kernel/linux.git/arch, branch v4.18.17

sparc: Throttle perf events properly.

2018-11-04T13:51:56+00:00

[ Upstream commit 455adb3174d2c8518cef1a61140c211f6ac224d2 ] Like x86 and arm, call perf_sample_event_took() in perf event NMI interrupt handler. Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

sparc: Fix syscall fallback bugs in VDSO.

2018-11-04T13:51:56+00:00

[ Upstream commit 776ca1543b5fe673aaf1beb244fcc2429d378083 ] First, the trap number for 32-bit syscalls is 0x10. Also, only negate the return value when syscall error is indicated by the carry bit being set. Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

sparc: Fix single-pcr perf event counter management.

2018-11-04T13:51:56+00:00

[ Upstream commit cfdc3170d214046b9509183fe9b9544dc644d40b ] It is important to clear the hw->state value for non-stopped events when they are added into the PMU. Otherwise when the event is scheduled out, we won't read the counter because HES_UPTODATE is still set. This breaks 'perf stat' and similar use cases, causing all the events to show zero. This worked for multi-pcr because we make explicit sparc_pmu_start() calls in calculate_multiple_pcrs(). calculate_single_pcr() doesn't do this because the idea there is to accumulate all of the counter settings into the single pcr value. So we have to add explicit hw->state handling there. Like x86, we use the PERF_HES_ARCH bit to track truly stopped events so that we don't accidently start them on a reload. Related to all of this, sparc_pmu_start() is missing a userpage update so add it. Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

sparc64: Wire up compat getpeername and getsockname.

2018-11-04T13:51:55+00:00

[ Upstream commit 1f2b5b8e2df4591fbca430aff9c5a072dcc0f408 ] Fixes: 8b30ca73b7cc ("sparc: Add all necessary direct socket system calls.") Reported-by: Joseph Myers Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

sparc64: Set %l4 properly on trap return after handling signals.

2018-11-04T13:51:55+00:00

[ Upstream commit d1f1f98c6d1708a90436e1a3b2aff5e93946731b ] If we did some signal processing, we have to reload the pt_regs tstate register because it's value may have changed. In doing so we also have to extract the %pil value contained in there anre load that into %l4. This value is at bit 20 and thus needs to be shifted down before we later write it into the %pil register. Most of the time this is harmless as we are returning to userspace and the %pil is zero for that case. Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

sparc64: Make proc_id signed.

2018-11-04T13:51:55+00:00

[ Upstream commit b3e1eb8e7ac9aaa283989496651d99267c4cad6c ] So that when it is unset, ie. '-1', userspace can see it properly. Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

sparc64: Make corrupted user stacks more debuggable.

2018-11-04T13:51:55+00:00

[ Upstream commit 5b4fc3882a649c9411dd0dcad2ddb78e911d340e ] Right now if we get a corrupted user stack frame we do a do_exit(SIGILL) which is not helpful. If under a debugger, this behavior causes the inferior process to exit. So the register and other state cannot be examined at the time of the event. Instead, conditionally log a rate limited kernel log message and then force a SIGSEGV. With bits and ideas borrowed (as usual) from powerpc. Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

sparc64: Export __node_distance.

2018-11-04T13:51:55+00:00

[ Upstream commit 2b4792eaa9f553764047d157365ed8b7787751a3 ] Some drivers reference it via node_distance(), for example the NVME host driver core. ERROR: "__node_distance" [drivers/nvme/host/nvme-core.ko] undefined! make[1]: *** [scripts/Makefile.modpost:92: __modpost] Error 1 Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

vmlinux.lds.h: Fix incomplete .text.exit discards

2018-11-04T13:51:51+00:00

[ Upstream commit 8dcf86caa1e3daf4a6ccf38e97f4f752b411f829 ] Enabling CONFIG_GCOV_PROFILE_ALL=y causes linker errors on ARM: `.text.exit' referenced in section `.ARM.exidx.text.exit': defined in discarded section `.text.exit' `.text.exit' referenced in section `.fini_array.00100': defined in discarded section `.text.exit' And related errors on NDS32: `.text.exit' referenced in section `.dtors.65435': defined in discarded section `.text.exit' The gcov compiler flags cause certain compiler versions to generate additional destructor-related sections that are not yet handled by the linker script, resulting in references between discarded and non-discarded sections. Since destructors are not used in the Linux kernel, fix this by discarding these additional sections. Reported-by: Arnd Bergmann Tested-by: Arnd Bergmann Acked-by: Arnd Bergmann Reported-by: Greentime Hu Tested-by: Masami Hiramatsu Signed-off-by: Peter Oberparleiter Signed-off-by: Stephen Rothwell Signed-off-by: Sasha Levin

KVM: x86: support CONFIG_KVM_AMD=y with CONFIG_CRYPTO_DEV_CCP_DD=m

2018-11-04T13:51:50+00:00

[ Upstream commit 853c110982eaff0d99dace3f66f1ba58b5bfd9d5 ] SEV requires access to the AMD cryptographic device APIs, and this does not work when KVM is builtin and the crypto driver is a module. Actually the Kconfig conditions for CONFIG_KVM_AMD_SEV try to disable SEV in that case, but it does not work because the actual crypto calls are not culled, only sev_hardware_setup() is. This patch adds two CONFIG_KVM_AMD_SEV checks that gate all the remaining SEV code; it fixes this particular configuration, and drops 5 KiB of code when CONFIG_KVM_AMD_SEV=n. Reported-by: Guenter Roeck Signed-off-by: Paolo Bonzini Signed-off-by: Sasha Levin