linux.git/arch/s390, branch v6.13

KVM: s390: Reject KVM_SET_GSI_ROUTING on ucontrol VMs

2025-01-07T15:36:11+00:00

Prevent null pointer dereference when processing
KVM_IRQ_ROUTING_S390_ADAPTER routing entries.
The ioctl cannot be processed for ucontrol VMs.

Fixes: f65470661f36 ("KVM: s390/interrupt: do not pin adapter interrupt pages")
Signed-off-by: Christoph Schlameuss 
Tested-by: Hariharan Mari 
Reviewed-by: Hariharan Mari 
Reviewed-by: Claudio Imbrenda 
Link: https://lore.kernel.org/r/20241216092140.329196-4-schlameuss@linux.ibm.com
Message-ID: <20241216092140.329196-4-schlameuss@linux.ibm.com>
Signed-off-by: Claudio Imbrenda

KVM: s390: Reject setting flic pfault attributes on ucontrol VMs

2025-01-07T15:36:10+00:00

Prevent null pointer dereference when processing the
KVM_DEV_FLIC_APF_ENABLE and KVM_DEV_FLIC_APF_DISABLE_WAIT ioctls in the
interrupt controller.

Fixes: 3c038e6be0e2 ("KVM: async_pf: Async page fault support on s390")
Reported-by: Claudio Imbrenda 
Signed-off-by: Christoph Schlameuss 
Reviewed-by: Hariharan Mari 
Reviewed-by: Claudio Imbrenda 
Link: https://lore.kernel.org/r/20241216092140.329196-2-schlameuss@linux.ibm.com
Message-ID: <20241216092140.329196-2-schlameuss@linux.ibm.com>
Signed-off-by: Claudio Imbrenda

KVM: s390: vsie: fix virtual/physical address in unpin_scb()

2025-01-07T15:36:10+00:00

In commit 77b533411595 ("KVM: s390: VSIE: sort out virtual/physical
address in pin_guest_page"), only pin_scb() has been updated. This
means that in unpin_scb() a virtual address was still used directly as
physical address without conversion. The resulting physical address is
obviously wrong and most of the time also invalid.

Since commit d0ef8d9fbebe ("KVM: s390: Use kvm_release_page_dirty() to
unpin "struct page" memory"), unpin_guest_page() will directly use
kvm_release_page_dirty(), instead of kvm_release_pfn_dirty(), which has
since been removed.

One of the checks that were performed by kvm_release_pfn_dirty() was to
verify whether the page was valid at all, and silently return
successfully without doing anything if the page was invalid.

When kvm_release_pfn_dirty() was still used, the invalid page was thus
silently ignored. Now the check is gone and the result is an Oops.
This also means that when running with a V!=R kernel, the page was not
released, causing a leak.

The solution is simply to add the missing virt_to_phys().

Fixes: 77b533411595 ("KVM: s390: VSIE: sort out virtual/physical address in pin_guest_page")
Reviewed-by: Janosch Frank 
Reviewed-by: Nico Boehr 
Link: https://lore.kernel.org/r/20241210083948.23963-1-imbrenda@linux.ibm.com
Signed-off-by: Claudio Imbrenda 
Message-ID: <20241210083948.23963-1-imbrenda@linux.ibm.com>

s390/mm: Consider KMSAN modules metadata for paging levels

2024-12-15T22:35:09+00:00

The calculation determining whether to use three- or four-level paging
didn't account for KMSAN modules metadata. Include this metadata in the
virtual memory size calculation to ensure correct paging mode selection
and avoiding potentially unnecessary physical memory size limitations.

Fixes: 65ca73f9fb36 ("s390/mm: define KMSAN metadata for vmalloc and modules")
Acked-by: Heiko Carstens 
Reviewed-by: Alexander Gordeev 
Reviewed-by: Ilya Leoshkevich 
Signed-off-by: Vasily Gorbik 
Signed-off-by: Alexander Gordeev

s390/ipl: Fix never less than zero warning

2024-12-11T17:05:49+00:00

DEFINE_IPL_ATTR_STR_RW() macro produces "unsigned 'len' is never less
than zero." warning when sys_vmcmd_on_*_store() callbacks are defined.

