Acked-by: Sebastian Ott <sebott@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Sun, 24 Jun 2018 07:21:59 +0000 (09:21 +0200)]
s390/dasd: correct numa_node in dasd_alloc_queue
The numa_node field of the tag_set struct has to be explicitly
initialized, otherwise it stays as 0, which is a valid numa node id and
cause memory allocation failure if node 0 is offline.
Acked-by: Stefan Haberland <sth@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
s390/zcrypt: Integrate ap_asm.h into include/asm/ap.h.
Move all the inline functions from the ap bus header
file ap_asm.h into the in-kernel api header file
arch/s390/include/asm/ap.h so that KVM can make use
of all the low level AP functions.
Signed-off-by: Harald Freudenberger <freude@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
s390/zcrypt: Show load of cards and queues in sysfs
Show the current load value of cards and queues in sysfs.
The load value for each card and queue is maintained by
the zcrypt device driver for dispatching and load
balancing requests over the available devices.
This patch provides the load value to userspace via a
new read only sysfs attribute 'load' per card and queue.
Signed-off-by: Harald Freudenberger <freude@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Tue, 12 Jun 2018 14:46:59 +0000 (16:46 +0200)]
s390/setup: do not reserve the decompressor code
Introduce PARMAREA_END, and use it for memblock reserve of low
memory, which is used for lowcore, kdump data mover code and page
buffer, early stack and parmarea. There is no need to reserve an
area between PARMAREA_END and the decompressor _ehead.
Arnd Bergmann [Mon, 18 Jun 2018 10:05:00 +0000 (12:05 +0200)]
s390: hypfs: use ktime_get_seconds() instead of get_seconds()
time_t and get_seconds() are deprecated because they will overflow on
32-bit architectures in the future. This is not a problem on 64-bit s390,
but we should use proper interfaces anyway.
Besides moving to the time64_t based interface, the CLOCK_MONOTONIC
based ktime_get_seconds() is preferred for kernel internal timekeeping
because it does not behave in unexpected ways during leap second changes
or settimeofday() calls.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Vasily Gorbik [Tue, 12 Jun 2018 22:00:25 +0000 (00:00 +0200)]
s390/decompressor: support uncompressed kernel
Implement uncompressed kernel support (when "None" is picked in kernel
compression mode list). In that case an actual decompression code is
skipped and control is passed from boot/head.S to startup_continue in
kernel/head64.S. To achieve that uncompressed kernel payload is
conditionally put at 0x100000 in bzImage.
In reality this is very close to classic uncompressed kernel "image",
but the decompressor has its own build and link process,
kernel/head64.S lives at 0x100000 rather than at 0x11000, and .bss
section is reused for both stages.
Vasily Gorbik [Tue, 12 Jun 2018 20:58:50 +0000 (22:58 +0200)]
s390/decompressor: reuse lib/mem.S for mem functions
Reusing arch/s390/lib/mem.S file solves a problem that sclp_early_core.c
and its dependencies have to be compiled with -march=z900 (no need to
compile compressed/misc.c with -march=z900). This also allows to avoid
mem functions duplicates, makes code a bit smaller and optimized mem
functions are utilized.
Re-compile ebcdic.c and sclp_early_core.c for the decompressor,
using proper decompressor CFLAGS. This also allows to potentially use
instrumentation for those files when built for the main kernel image.
With kbuild there is no easy way to re-compile a source file from
another directory. Bypass ugly rules and Makefile meta-programming
with relative path includes of original files.
Vasily Gorbik [Tue, 12 Jun 2018 13:14:51 +0000 (15:14 +0200)]
s390/als: avoid .init.* sections usage
Since als.c is the part of the decompressor only, there is no point in
using init sections for code and data. That's just creating extra
sections in the decompressor image.
Vasily Gorbik [Tue, 12 Jun 2018 13:11:40 +0000 (15:11 +0200)]
s390/boot: make head.S and als.c be part of the decompressor only
Since uncompressed kernel image does not have to be bootable anymore,
move head.S, head_kdump.S and als.c to boot/ folder and compile them
in just in the decompressor.
Vasily Gorbik [Tue, 12 Jun 2018 13:11:40 +0000 (15:11 +0200)]
s390/decompressor: trim the kernel image up to 1M
Move head64.S main kernel entry point "startup_continue" to 0x100000 and
trim everything which is below 1M during build. So, that the decompressor
would unpack the main kernel image, move it to 0x100000 and jump to
startup_continue.
Vasily Gorbik [Mon, 11 Jun 2018 12:35:01 +0000 (14:35 +0200)]
s390: remove uncompressed kernel image build
Dropping support for uncompressed kernel "image" build. Having
both image and bzImage makes it complicated to add new code to an
early boot phase (which is part of vmlinux for uncompressed kernel and
a separate arch/s390/boot/compressed/vmlinux for bzImage).
e.g. sclp_early_core.o is used for both, the decompressor phase and the
main kernel. The fact of having uncompressed kernel "image" forces us
to have a single object file and sacrifice instrumentation flags on such
files (so that we could use them early). The story gets much more
complicated with the need to utilize some of the string functions.
With bzImage only support, we have 2 separate boot stages each built
and linked separately, which allows to reuse some shared code, but
recompile with appropriate flags.
Vasily Gorbik [Fri, 15 Jun 2018 10:28:05 +0000 (12:28 +0200)]
s390/decompressor: correct build flags
The decompressor requires its own set of cc and asm flags, to avoid
building with features which do not make sense at such an early boot stage
(e.g. expoline, ftrace).
Currently cc flags are already set for the decompressor, but "cflags-y"
is not exported and hence empty. To fix that and to add asm flags, define
and export KBUILD_AFLAGS_DECOMPRESSOR and KBUILD_CFLAGS_DECOMPRESSOR
and rely on them in the decompressor's Makefile.
Vasily Gorbik [Fri, 15 Jun 2018 08:22:18 +0000 (10:22 +0200)]
s390: disable asm code expolines if cc does not support it
To avoid a mixture of asm code with expolines and c code without them,
propagate CC_USING_EXPOLINE to KBUILD_AFLAGS and use it to detect
whether asm code should have expolines or not.
Julian Wiedmann [Wed, 16 May 2018 07:37:25 +0000 (09:37 +0200)]
s390/qdio: reset old sbal_state flags
When allocating a new AOB fails, handle_outbound() is still capable of
transmitting the selected buffer (just without async completion).
But if a previous transfer on this queue slot used async completion, its
sbal_state flags field is still set to QDIO_OUTBUF_STATE_FLAG_PENDING.
So when the upper layer driver sees this stale flag, it expects an async
completion that never happens.
Fix this by unconditionally clearing the flags field.
Fixes: 104ea556ee7f ("qdio: support asynchronous delivery of storage blocks") Cc: <stable@vger.kernel.org> #v3.2+ Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
projects / mirror_ubuntu-disco-kernel.git / log