kernel.git/arch/s390/kernel, branch linux-6.10.y

s390/ftrace: Avoid calling unwinder in ftrace_return_address()

2024-10-04T14:33:47Z

commit a84dd0d8ae24bdc6da341187fc4c1a0adfce2ccc upstream. ftrace_return_address() is called extremely often from performance-critical code paths when debugging features like CONFIG_TRACE_IRQFLAGS are enabled. For example, with debug_defconfig, ftrace selftests on my LPAR currently execute ftrace_return_address() as follows: ftrace_return_address(0) - 0 times (common code uses __builtin_return_address(0) instead) ftrace_return_address(1) - 2,986,805,401 times (with this patch applied) ftrace_return_address(2) - 140 times ftrace_return_address(>2) - 0 times The use of __builtin_return_address(n) was replaced by return_address() with an unwinder call by commit cae74ba8c295 ("s390/ftrace: Use unwinder instead of __builtin_return_address()") because __builtin_return_address(n) simply walks the stack backchain and doesn't check for reaching the stack top. For shallow stacks with fewer than "n" frames, this results in reads at low addresses and random memory accesses. While calling the fully functional unwinder "works", it is very slow for this purpose. Moreover, potentially following stack switches and walking past IRQ context is simply wrong thing to do for ftrace_return_address(). Reimplement return_address() to essentially be __builtin_return_address(n) with checks for reaching the stack top. Since the ftrace_return_address(n) argument is always a constant, keep the implementation in the header, allowing both GCC and Clang to unroll the loop and optimize it to the bare minimum. Fixes: cae74ba8c295 ("s390/ftrace: Use unwinder instead of __builtin_return_address()") Cc: stable@vger.kernel.org Reported-by: Sumanth Korikkar Reviewed-by: Heiko Carstens Acked-by: Sumanth Korikkar Signed-off-by: Vasily Gorbik Signed-off-by: Greg Kroah-Hartman

s390/vmlinux.lds.S: Move ro_after_init section behind rodata section

2024-09-12T09:13:04Z

[ Upstream commit 75c10d5377d8821efafed32e4d72068d9c1f8ec0 ] The .data.rel.ro and .got section were added between the rodata and ro_after_init data section, which adds an RW mapping in between all RO mapping of the kernel image: ---[ Kernel Image Start ]--- 0x000003ffe0000000-0x000003ffe0e00000 14M PMD RO X 0x000003ffe0e00000-0x000003ffe0ec7000 796K PTE RO X 0x000003ffe0ec7000-0x000003ffe0f00000 228K PTE RO NX 0x000003ffe0f00000-0x000003ffe1300000 4M PMD RO NX 0x000003ffe1300000-0x000003ffe1331000 196K PTE RO NX 0x000003ffe1331000-0x000003ffe13b3000 520K PTE RW NX <--- 0x000003ffe13b3000-0x000003ffe13d5000 136K PTE RO NX 0x000003ffe13d5000-0x000003ffe1400000 172K PTE RW NX 0x000003ffe1400000-0x000003ffe1500000 1M PMD RW NX 0x000003ffe1500000-0x000003ffe1700000 2M PTE RW NX 0x000003ffe1700000-0x000003ffe1800000 1M PMD RW NX 0x000003ffe1800000-0x000003ffe187e000 504K PTE RW NX ---[ Kernel Image End ]--- Move the ro_after_init data section again right behind the rodata section to prevent interleaving RO and RW mappings: ---[ Kernel Image Start ]--- 0x000003ffe0000000-0x000003ffe0e00000 14M PMD RO X 0x000003ffe0e00000-0x000003ffe0ec7000 796K PTE RO X 0x000003ffe0ec7000-0x000003ffe0f00000 228K PTE RO NX 0x000003ffe0f00000-0x000003ffe1300000 4M PMD RO NX 0x000003ffe1300000-0x000003ffe1353000 332K PTE RO NX 0x000003ffe1353000-0x000003ffe1400000 692K PTE RW NX 0x000003ffe1400000-0x000003ffe1500000 1M PMD RW NX 0x000003ffe1500000-0x000003ffe1700000 2M PTE RW NX 0x000003ffe1700000-0x000003ffe1800000 1M PMD RW NX 0x000003ffe1800000-0x000003ffe187e000 504K PTE RW NX ---[ Kernel Image End ]--- Reviewed-by: Alexander Gordeev Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik Signed-off-by: Sasha Levin

