kernel/arch/s390/boot/boot.h, branch linux-rolling-stable

s390: Add stackprotector support

2025-11-24T10:45:21Z

Stackprotector support was previously unavailable on s390 because by default compilers generate code which is not suitable for the kernel: the canary value is accessed via thread local storage, where the address of thread local storage is within access registers 0 and 1. Using those registers also for the kernel would come with a significant performance impact and more complicated kernel entry/exit code, since access registers contents would have to be exchanged on every kernel entry and exit. With the upcoming gcc 16 release new compiler options will become available which allow to generate code suitable for the kernel. [1] Compiler option -mstack-protector-guard=global instructs gcc to generate stackprotector code that refers to a global stackprotector canary value via symbol __stack_chk_guard. Access to this value is guaranteed to occur via larl and lgrl instructions. Furthermore, compiler option -mstack-protector-guard-record generates a section containing all code addresses that reference the canary value. To allow for per task canary values the instructions which load the address of __stack_chk_guard are patched so they access a lowcore field instead: a per task canary value is available within the task_struct of each task, and is written to the per-cpu lowcore location on each context switch. Also add sanity checks and debugging option to be consistent with other kernel code patching mechanisms. Full debugging output can be enabled with the following kernel command line options: debug_stackprotector bootdebug ignore_loglevel earlyprintk dyndbg="file stackprotector.c +p" Example debug output: stackprot: 0000021e402d4eda: c010005a9ae3 -> c01f00070240 where ": -> ". [1] gcc commit 0cd1f03939d5 ("s390: Support global stack protector") Reviewed-by: Sven Schnelle Signed-off-by: Heiko Carstens

s390/boot: Add common boot_panic() code

2025-08-28T14:52:54Z

Introduce a common boot_panic() helper macro, and use it to get rid of three more or less identical implementations. Signed-off-by: Heiko Carstens Reviewed-by: Alexander Gordeev Signed-off-by: Alexander Gordeev

s390/boot: Introduce jump_to_kernel() function

2025-07-10T12:42:10Z

Introduce a global function that jumps from the decompressor to the decompressed kernel. Put its address into svc_old_psw, from where GDB can take it without loading decompressor symbols. It should be available throughout the entire decompressor execution, because it's placed there statically, and nothing in the decompressor uses the SVC instruction. Acked-by: Heiko Carstens Signed-off-by: Ilya Leoshkevich Tested-by: Alexander Gordeev Link: https://lore.kernel.org/r/20250625154220.75300-2-iii@linux.ibm.com Signed-off-by: Alexander Gordeev

s390: Replace ASSEMBLY with ASSEMBLER in non-uapi headers

2025-06-16T14:23:02Z

While the GCC and Clang compilers already define __ASSEMBLER__ automatically when compiling assembler code, __ASSEMBLY__ is a macro that only gets defined by the Makefiles in the kernel. This is bad since macros starting with two underscores are names that are reserved by the C language. It can also be very confusing for the developers when switching between userspace and kernelspace coding, or when dealing with uapi headers that rather should use __ASSEMBLER__ instead. So let's now standardize on the __ASSEMBLER__ macro that is provided by the compilers. This is a completely mechanical patch (done with a simple "sed -i" statement), with some manual fixups done later while rebasing the patch. Signed-off-by: Thomas Huth Acked-by: Heiko Carstens Link: https://lore.kernel.org/r/20250611140046.137739-3-thuth@redhat.com Signed-off-by: Alexander Gordeev

s390/boot: Add exception table support

2025-03-04T16:25:22Z

The early boot code contains various open-coded inline assemblies with exception handling. In order to handle possible exceptions each of them changes the program check new psw, and restores it. In order to simplify the various inline assemblies add simple exception table support: the program check handler is called with a fully populated pt_regs on the stack and may change the psw and register members. When the program check handler returns the psw and registers from pt_regs will be used to continue execution. The program check handler searches the exception table for an entry which matches the address of the program check. If such an entry is found the psw address within pt_regs on the stack is replaced with a fixup address, and execution continues at the new address. If no entry is found the psw is changed to a disabled wait psw and execution stops. Before entering the C part of the program check handler the address of the program check new psw is replaced to a minimalistic handler. This is supposed to help against program check loops. If an exception happens while in program check processing the register contents of the original exception are restored and a disabled wait psw is loaded. Acked-by: Alexander Gordeev Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik

s390/boot: Pass pt_regs to program check handler

2025-03-04T16:25:22Z

Setup a pt_regs structure on the stack, poplulate it in low level assembler code, and pass it to print_pgm_check_info(). This way there is no need to access then lowcore from print_pgm_check_info() anymore, and the function looks like a normal program check handler function. Acked-by: Alexander Gordeev Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik

s390/alternatives: Add debug functionality

2025-03-04T16:18:08Z

Similar to x86 and loongarch add a "debug-alternative" command line parameter, which allows for alternative debugging. The parameter itself comes with architecture specific semantics: "debug-alternative" -> print debug message for every single alternative "debug-alternative=0;2" -> print debug message for all alternatives with type 0 and 2 "debug-alternative=0:0-7" -> print debug message for all alternatives with type 0 which have a facility number within the range of 0-7 "debug-alternative=0:!8;1" -> print debug message for all alternatives with type 0, for all facility numbers, except facility 8, and in addition print all alternatives with type 1 A defconfig build currently results in a kernel with more than 20.000 alternatives, where the majority is for the niai alternative (spinlocks), and the relocated lowcore alternative. The following kernel command like options limit alternative debug output, and enable dynamic debug messages: debug-alternative=0:!49;1:!0 earlyprintk bootdebug ignore_loglevel loglevel=8 dyndbg="file alternative.c +p" This results in output like this: alt: [0/ 11] 0000021b9ce8680c: c0f400000089 -> c00400000000 alt: [0/ 64] 0000021b9ce87e60: c0f400000043 -> c00400000000 alt: [0/133] 0000021b9ce88c56: c0f400000027 -> c00400000000 alt: [0/ 74] 0000021b9ce89410: c0f40000002a -> c00400000000 alt: [0/ 40] 0000021b9dc3720a: 47000000 -> b280d398 alt: [0/193] 0000021b9dc37306: 47000000 -> b201d2b0 alt: [0/193] 0000021b9dc37354: c00400000000 -> d20720c0d2b0 alt: [1/ 5] 0000038d720d7bf2: c0f400000016 -> c00400000000 With [/]

oldcode -> newcode Alternative data depends on the alternative type: for type 0 (ALT_TYPE_FACILITY) data is the facility. For type 1 (ALT_TYPE_FEATURE) data is the corresponding machine feature. Acked-by: Vasily Gorbik Signed-off-by: Heiko Carstens Signed-off-by: Vasily Gorbik

s390/cpufeature: Convert MACHINE_HAS_EDAT2 to cpu_has_edat2()

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_EDAT1 to cpu_has_edat1()