* Peter Zijlstra <peterz@infradead.org> wrote:
> On Fri, Mar 28, 2025 at 02:26:26PM +0100, Ingo Molnar wrote:
> > So the temp_mm_state_t abstraction used by use_temporary_mm() and
> > unuse_temporary_mm() is super confusing:
>
> I thing I see what you're saying, but we also still have these patches
> pending:
>
> https://lkml.kernel.org/r/20241119162527.952745944@infradead.org
>
> :-(
Yeah, so I think we should do your queue on top of mine, the
whole temp_mm_state_t abstraction was rather nonsensical,
and we shouldn't be iterating confusing code...
I've ported patches #1 and #3 from your queue on top, see
attached below - these should be the two that represent 99%
of the conflicts between these two efforts AFAICS.
Does that work for you?
Thanks,
Ingo
================>
From: Peter Zijlstra <peterz@infradead.org>
Date: Tue, 19 Nov 2024 17:25:28 +0100
Subject: [PATCH] x86/mm: Add mm argument to unuse_temporary_mm()
In commit 209954cbc7d0 ("x86/mm/tlb: Update mm_cpumask lazily")
unuse_temporary_mm() grew the assumption that it gets used on
poking_nn exclusively. While this is currently true, lets not hard
code this assumption.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20241119163035.322525475@infradead.org
---
arch/x86/kernel/alternative.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c
index 5b1a6252a4b9..cfffcb80f564 100644
--- a/arch/x86/kernel/alternative.c
+++ b/arch/x86/kernel/alternative.c
@@ -2161,14 +2161,14 @@ static inline struct mm_struct *use_temporary_mm(struct mm_struct *temp_mm)
__ro_after_init struct mm_struct *text_poke_mm;
__ro_after_init unsigned long text_poke_mm_addr;
-static inline void unuse_temporary_mm(struct mm_struct *prev_mm)
+static inline void unuse_temporary_mm(struct mm_struct *mm, struct mm_struct *prev_mm)
{
lockdep_assert_irqs_disabled();
switch_mm_irqs_off(NULL, prev_mm, current);
/* Clear the cpumask, to indicate no TLB flushing is needed anywhere */
- cpumask_clear_cpu(raw_smp_processor_id(), mm_cpumask(text_poke_mm));
+ cpumask_clear_cpu(raw_smp_processor_id(), mm_cpumask(mm));
/*
* Restore the breakpoints if they were disabled before the temporary mm
@@ -2275,7 +2275,7 @@ static void *__text_poke(text_poke_f func, void *addr, const void *src, size_t l
* instruction that already allows the core to see the updated version.
* Xen-PV is assumed to serialize execution in a similar manner.
*/
- unuse_temporary_mm(prev_mm);
+ unuse_temporary_mm(text_poke_mm, prev_mm);
/*
* Flushing the TLB might involve IPIs, which would require enabled
======================================>
From: Andy Lutomirski <luto@kernel.org>
Date: Tue, 19 Nov 2024 17:25:30 +0100
Subject: [PATCH] x86/mm: Make use/unuse_temporary_mm() non-static
This prepares them for use outside of the alternative machinery.
The code is unchanged.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/d1205bc7e165e249c52b7fe8cb1254f06e8a0e2a.1641659630.git.luto@kernel.org
Link: https://lore.kernel.org/r/20241119163035.533822339@infradead.org
---
arch/x86/include/asm/mmu_context.h | 3 ++
arch/x86/kernel/alternative.c | 64 --------------------------------------
arch/x86/mm/tlb.c | 64 ++++++++++++++++++++++++++++++++++++++
3 files changed, 67 insertions(+), 64 deletions(-)
diff --git a/arch/x86/include/asm/mmu_context.h b/arch/x86/include/asm/mmu_context.h
index 2398058b6e83..b103e1709a67 100644
--- a/arch/x86/include/asm/mmu_context.h
+++ b/arch/x86/include/asm/mmu_context.h
@@ -272,4 +272,7 @@ unsigned long __get_current_cr3_fast(void);
#include <asm-generic/mmu_context.h>
+extern struct mm_struct *use_temporary_mm(struct mm_struct *temp_mm);
+extern void unuse_temporary_mm(struct mm_struct *mm, struct mm_struct *prev_mm);
+
#endif /* _ASM_X86_MMU_CONTEXT_H */
diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c
index cfffcb80f564..25abadaf8751 100644
--- a/arch/x86/kernel/alternative.c
+++ b/arch/x86/kernel/alternative.c
@@ -2111,73 +2111,9 @@ void __init_or_module text_poke_early(void *addr, const void *opcode,
}
}
-/*
- * Using a temporary mm allows to set temporary mappings that are not accessible
- * by other CPUs. Such mappings are needed to perform sensitive memory writes
- * that override the kernel memory protections (e.g., W^X), without exposing the
- * temporary page-table mappings that are required for these write operations to
- * other CPUs. Using a temporary mm also allows to avoid TLB shootdowns when the
- * mapping is torn down.