Reported-by: kernel test robot 
Closes: https://lore.kernel.org/oe-kbuild-all/202412081614.5uel8F6W-lkp@intel.com/
Fixes: 247576bf624a ("s390/ipl: Do not accept z/VM CP diag X'008' cmds longer than max length")
Reviewed-by: Heiko Carstens 
Signed-off-by: Alexander Gordeev

s390/mm: Fix DirectMap accounting

2024-12-10T14:40:41+00:00

With uncoupling of physical and virtual address spaces population of
the identity mapping was changed to use the type POPULATE_IDENTITY
instead of POPULATE_DIRECT. This breaks DirectMap accounting:

> cat /proc/meminfo
DirectMap4k:       55296 kB
DirectMap1M:    18446744073709496320 kB

Adjust all locations of update_page_count() in vmem.c to use
POPULATE_IDENTITY instead of POPULATE_DIRECT as well. With this
accounting is correct again:

> cat /proc/meminfo
DirectMap4k:       54264 kB
DirectMap1M:     8334336 kB

Fixes: c98d2ecae08f ("s390/mm: Uncouple physical vs virtual address spaces")
Cc: stable@vger.kernel.org
Reviewed-by: Alexander Gordeev 
Signed-off-by: Heiko Carstens 
Signed-off-by: Alexander Gordeev

module: Convert symbol namespace to string literal

2024-12-02T19:34:44+00:00

Clean up the existing export namespace code along the same lines of
commit 33def8498fdd ("treewide: Convert macro and uses of __section(foo)
to __section("foo")") and for the same reason, it is not desired for the
namespace argument to be a macro expansion itself.

Scripted using

  git grep -l -e MODULE_IMPORT_NS -e EXPORT_SYMBOL_NS | while read file;
  do
    awk -i inplace '
      /^#define EXPORT_SYMBOL_NS/ {
        gsub(/__stringify\(ns\)/, "ns");
        print;
        next;
      }
      /^#define MODULE_IMPORT_NS/ {
        gsub(/__stringify\(ns\)/, "ns");
        print;
        next;
      }
      /MODULE_IMPORT_NS/ {
        $0 = gensub(/MODULE_IMPORT_NS\(([^)]*)\)/, "MODULE_IMPORT_NS(\"\\1\")", "g");
      }
      /EXPORT_SYMBOL_NS/ {
        if ($0 ~ /(EXPORT_SYMBOL_NS[^(]*)\(([^,]+),/) {
  	if ($0 !~ /(EXPORT_SYMBOL_NS[^(]*)\(([^,]+), ([^)]+)\)/ &&
  	    $0 !~ /(EXPORT_SYMBOL_NS[^(]*)\(\)/ &&
  	    $0 !~ /^my/) {
  	  getline line;
  	  gsub(/[[:space:]]*\\$/, "");
  	  gsub(/[[:space:]]/, "", line);
  	  $0 = $0 " " line;
  	}

  	$0 = gensub(/(EXPORT_SYMBOL_NS[^(]*)\(([^,]+), ([^)]+)\)/,
  		    "\\1(\\2, \"\\3\")", "g");
        }
      }
      { print }' $file;
  done

Requested-by: Masahiro Yamada 
Signed-off-by: Peter Zijlstra (Intel) 
Link: https://mail.google.com/mail/u/2/#inbox/FMfcgzQXKWgMmjdFwwdsfgxzKpVHWPlc
Acked-by: Greg KH 
Signed-off-by: Linus Torvalds

Merge tag 's390-6.13-2' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux

2024-11-29T18:40:52+00:00

Pull more s390 updates from Heiko Carstens:

