Re: [PATCH] x86/asm: Merge KSTK_ESP() implementations

["H. Peter Anvin" <hpa@zytor.com>]

On March 3, 2025 4:25:06 PM PST, Brian Gerst <brgerst@gmail.com> wrote:
>On Mon, Mar 3, 2025 at 7:01 PM H. Peter Anvin <hpa@zytor.com> wrote:
>>
>> On March 3, 2025 10:31:11 AM PST, Brian Gerst <brgerst@gmail.com> wrote:
>> >Commit 263042e4630a ("Save user RSP in pt_regs->sp on SYSCALL64
>> >fastpath") simplified the 64-bit implementation of KSTK_ESP() which is
>> >now identical to 32-bit.  Merge them into a common definition.
>> >
>> >No functional change.
>> >
>> >Signed-off-by: Brian Gerst <brgerst@gmail.com>
>> >---
>> > arch/x86/include/asm/processor.h | 5 +----
>> > arch/x86/kernel/process_64.c     | 5 -----
>> > 2 files changed, 1 insertion(+), 9 deletions(-)
>> >
>> >diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
>> >index a969bea1ed07..55f0e48413b0 100644
>> >--- a/arch/x86/include/asm/processor.h
>> >+++ b/arch/x86/include/asm/processor.h
>> >@@ -652,8 +652,6 @@ static __always_inline void prefetchw(const void *x)
>> >       .sysenter_cs            = __KERNEL_CS,                            \
>> > }
>> >
>> >-#define KSTK_ESP(task)                (task_pt_regs(task)->sp)
>> >-
>> > #else
>> > extern unsigned long __top_init_kernel_stack[];
>> >
>> >@@ -661,8 +659,6 @@ extern unsigned long __top_init_kernel_stack[];
>> >       .sp     = (unsigned long)&__top_init_kernel_stack,              \
>> > }
>> >
>> >-extern unsigned long KSTK_ESP(struct task_struct *task);
>> >-
>> > #endif /* CONFIG_X86_64 */
>> >
>> > extern void start_thread(struct pt_regs *regs, unsigned long new_ip,
>> >@@ -676,6 +672,7 @@ extern void start_thread(struct pt_regs *regs, unsigned long new_ip,
>> > #define TASK_UNMAPPED_BASE            __TASK_UNMAPPED_BASE(TASK_SIZE_LOW)
>> >
>> > #define KSTK_EIP(task)                (task_pt_regs(task)->ip)
>> >+#define KSTK_ESP(task)                (task_pt_regs(task)->sp)
>> >
>> > /* Get/set a process' ability to use the timestamp counter instruction */
>> > #define GET_TSC_CTL(adr)      get_tsc_mode((adr))
>> >diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
>> >index 4ca73ddfb30b..f983d2a57ac3 100644
>> >--- a/arch/x86/kernel/process_64.c
>> >+++ b/arch/x86/kernel/process_64.c
>> >@@ -979,8 +979,3 @@ long do_arch_prctl_64(struct task_struct *task, int option, unsigned long arg2)
>> >
>> >       return ret;
>> > }
>> >-
>> >-unsigned long KSTK_ESP(struct task_struct *task)
>> >-{
>> >-      return task_pt_regs(task)->sp;
>> >-}
>> >
>> >base-commit: 693c8502970a533363e9ece482c80bb6db0c12a5
>>
>> Why using the macro version?
>
>Why call an out-of-line function?  I guess it could be an inline
>function (along with KSK_EIP()).
>
>
>Brian Gerst
>

Ah yes, not out of line obviously.