[PATCH v2] kmemleak: skip scanning holes in the .bss section

[Catalin Marinas]

On Thu, Mar 21, 2019 at 12:15:46AM +1100, Michael Ellerman wrote:
> Catalin Marinas <catalin.marinas at arm.com> writes:
> > On Wed, Mar 13, 2019 at 10:57:17AM -0400, Qian Cai wrote:
> >> @@ -1531,7 +1547,14 @@ static void kmemleak_scan(void)
> >>  
> >>  	/* data/bss scanning */
> >>  	scan_large_block(_sdata, _edata);
> >> -	scan_large_block(__bss_start, __bss_stop);
> >> +
> >> +	if (bss_hole_start) {
> >> +		scan_large_block(__bss_start, bss_hole_start);
> >> +		scan_large_block(bss_hole_stop, __bss_stop);
> >> +	} else {
> >> +		scan_large_block(__bss_start, __bss_stop);
> >> +	}
> >> +
> >>  	scan_large_block(__start_ro_after_init, __end_ro_after_init);
> >
> > I'm not a fan of this approach but I couldn't come up with anything
> > better. I was hoping we could check for PageReserved() in scan_block()
> > but on arm64 it ends up not scanning the .bss at all.
> >
> > Until another user appears, I'm ok with this patch.
> >
> > Acked-by: Catalin Marinas <catalin.marinas at arm.com>
> 
> I actually would like to rework this kvm_tmp thing to not be in bss at
> all. It's a bit of a hack and is incompatible with strict RWX.
> 
> If we size it a bit more conservatively we can hopefully just reserve
> some space in the text section for it.
> 
> I'm not going to have time to work on that immediately though, so if
> people want this fixed now then this patch could go in as a temporary
> solution.

I think I have a simpler idea. Kmemleak allows punching holes in
allocated objects, so just turn the data/bss sections into dedicated
kmemleak objects. This happens when kmemleak is initialised, before the
initcalls are invoked. The kvm_free_tmp() would just free the
corresponding part of the bss.

Patch below, only tested briefly on arm64. Qian, could you give it a try
on powerpc? Thanks.

--------8<------------------------------
diff --git a/arch/powerpc/kernel/kvm.c b/arch/powerpc/kernel/kvm.c
index 683b5b3805bd..c4b8cb3c298d 100644
--- a/arch/powerpc/kernel/kvm.c
+++ b/arch/powerpc/kernel/kvm.c
@@ -712,6 +712,8 @@ static void kvm_use_magic_page(void)
 
 static __init void kvm_free_tmp(void)
 {
+	kmemleak_free_part(&kvm_tmp[kvm_tmp_index],
+			   ARRAY_SIZE(kvm_tmp) - kvm_tmp_index);
 	free_reserved_area(&kvm_tmp[kvm_tmp_index],
 			   &kvm_tmp[ARRAY_SIZE(kvm_tmp)], -1, NULL);
 }
diff --git a/mm/kmemleak.c b/mm/kmemleak.c
index 707fa5579f66..0f6adcbfc2c7 100644
--- a/mm/kmemleak.c
+++ b/mm/kmemleak.c
@@ -1529,11 +1529,6 @@ static void kmemleak_scan(void)
 	}
 	rcu_read_unlock();
 
-	/* data/bss scanning */
-	scan_large_block(_sdata, _edata);
-	scan_large_block(__bss_start, __bss_stop);
-	scan_large_block(__start_ro_after_init, __end_ro_after_init);
-
 #ifdef CONFIG_SMP
 	/* per-cpu sections scanning */
 	for_each_possible_cpu(i)
@@ -2071,6 +2066,15 @@ void __init kmemleak_init(void)
 	}
 	local_irq_restore(flags);
 
+	/* register the data/bss sections */
+	create_object((unsigned long)_sdata, _edata - _sdata,
+		      KMEMLEAK_GREY, GFP_ATOMIC);
+	create_object((unsigned long)__bss_start, __bss_stop - __bss_start,
+		      KMEMLEAK_GREY, GFP_ATOMIC);
+	create_object((unsigned long)__start_ro_after_init,
+		      __end_ro_after_init - __start_ro_after_init,
+		      KMEMLEAK_GREY, GFP_ATOMIC);
+
 	/*
 	 * This is the point where tracking allocations is safe. Automatic
 	 * scanning is started during the late initcall. Add the early logged