s390/boot: Fix KASLR base offset off by __START_KERNEL bytes

2024-08-29T15:36:06Z

[ Upstream commit 1642285e511c2a40b14e87a41aa8feace6123036 ] Symbol offsets to the KASLR base do not match symbol address in the vmlinux image. That is the result of setting the KASLR base to the beginning of .text section as result of an optimization. Revert that optimization and allocate virtual memory for the whole kernel image including __START_KERNEL bytes as per the linker script. That allows keeping the semantics of the KASLR base offset in sync with other architectures. Rename __START_KERNEL to TEXT_OFFSET, since it represents the offset of the .text section within the kernel image, rather than a virtual address. Still skip mapping TEXT_OFFSET bytes to save memory on pgtables and provoke exceptions in case an attempt to access this area is made, as no kernel symbol may reside there. In case CONFIG_KASAN is enabled the location counter might exceed the value of TEXT_OFFSET, while the decompressor linker script forcefully resets it to TEXT_OFFSET, which leads to a sections overlap link failure. Use MAX() expression to avoid that. Reported-by: Omar Sandoval Closes: https://lore.kernel.org/linux-s390/ZnS8dycxhtXBZVky@telecaster.dhcp.thefacebook.com/ Fixes: 56b1069c40c7 ("s390/boot: Rework deployment of the kernel image") Signed-off-by: Alexander Gordeev Acked-by: Vasily Gorbik Signed-off-by: Vasily Gorbik Signed-off-by: Sasha Levin

s390/fpu: Re-add exception handling in load_fpu_state()

2024-08-11T10:57:57Z

commit 4734406c39238cbeafe66f0060084caa3247ff53 upstream. With the recent rewrite of the fpu code exception handling for the lfpc instruction within load_fpu_state() was erroneously removed. Add it again to prevent that loading invalid floating point register values cause an unhandled specification exception. Fixes: 8c09871a950a ("s390/fpu: limit save and restore to used registers") Cc: stable@vger.kernel.org Reported-by: Aristeu Rozanski Tested-by: Aristeu Rozanski Reviewed-by: Vasily Gorbik Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik Signed-off-by: Greg Kroah-Hartman

s390/cpum_cf: Fix endless loop in CF_DIAG event stop

2024-08-03T07:01:07Z

[ Upstream commit e6ce1f12d777f6ee22b20e10ae6a771e7e6f44f5 ] Event CF_DIAG reads out complete counter sets using stcctm instruction. This is done at event start time when the process starts execution and at event stop time when the process is removed from the CPU. During removal the difference of each counter in the counter sets is calculated and saved as raw data in the ring buffer. This works fine unless the number of counters in a counter set is zero. This may happen for the extended counter set. This set is machine specific and the size of the counter set can be zero even when extended counter set is authorized for read access. This case is not handled. cfdiag_diffctr() checks authorization of the extended counter set. If true the functions assumes the extended counter set has been saved in a data buffer. However this is not the case, cfdiag_getctrset() does not save a counter set with counter set size of zero. This mismatch causes an endless loop in the counter set readout during event stop handling. The calculation of the difference of the counters in each counter now verifies the size of the counter set is non-zero. A counter set with size zero is skipped. Fixes: a029a4eab39e ("s390/cpumf: Allow concurrent access for CPU Measurement Counter Facility") Signed-off-by: Thomas Richter Acked-by: Sumanth Korikkar Acked-by: Heiko Carstens Cc: Heiko Carstens Cc: Vasily Gorbik Cc: Alexander Gordeev Signed-off-by: Vasily Gorbik Signed-off-by: Sasha Levin

