Tags give the ability to mark specific points in history as being important
-
-
-
-
-
-
-
-
-
-
x86-boot-2024-03-12
2e2bc42c · ·x86/boot changes for v6.9: - Continuing work by Ard Biesheuvel to improve the x86 early startup code, with the long-term goal to make it position independent: - Get rid of early accesses to global objects, either by moving them to the stack, deferring the access until later, or dropping the globals entirely. - Move all code that runs early via the 1:1 mapping into .head.text, and move code that does not out of it, so that build time checks can be added later to ensure that no inadvertent absolute references were emitted into code that does not tolerate them. - Remove fixup_pointer() and occurrences of __pa_symbol(), which rely on the compiler emitting absolute references, which is not guaranteed. - Improve the early console code. - Add early console message about ignored NMIs, so that users are at least warned about their existence - even if we cannot do anything about them. - Improve the kexec code's kernel load address handling. - Enable more X86S (simplified x86) bits. - Simplify early boot GDT handling - Micro-optimize the boot code a bit - Misc cleanups. Signed-off-by: Ingo Molnar <mingo@kernel.org> -
x86-entry-2024-03-11
bb998361 · ·A single update for the x86 entry code: The current CR3 handling for kernel page table isolation in the paranoid return paths which are relevant for #NMI, #MCE, #VC, #DB and #DF is unconditionally writing CR3 with the value retrieved on exception entry. In the vast majority of cases when returning to the kernel this is a pointless exercise because CR3 was not modified on exception entry. The only situation where this is necessary is when the exception interrupts a entry from user before switching to kernel CR3 or interrupts an exit to user after switching back to user CR3. As CR3 writes can be expensive on some systems this becomes measurable overhead with high frequency #NMIs such as perf. Avoid this overhead by checking the CR3 value, which was saved on entry, and write it back to CR3 only when it us a user CR3.
-
-
x86_mm_for_6.9
82ace185 · ·* Add a warning when memory encryption conversions fail. These operations require VMM cooperation, even in CoCo environments where the VMM is untrusted. While it's _possible_ that memory pressure could trigger the new warning, the odds are that a guest would only see this from an attacking VMM. * Simplify page fault code by re-enabling interrupts unconditionally * Avoid truncation issues when pfns are passed in to pfn_to_kaddr() with small (<64-bit) types.
-
x86_misc_for_v6.9_rc1
d54e56f3 · ·- Fix a wrong check in the function reporting whether a CPU executes (or not) a NMI handler - Ratelimit unknown NMIs messages in order to not potentially slow down the machine - Other fixlets
-
x86_sev_for_v6.9_rc1
c0935fca · ·- Add the x86 part of the SEV-SNP host support. This will allow the kernel to be used as a KVM hypervisor capable of running SNP (Secure Nested Paging) guests. Roughly speaking, SEV-SNP is the ultimate goal of the AMD confidential computing side, providing the most comprehensive confidential computing environment up to date. This is the x86 part and there is a KVM part which did not get ready in time for the merge window so latter will be forthcoming in the next cycle. - Rework the early code's position-dependent SEV variable references in order to allow building the kernel with clang and -fPIE/-fPIC and -mcmodel=kernel - The usual set of fixes, cleanups and improvements all over the place
-
x86_cache_for_v6.9_rc1
c0d848fc · ·- Rework different aspects of the resctrl code like adding arch-specific accessors and splitting the locking, in order to accomodate ARM's MPAM implementation of hw resource control and be able to use the same filesystem control interface like on x86. Work by James Morse - Improve the memory bandwidth throttling heuristic to handle workloads with not too regular load levels which end up penalized unnecessarily - Use CPUID to detect the memory bandwidth enforcement limit on AMD - The usual set of fixes
-
x86-core-2024-03-11
35ce6492 · ·Core x86 changes for v6.9: - The biggest change is the rework of the percpu code, to support the 'Named Address Spaces' GCC feature, by Uros Bizjak: - This allows C code to access GS and FS segment relative memory via variables declared with such attributes, which allows the compiler to better optimize those accesses than the previous inline assembly code. - The series also includes a number of micro-optimizations for various percpu access methods, plus a number of cleanups of %gs accesses in assembly code. - These changes have been exposed to linux-next testing for the last ~5 months, with no known regressions in this area. - Fix/clean up __switch_to()'s broken but accidentally working handling of FPU switching - which also generates better code. - Propagate more RIP-relative addressing in assembly code, to generate slightly better code. - Rework the CPU mitigations Kconfig space to be less idiosyncratic, to make it easier for distros to follow & maintain these options. - Rework the x86 idle code to cure RCU violations and to clean up the logic. - Clean up the vDSO Makefile logic. - Misc cleanups and fixes. [ Please note that there's a higher number of merge commits in this branch (three) than is usual in x86 topic trees. This happened due to the long testing lifecycle of the percpu changes that involved 3 merge windows, which generated a longer history and various interactions with other core x86 changes that we felt better about to carry in a single branch. ] Signed-off-by: Ingo Molnar <mingo@kernel.org> -
x86-cleanups-2024-03-11
774a86f1 · ·Misc cleanups, including a large series from Thomas Gleixner to cure Sparse warnings. Signed-off-by: Ingo Molnar <mingo@kernel.org>
-
x86-build-2024-03-11
103bf75f · ·x86/build changes for v6.9: - Reduce <asm/bootparam.h> dependencies - Simplify <asm/efi.h> - Unify *_setup_data definitions into <asm/setup_data.h> - Reduce the size of <asm/bootparam.h> Signed-off-by: Ingo Molnar <mingo@kernel.org>
-
x86-asm-2024-03-11
07a5d4bc · ·Two changes to simplify the x86 decoder logic a bit. Signed-off-by: Ingo Molnar <mingo@kernel.org>