kernel.git/arch/s390/include/asm/pgtable.h, branch linux-6.15.y

s390: Remove 2k vs 4k page table leftovers

2025-03-18T16:13:05Z

Since commit d08d4e7cd6bf ("s390/mm: use full 4KB page for 2KB PTE") always 4k page tables are allocated, however there is still some (now) obsolete code left which deals with switching from 2k to 4k page tables for qemu/kvm processes. Reviewed-by: Alexander Gordeev Remove the not needed code. Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik

Merge branch 'strict-mm-typechecks-support' into features

2025-03-11T14:29:44Z

Heiko writes: "The recent large kernel Rust thread where Linus commented about that structures may be returned in registers [1] made me again aware that this is not true for s390 where the ABI defines that structures are returned in a return value buffer allocated by the caller. This was also mentioned by Alexander Gordeev a couple of weeks ago. In theory the -freg-struct-return compiler flag would allow to return small structures in registers, however that has not been implemented for s390. Juergen Christ did an experimental gcc implementation which shows the benefit of such a change (bloat-o-meter): add/remove: 3/2 grow/shrink: 12/441 up/down: 740/-7182 (-6442) This result is not very impressive, and doesn't seem to justify a new ABI for the kernel. However there is still the existing STRICT_MM_TYPECHECKS which can be used to change some mm types from structures to simple scalar types. Changing the mm types results in: add/remove: 2/8 grow/shrink: 25/116 up/down: 3902/-6204 (-2302) Which is already a third of the possible savings which would be the result of the described ABI change. Therefore add support for a configurable STRICT_MM_TYPECHECKS which allows to generate better code, but also allows to have type checking for debug builds." [1] https://lore.kernel.org/all/CAHk-=wgb1g9VVHRaAnJjrfRFWAOVT2ouNOMqt0js8h3D6zvHDw@mail.gmail.com/ * strict-mm-typechecks-support: s390/mm: Add configurable STRICT_MM_TYPECHECKS s390/mm: Convert pgste_val() into function s390/mm: Convert pgprot_val() into function s390/mm: Use pgprot_val() instead of open coding Signed-off-by: Vasily Gorbik

s390: Remove ioremap_wt() and pgprot_writethrough()

2025-03-11T14:28:58Z

It turns out that while s390 architecture calls its memory-I/O mapping variants write-through and write-back the implementation of ioremap_wt() and pgprot_writethrough() does not match Linux notion of ioremap_wt(). In particular Linux expects ioremap_wt() to be weaker still than ioremap_wc(), allowing not just gathering and re-ordering but also reads to be served from cache. Instead s390's implementation is equivalent to normal ioremap() while its ioremap_wc() allows re-ordering. Note that there are no known users of ioremap_wt() on s390 and the resulting behavior is in line with asm-generic defining ioremap_wt() as ioremap(), if undefined, so no breakage is expected. As s390 does not have a mapping type matching the Linux notion of ioremap_wt() and pgprot_writethrough(), simply drop them and rely on the asm-generic fallbacks instead. Fixes: b02002cc4c0f ("s390/pci: Implement ioremap_wc/prot() with MIO") Fixes: b43b3fff042d ("s390: mm: convert to GENERIC_IOREMAP") Acked-by: Heiko Carstens Signed-off-by: Niklas Schnelle Signed-off-by: Vasily Gorbik

s390/mm: Convert pgste_val() into function

2025-03-11T14:27:34Z

Similar to all other *_val() functions convert the last remaining architecture specific mm primitive pgste_val() into a function. Add set_pgste_bit() and clear_pgste_bit() helper functions which allow to clear and set pgste bits. This is also similar to e.g. set_pte_bit() and other helper functions. Acked-by: Alexander Gordeev Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik

s390/cpufeature: Convert MACHINE_HAS_EDAT1 to cpu_has_edat1()

2025-03-04T16:18:05Z

Convert MACHINE_HAS_... to cpu_has_...() which uses test_facility() instead of testing the machine_flags lowcore member if the feature is present. test_facility() generates better code since it results in a static branch without accessing memory. The branch is patched via alternatives by the decompressor depending on the availability of the required facility. Reviewed-by: Vasily Gorbik Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik

s390/cpufeature: Convert MACHINE_HAS_RDP to cpu_has_rdp()

2025-03-04T16:18:04Z

KVM: s390: remove the last user of page->index

2025-01-31T11:03:53Z

Shadow page tables use page->index to keep the g2 address of the guest page table being shadowed. Instead of keeping the information in page->index, split the address and smear it over the 16-bit softbits areas of 4 PGSTEs. This removes the last s390 user of page->index. Reviewed-by: Steffen Eiden Reviewed-by: Christoph Schlameuss Link: https://lore.kernel.org/r/20250123144627.312456-16-imbrenda@linux.ibm.com Signed-off-by: Claudio Imbrenda Message-ID: <20250123144627.312456-16-imbrenda@linux.ibm.com>

KVM: s390: move PGSTE softbits

2025-01-31T11:03:53Z

Move the softbits in the PGSTEs to the other usable area. This leaves the 16-bit block of usable bits free, which will be used in the next patch for something else. Reviewed-by: Steffen Eiden Reviewed-by: Christoph Schlameuss Link: https://lore.kernel.org/r/20250123144627.312456-15-imbrenda@linux.ibm.com Signed-off-by: Claudio Imbrenda Message-ID: <20250123144627.312456-15-imbrenda@linux.ibm.com>

s390/mm: Simplify noexec page protection handling

2024-12-17T11:46:13Z

By default page protection definitions like PAGE_RX have the _PAGE_NOEXEC bit set. For older machines without the instruction execution protection facility this bit is not allowed to be used in page table entries, and therefore must be removed. This is done at a couple of page table walkers, but also at some but not all page table modification functions like ptep_modify_prot_commit(). Avoid all of this and change the page, segment and region3 protection definitions so that the noexec bit is masked out automatically if the instruction execution-protection facility is not available. This is similar to what also various other architectures do which had to solve the same problem. Reviewed-by: Gerald Schaefer Acked-by: Alexander Gordeev Signed-off-by: Heiko Carstens Signed-off-by: Alexander Gordeev

s390/mm: Remove unused PAGE_KERNEL_EXEC and friends

2024-12-17T11:46:13Z

Remove unused PAGE_KERNEL_EXEC, SEGMENT_KERNEL_EXEC, and REGION3_KERNEL_EXEC. Reviewed-by: Gerald Schaefer Acked-by: Alexander Gordeev Signed-off-by: Heiko Carstens Signed-off-by: Alexander Gordeev