kernel.git/arch/xtensa/kernel/entry.S, branch linux-5.18.y

xtensa: move trace_hardirqs_off call back to entry.S

2022-06-09T08:29:39Z

[ Upstream commit de4415d0bac91192ee9c74e849bc61429efa9b42 ] Context tracking call must be done after hardirq tracking call, otherwise lockdep_assert_irqs_disabled called from rcu_eqs_exit gives a warning. To avoid context tracking logic duplication for IRQ/exception entry paths move trace_hardirqs_off call back to common entry code. Signed-off-by: Max Filippov Signed-off-by: Sasha Levin

xtensa: clean up kernel exit assembly code

2022-03-19T20:08:41Z

Don't use numeric labels for complex branching logic. Mark each branch with named local label and use them. Rearrange exit back to kernel mode to avoid conditional label definition. Signed-off-by: Max Filippov

xtensa: rearrange NMI exit path

2022-03-19T20:08:32Z

NMI exit path to userspace should neither check TIF_DB_DISABLED nor call check_tlb_sanity because NMI shouldn't touch anything related to userspace. Drop kernel/userspace check in NMI exit path. Signed-off-by: Max Filippov

xtensa: rename PT_SIZE to PT_KERNEL_SIZE

2022-03-07T20:02:48Z

PT_SIZE is used by the xtensa port to designate kernel exception frame size. In preparation for struct pt_regs size change rename PT_SIZE to PT_KERNEL_SIZE for clarity and change its definition to always cover only the kernel exception frame. Signed-off-by: Max Filippov Reviewed-by: Kees Cook

exit/xtensa: In arch/xtensa/entry.S:Linvalid_mask call make_task_dead

2022-01-08T16:46:31Z

There have historically been two big uses of do_exit. The first is it's design use to be the guts of the exit(2) system call. The second use is to terminate a task after something catastrophic has happened like a NULL pointer in kernel code. The function make_task_dead has been added to accomidate the second use. The call to do_exit in Linvalidmask is clearly not a normal userspace exit. As failure handling there are two possible ways to go. If userspace can trigger the issue force_exit_sig should be called. Otherwise make_task_dead probably from the implementation of die is appropriate. Replace the call of do_exit in Linvalidmask with make_task_dead as I don't know xtensa and especially xtensa assembly language well enough to do anything else. Link: https://lkml.kernel.org/r/YdUmN7n4W5YETUhW@zeniv-ca.linux.org.uk Suggested-by: Al Viro Signed-off-by: "Eric W. Biederman"

xtensa: only build windowed register support code when needed

2021-10-19T05:19:35Z

There's no need in window overflow/underflow/alloca exception handlers or window spill code when neither kernel nor userspace support windowed registers. Don't build or link it. Signed-off-by: Max Filippov

xtensa: use register window specific opcodes only when present

2021-10-19T05:19:35Z

xtensa core may be configured without register windows support, don't use register window specific opcodes in that case. Use window register specific opcodes to initialize hardware or reset core to a known state regardless of the chosen ABI. Signed-off-by: Max Filippov

xtensa: implement call0 ABI support in assembly

2021-10-19T05:19:35Z

Replace hardcoded register and opcode names with ABI-agnostic macros. Add register save/restore code where necessary. Conditionalize windowed only or call0 only code. Add stack initialization matching _switch_to epilogue to copy_thread. Signed-off-by: Max Filippov

xtensa: add support for TIF_NOTIFY_SIGNAL

2020-11-12T15:45:56Z

Wire up TIF_NOTIFY_SIGNAL handling for xtensa. Thanks to Max Filippov for making the asm correct. Cc: linux-xtensa@linux-xtensa.org Signed-off-by: Jens Axboe

xtensa: add missing exclusive access state management

2020-07-31T21:15:57Z

The result of the s32ex opcode is recorded in the ATOMCTL special register and must be retrieved with the getex opcode. Context switch between s32ex and getex may trash the ATOMCTL register and result in duplicate update or missing update of the atomic variable. Add atomctl8 field to the struct thread_info and use getex to swap ATOMCTL bit 8 as a part of context switch. Clear exclusive access monitor on kernel entry. Cc: stable@vger.kernel.org Fixes: f7c34874f04a ("xtensa: add exclusive atomics support") Signed-off-by: Max Filippov