linux-stable.git/arch, branch v5.8.2

s390/gmap: improve THP splitting

2020-08-19T06:27:09+00:00

commit ba925fa35057a062ac98c3e8138b013ce4ce351c upstream.

During s390_enable_sie(), we need to take care of splitting all qemu user
process THP mappings. This is currently done with follow_page(FOLL_SPLIT),
by simply iterating over all vma ranges, with PAGE_SIZE increment.

This logic is sub-optimal and can result in a lot of unnecessary overhead,
especially when using qemu and ASAN with large shadow map. Ilya reported
significant system slow-down with one CPU busy for a long time and overall
unresponsiveness.

Fix this by using walk_page_vma() and directly calling split_huge_pmd()
only for present pmds, which greatly reduces overhead.

Cc:  # v5.4+
Reported-by: Ilya Leoshkevich 
Tested-by: Ilya Leoshkevich 
Acked-by: Christian Borntraeger 
Signed-off-by: Gerald Schaefer 
Signed-off-by: Heiko Carstens 
Signed-off-by: Greg Kroah-Hartman

s390/numa: set node distance to LOCAL_DISTANCE

2020-08-19T06:27:09+00:00

commit 535e4fc623fab2e09a0653fc3a3e17f382ad0251 upstream.

The node distance is hardcoded to 0, which causes a trouble
for some user-level applications. In particular, "libnuma"
expects the distance of a node to itself as LOCAL_DISTANCE.
This update removes the offending node distance override.

Cc:  # 4.4
Fixes: 3a368f742da1 ("s390/numa: add core infrastructure")
Signed-off-by: Alexander Gordeev 
Signed-off-by: Heiko Carstens 
Signed-off-by: Greg Kroah-Hartman

irqdomain/treewide: Free firmware node after domain removal

2020-08-19T06:27:08+00:00

commit ec0160891e387f4771f953b888b1fe951398e5d9 upstream.

Commit 711419e504eb ("irqdomain: Add the missing assignment of
domain->fwnode for named fwnode") unintentionally caused a dangling pointer
page fault issue on firmware nodes that were freed after IRQ domain
allocation. Commit e3beca48a45b fixed that dangling pointer issue by only
freeing the firmware node after an IRQ domain allocation failure. That fix
no longer frees the firmware node immediately, but leaves the firmware node
allocated after the domain is removed.

The firmware node must be kept around through irq_domain_remove, but should be
freed it afterwards.

Add the missing free operations after domain removal where where appropriate.

Fixes: e3beca48a45b ("irqdomain/treewide: Keep firmware node unconditionally allocated")
Signed-off-by: Jon Derrick 
Signed-off-by: Thomas Gleixner 
Reviewed-by: Andy Shevchenko 
Acked-by: Bjorn Helgaas 	# drivers/pci
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/1595363169-7157-1-git-send-email-jonathan.derrick@intel.com
Signed-off-by: Greg Kroah-Hartman

ARM: 8992/1: Fix unwind_frame for clang-built kernels

2020-08-19T06:27:08+00:00

commit b4d5ec9b39f8b31d98f65bc5577b5d15d93795d7 upstream.

Since clang does not push pc and sp in function prologues, the current
implementation of unwind_frame does not work. By using the previous
frame's lr/fp instead of saved pc/sp we get valid unwinds on clang-built
kernels.

The bounds check on next frame pointer must be changed as well since
there are 8 less bytes between frames.

This fixes /proc//stack.

Link: https://github.com/ClangBuiltLinux/linux/issues/912

Reported-by: Miles Chen 
Tested-by: Miles Chen 
Cc: stable@vger.kernel.org
Reviewed-by: Nick Desaulniers 
Signed-off-by: Nathan Huckleberry 
Signed-off-by: Russell King 
Signed-off-by: Greg Kroah-Hartman

ARM: dts: exynos: Extend all Exynos5800 A15's OPPs with max voltage data

2020-08-19T06:27:07+00:00

commit d644853ff8fcbb7a4e3757f9d8ccc39d930b7e3c upstream.

On Exynos5422/5800 the regulator supply for the A15 cores ("vdd_arm") is
coupled with the regulator supply for the SoC internal circuits
("vdd_int"), thus all operating points that modify one of those supplies
have to specify a triplet of the min/target/max values to properly work
with regulator coupling.

Fixes: eaffc4de16c6 ("ARM: dts: exynos: Add missing CPU frequencies for Exynos5422/5800")
Cc: 
Signed-off-by: Marek Szyprowski 
Signed-off-by: Krzysztof Kozlowski 
Signed-off-by: Greg Kroah-Hartman

parisc: Implement __smp_store_release and __smp_load_acquire barriers

2020-08-19T06:27:07+00:00

commit e96ebd589debd9a6a793608c4ec7019c38785dea upstream.

This patch implements the __smp_store_release and __smp_load_acquire barriers
using ordered stores and loads.  This avoids the sync instruction present in
the generic implementation.

Cc:  # 4.14+
Signed-off-by: Dave Anglin 
Signed-off-by: Helge Deller 
Signed-off-by: Greg Kroah-Hartman

parisc: Do not use an ordered store in pa_tlb_lock()

2020-08-19T06:27:07+00:00

commit e72b23dec1da5e62a0090c5da1d926778284e230 upstream.

No need to use an ordered store in pa_tlb_lock() and update the comment
regarng usage of the sid register to unlocak a spinlock in
tlb_unlock0().

Signed-off-by: John David Anglin 
Signed-off-by: Helge Deller 
Cc:  # v5.0+
Signed-off-by: Greg Kroah-Hartman

Revert "parisc: Revert "Release spinlocks using ordered store""

2020-08-19T06:27:07+00:00

commit 157e9afcc4fa25068b0e8743bc254a9b56010e13 upstream.

This reverts commit 86d4d068df573a8c2105554624796c086d6bec3d.

Signed-off-by: Helge Deller 
Cc:  # v5.0+
Signed-off-by: Greg Kroah-Hartman

Revert "parisc: Use ldcw instruction for SMP spinlock release barrier"

2020-08-19T06:27:07+00:00

commit 6e9f06ee6c9566f3606d93182ac8f803a148504b upstream.

This reverts commit 9e5c602186a692a7e848c0da17aed40f49d30519.
No need to use the ldcw instruction as SMP spinlock release barrier.
Revert it to gain back speed again.

Signed-off-by: Helge Deller 
Cc:  # v5.2+
Signed-off-by: Greg Kroah-Hartman

Revert "parisc: Drop LDCW barrier in CAS code when running UP"

2020-08-19T06:27:07+00:00

commit 462fb756c7de1ffe5bc6099149136031c2d9c02a upstream.

This reverts commit e6eb5fe9123f05dcbf339ae5c0b6d32fcc0685d5.
We need to optimize it differently. A follow up patch will correct it.

Signed-off-by: Helge Deller 
Cc:  # v5.2+
Signed-off-by: Greg Kroah-Hartman