 - Add swap entry for hugetlbfs support

 - Add PTE_MARKER support for hugetlbs mappings; this fixes a regression
   (possible page fault loop) which was introduced when support for
   UFFDIO_POISON for hugetlbfs was added

 - Add ARCH_HAS_PREEMPT_LAZY and PREEMPT_DYNAMIC support

 - Mark IRQ entries in entry code, so that stack tracers can filter out
   the non-IRQ parts of stack traces. This fixes stack depot capacity
   limit warnings, since without filtering the number of unique stack
   traces is huge

 - In PCI code fix leak of struct zpci_dev object, and fix potential
   double remove of hotplug slot

 - Fix pagefault_disable() / pagefault_enable() unbalance in
   arch_stack_user_walk_common()

 - A couple of inline assembly optimizations, more cmpxchg() to
   try_cmpxchg() conversions, and removal of usages of xchg() and
   cmpxchg() on one and two byte memory areas

 - Various other small improvements and cleanups

* tag 's390-6.13-2' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (27 commits)
  Revert "s390/mm: Allow large pages for KASAN shadow mapping"
  s390/spinlock: Use flag output constraint for arch_cmpxchg_niai8()
  s390/spinlock: Use R constraint for arch_load_niai4()
  s390/spinlock: Generate shorter code for arch_spin_unlock()
  s390/spinlock: Remove condition code clobber from arch_spin_unlock()
  s390/spinlock: Use symbolic names in inline assemblies
  s390: Support PREEMPT_DYNAMIC
  s390/pci: Fix potential double remove of hotplug slot
  s390/pci: Fix leak of struct zpci_dev when zpci_add_device() fails
  s390/mm/hugetlbfs: Add missing includes
  s390/mm: Add PTE_MARKER support for hugetlbfs mappings
  s390/mm: Introduce region-third and segment table swap entries
  s390/mm: Introduce region-third and segment table entry present bits
  s390/mm: Rearrange region-third and segment table entry SW bits
  KVM: s390: Increase size of union sca_utility to four bytes
  KVM: s390: Remove one byte cmpxchg() usage
  KVM: s390: Use try_cmpxchg() instead of cmpxchg() loops
  s390/ap: Replace xchg() with WRITE_ONCE()
  s390/mm: Allow large pages for KASAN shadow mapping
  s390: Add ARCH_HAS_PREEMPT_LAZY support
  ...

Revert "s390/mm: Allow large pages for KASAN shadow mapping"

2024-11-29T12:41:11+00:00

This reverts commit ff123eb7741638d55abf82fac090bb3a543c1e74.

Allowing large pages for KASAN shadow mappings isn't inherently wrong,
but adding POPULATE_KASAN_MAP_SHADOW to large_allowed() exposes an issue
in can_large_pud() and can_large_pmd().

Since commit d8073dc6bc04 ("s390/mm: Allow large pages only for aligned
physical addresses"), both can_large_pud() and can_large_pmd() call _pa()
to check if large page physical addresses are aligned. However, _pa()
has a side effect: it allocates memory in POPULATE_KASAN_MAP_SHADOW
mode. This results in massive memory leaks.

The proper fix would be to address both large_allowed() and _pa()'s side
effects, but for now, revert this change to avoid the leaks.

Fixes: ff123eb77416 ("s390/mm: Allow large pages for KASAN shadow mapping")
Signed-off-by: Vasily Gorbik 
Signed-off-by: Heiko Carstens

s390/spinlock: Use flag output constraint for arch_cmpxchg_niai8()

2024-11-28T13:14:34+00:00

Add a new variant of arch_cmpxchg_niai8() which makes use of the flag
output constraint, which allows the compiler to generate slightly better
code. Also rename arch_cmpxchg_niai8() to arch_try_cmpxchg_niai8() which
reflects the purpose of the function and makes it consistent with other
"try" variants.

Signed-off-by: Heiko Carstens