Review of KPTI patchset

[Mathieu Desnoyers]

Hi Thomas,

Here is some feedback on the KPTI patchset. Sorry for not replying to the
patch, I was not CC'd on the original email, and don't have it in my inbox.

I notice that fill_ldt() sets the desc->type with "|= 1", whereas all
other operations on the desc type are done with a type enum based on
clearly defined bits. Is the hardcoded "1" on purpose ?


arch/x86/include/asm/processor.h:

"+ * With page table isolation enabled, we map the LDT in ... [stay tuned]"

I look forward to publication of the next chapter containing the rest of
this sentence. When is it due ? ;)


arch/x86/include/asm/tlbflush.h:

@@ -135,6 +193,24 @@ struct tlb_state {

bool invalidate_other;

^ I thought using bool in structures was discouraged ? Arguably, there are other
booleans in that structure already.


+static void free_ldt_pgtables(struct mm_struct *mm)
+{
+#ifdef CONFIG_PAGE_TABLE_ISOLATION
+	struct mmu_gather tlb;
+	unsigned long start = LDT_BASE_ADDR;
+	unsigned long end = start + (1UL << PGDIR_SHIFT);
+
+	if (!static_cpu_has(X86_FEATURE_PTI))
+		return;
+
+	tlb_gather_mmu(&tlb, mm, start, end);
+	free_pgd_range(&tlb, start, end, start, end);
+	tlb_finish_mmu(&tlb, start, end);
+#endif

^ AFAIK, the usual approach is to move the #ifdef outside of the function body,
and have one empty function.


@@ -156,6 +271,12 @@ int ldt_dup_context(struct mm_struct *old_mm, struct mm_struct *mm)
 	       new_ldt->nr_entries * LDT_ENTRY_SIZE);
 	finalize_ldt_struct(new_ldt);
 
+	retval = map_ldt_struct(mm, new_ldt, 0);
+	if (retval) {
+		free_ldt_pgtables(mm);
+		free_ldt_struct(new_ldt);
+		goto out_unlock;
+	}
 	mm->context.ldt = new_ldt;
 
 out_unlock:

^ I don't get why it does "free_ldt_pgtables(mm)" on the mm argument, but
it's not done in other error paths. Perhaps it's OK, but ownership seems
non-obvious.


@@ -287,6 +413,18 @@ static int write_ldt(void __user *ptr, unsigned long bytecount, int oldmode)
 	new_ldt->entries[ldt_info.entry_number] = ldt;
 	finalize_ldt_struct(new_ldt);
 
+	/*
+	 * If we are using PTI, map the new LDT into the userspace pagetables.
+	 * If there is already an LDT, use the other slot so that other CPUs
+	 * will continue to use the old LDT until install_ldt() switches
+	 * them over to the new LDT.
+	 */
+	error = map_ldt_struct(mm, new_ldt, old_ldt ? !old_ldt->slot : 0);
+	if (error) {
+		free_ldt_struct(old_ldt);
+		goto out_unlock;
+	}
+

^ is it really "old_ldt" that we want freed on error here ? Or should it be
"new_ldt" ?


--- a/arch/x86/mm/cpu_entry_area.c
+++ b/arch/x86/mm/cpu_entry_area.c
@@ -38,6 +38,32 @@ cea_map_percpu_pages(void *cea_vaddr, void *ptr, int pages, pgprot_t prot)
 		cea_set_pte(cea_vaddr, per_cpu_ptr_to_phys(ptr), prot);
 }
 
+static void percpu_setup_debug_store(int cpu)
+{
+#ifdef CONFIG_CPU_SUP_INTEL

^ This ifdef should ideally be outside of the function, defining
an empty function in the #else.


+	/*
+	 * Force the population of PMDs for not yet allocated per cpu
+	 * memory like debug store buffers.
+	 */
+	npages = sizeof(struct debug_store_buffers) / PAGE_SIZE;
+	for (; npages; npages--, cea += PAGE_SIZE)
+		cea_set_pte(cea, 0, PAGE_NONE);

^ the code above (in percpu_setup_debug_store()) depends on having
struct debug_store_buffers's size being a multiple of PAGE_SIZE. A
comment should be added near the structure declaration to document
this requirement.


+static void __init pti_setup_espfix64(void)
+{
+#ifdef CONFIG_X86_ESPFIX64
+	pti_clone_p4d(ESPFIX_BASE_ADDR);
+#endif
+}

Seeing how this ifdef within function layout is everywhere in the patch,
I start to wonder whether I missed a coding style guideline somewhere... ?


+/*
+ * We get here when we do something requiring a TLB invalidation
+ * but could not go invalidate all of the contexts.  We do the
+ * necessary invalidation by clearing out the 'ctx_id' which
+ * forces a TLB flush when the context is loaded.
+ */
+void clear_asid_other(void)
+{
+	u16 asid;
+
+	/*
+	 * This is only expected to be set if we have disabled
+	 * kernel _PAGE_GLOBAL pages.
+	 */
+	if (!static_cpu_has(X86_FEATURE_PTI)) {
+		WARN_ON_ONCE(1);
+		return;
+	}
+
+	for (asid = 0; asid < TLB_NR_DYN_ASIDS; asid++) {
+		/* Do not need to flush the current asid */
+		if (asid == this_cpu_read(cpu_tlbstate.loaded_mm_asid))
+			continue;
+		/*
+		 * Make sure the next time we go to switch to
+		 * this asid, we do a flush:
+		 */
+		this_cpu_write(cpu_tlbstate.ctxs[asid].ctx_id, 0);
+	}
+	this_cpu_write(cpu_tlbstate.invalidate_other, false);
+}

Can this be called with preemption enabled ? If so, what happens
if migrated ?

Thanks,

Mathieu























-- 
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com