s390/setup: Fix pa/va for modules under non-GPL licenses

2024-08-03T07:01:07Z

[ Upstream commit e188e5d5ffd01d484b5255b88739fcf67b300223 ] The struct vm_layout contains fields used in __pa/__va calculations. Such fundamental things have to be exported with EXPORT_SYMBOL to avoid breakages of out-of-tree modules under non-GPL licenses. Fixes: 7de0446f0b26 ("s390/boot: Make identity mapping base address explicit") Acked-by: Heiko Carstens Acked-by: Alexander Gordeev Signed-off-by: Vasily Gorbik Signed-off-by: Sasha Levin

s390/uv: Don't call folio_wait_writeback() without a folio reference

2024-08-03T06:59:53Z

[ Upstream commit 3f29f6537f54d74e64bac0a390fb2e26da25800d ] folio_wait_writeback() requires that no spinlocks are held and that a folio reference is held, as documented. After we dropped the PTL, the folio could get freed concurrently. So grab a temporary reference. Fixes: 214d9bbcd3a6 ("s390/mm: provide memory management functions for protected KVM guests") Reviewed-by: Claudio Imbrenda Signed-off-by: David Hildenbrand Link: https://lore.kernel.org/r/20240508182955.358628-2-david@redhat.com Signed-off-by: Heiko Carstens Signed-off-by: Alexander Gordeev Signed-off-by: Sasha Levin

s390: remove native mmap2() syscall

2024-06-25T13:57:37Z

The mmap2() syscall has never been used on 64-bit s390x and should have been removed as part of 5a79859ae0f3 ("s390: remove 31 bit support"). Remove it now. Acked-by: Heiko Carstens Signed-off-by: Arnd Bergmann

syscalls: fix compat_sys_io_pgetevents_time64 usage

2024-06-25T13:57:20Z

Using sys_io_pgetevents() as the entry point for compat mode tasks works almost correctly, but misses the sign extension for the min_nr and nr arguments. This was addressed on parisc by switching to compat_sys_io_pgetevents_time64() in commit 6431e92fc827 ("parisc: io_pgetevents_time64() needs compat syscall in 32-bit compat mode"), as well as by using more sophisticated system call wrappers on x86 and s390. However, arm64, mips, powerpc, sparc and riscv still have the same bug. Change all of them over to use compat_sys_io_pgetevents_time64() like parisc already does. This was clearly the intention when the function was originally added, but it got hooked up incorrectly in the tables. Cc: stable@vger.kernel.org Fixes: 48166e6ea47d ("y2038: add 64-bit time_t syscalls to all 32-bit architectures") Acked-by: Heiko Carstens # s390 Signed-off-by: Arnd Bergmann

s390/crash: Do not use VM info if os_info does not have it

2024-06-05T15:03:24Z

The virtual memory information stored in os_info area is required for creation of the kernel image PT_LOAD program header for kernels since commit a2ec5bec56dd ("s390/mm: uncouple physical vs virtual address spaces"). By contrast, if such information in os_info is absent the PT_LOAD program header should not be created. Currently the proper PT_LOAD program header is created for kernels that contain the virtual memory information, but for kernels without one an invalid header of zero size is created. That in turn leads to stand-alone dump failures. Use OS_INFO_KASLR_OFFSET variable to check whether os_info is present or not (same as crash and makedumpfile tools do) and based on that create or do not create the kernel image PT_LOAD program header. Fixes: f4cac27dc0d6 ("s390/crash: Use old os_info to create PT_LOAD headers") Tested-by: Mikhail Zaslonko Acked-by: Mikhail Zaslonko Acked-by: Heiko Carstens Signed-off-by: Alexander Gordeev