- *
- * Context: The temporary mm needs to be used exclusively by a single core. To
- * harden security IRQs must be disabled while the temporary mm is
- * loaded, thereby preventing interrupt handler bugs from overriding
- * the kernel memory protection.
- */
-static inline struct mm_struct *use_temporary_mm(struct mm_struct *temp_mm)
-{
- struct mm_struct *prev_mm;
-
- lockdep_assert_irqs_disabled();
-
- /*
- * Make sure not to be in TLB lazy mode, as otherwise we'll end up
- * with a stale address space WITHOUT being in lazy mode after
- * restoring the previous mm.
- */
- if (this_cpu_read(cpu_tlbstate_shared.is_lazy))
- leave_mm();
-
- prev_mm = this_cpu_read(cpu_tlbstate.loaded_mm);
- switch_mm_irqs_off(NULL, temp_mm, current);
-
- /*
- * If breakpoints are enabled, disable them while the temporary mm is
- * used. Userspace might set up watchpoints on addresses that are used
- * in the temporary mm, which would lead to wrong signals being sent or
- * crashes.
- *
- * Note that breakpoints are not disabled selectively, which also causes
- * kernel breakpoints (e.g., perf's) to be disabled. This might be
- * undesirable, but still seems reasonable as the code that runs in the
- * temporary mm should be short.
- */
- if (hw_breakpoint_active())
- hw_breakpoint_disable();
-
- return prev_mm;
-}
-
__ro_after_init struct mm_struct *text_poke_mm;
__ro_after_init unsigned long text_poke_mm_addr;
-static inline void unuse_temporary_mm(struct mm_struct *mm, struct mm_struct *prev_mm)
-{
- lockdep_assert_irqs_disabled();
-
- switch_mm_irqs_off(NULL, prev_mm, current);
-
- /* Clear the cpumask, to indicate no TLB flushing is needed anywhere */
- cpumask_clear_cpu(raw_smp_processor_id(), mm_cpumask(mm));
-
- /*
- * Restore the breakpoints if they were disabled before the temporary mm
- * was loaded.
- */
- if (hw_breakpoint_active())
- hw_breakpoint_restore();
-}
-
static void text_poke_memcpy(void *dst, const void *src, size_t len)
{
memcpy(dst, src, len);
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 0925768d00cb..06a1ad39be74 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -972,6 +972,70 @@ void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
this_cpu_write(cpu_tlbstate_shared.is_lazy, true);
}
+/*
+ * Using a temporary mm allows to set temporary mappings that are not accessible
+ * by other CPUs. Such mappings are needed to perform sensitive memory writes
+ * that override the kernel memory protections (e.g., W^X), without exposing the
+ * temporary page-table mappings that are required for these write operations to
+ * other CPUs. Using a temporary mm also allows to avoid TLB shootdowns when the
+ * mapping is torn down.
+ *
+ * Context: The temporary mm needs to be used exclusively by a single core. To
+ * harden security IRQs must be disabled while the temporary mm is
+ * loaded, thereby preventing interrupt handler bugs from overriding
+ * the kernel memory protection.
+ */
+struct mm_struct *use_temporary_mm(struct mm_struct *temp_mm)
+{
+ struct mm_struct *prev_mm;
+
+ lockdep_assert_irqs_disabled();
+
+ /*
+ * Make sure not to be in TLB lazy mode, as otherwise we'll end up
+ * with a stale address space WITHOUT being in lazy mode after
+ * restoring the previous mm.
+ */
+ if (this_cpu_read(cpu_tlbstate_shared.is_lazy))
+ leave_mm();
+
+ prev_mm = this_cpu_read(cpu_tlbstate.loaded_mm);
+ switch_mm_irqs_off(NULL, temp_mm, current);
+
+ /*
+ * If breakpoints are enabled, disable them while the temporary mm is
+ * used. Userspace might set up watchpoints on addresses that are used
+ * in the temporary mm, which would lead to wrong signals being sent or
+ * crashes.
+ *
+ * Note that breakpoints are not disabled selectively, which also causes
+ * kernel breakpoints (e.g., perf's) to be disabled. This might be
+ * undesirable, but still seems reasonable as the code that runs in the
+ * temporary mm should be short.
+ */
+ if (hw_breakpoint_active())
+ hw_breakpoint_disable();
+
+ return prev_mm;
+}
+
+void unuse_temporary_mm(struct mm_struct *mm, struct mm_struct *prev_mm)
+{
+ lockdep_assert_irqs_disabled();
+
+ switch_mm_irqs_off(NULL, prev_mm, current);
+
+ /* Clear the cpumask, to indicate no TLB flushing is needed anywhere */
+ cpumask_clear_cpu(raw_smp_processor_id(), mm_cpumask(mm));
+
+ /*
+ * Restore the breakpoints if they were disabled before the temporary mm
+ * was loaded.
+ */
+ if (hw_breakpoint_active())
+ hw_breakpoint_restore();
+}
+
/*
* Call this when reinitializing a CPU. It fixes the following potential
* problems: