| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <aBzt3HIU3l0sp81_@pc636>
Date: Thu, 8 May 2025 19:46:04 +0200
From: Uladzislau Rezki <urezki@...il.com>
To: Jeongjun Park <aha310510@...il.com>
Cc: Uladzislau Rezki <urezki@...il.com>, akpm@...ux-foundation.org,
edumazet@...gle.com, linux-mm@...ck.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v5] mm/vmalloc: fix data race in show_numa_info()
On Fri, May 09, 2025 at 01:26:14AM +0900, Jeongjun Park wrote:
> Uladzislau Rezki <urezki@...il.com> wrote:
> >
> > On Fri, May 09, 2025 at 01:07:59AM +0900, Jeongjun Park wrote:
> > > The following data-race was found in show_numa_info():
> > >
> > > ==================================================================
> > > BUG: KCSAN: data-race in vmalloc_info_show / vmalloc_info_show
> > >
> > > read to 0xffff88800971fe30 of 4 bytes by task 8289 on cpu 0:
> > > show_numa_info mm/vmalloc.c:4936 [inline]
> > > vmalloc_info_show+0x5a8/0x7e0 mm/vmalloc.c:5016
> > > seq_read_iter+0x373/0xb40 fs/seq_file.c:230
> > > proc_reg_read_iter+0x11e/0x170 fs/proc/inode.c:299
> > > ....
> > >
> > > write to 0xffff88800971fe30 of 4 bytes by task 8287 on cpu 1:
> > > show_numa_info mm/vmalloc.c:4934 [inline]
> > > vmalloc_info_show+0x38f/0x7e0 mm/vmalloc.c:5016
> > > seq_read_iter+0x373/0xb40 fs/seq_file.c:230
> > > proc_reg_read_iter+0x11e/0x170 fs/proc/inode.c:299
> > > ....
> > >
> > > value changed: 0x0000008f -> 0x00000000
> > > ==================================================================
> > >
> > > According to this report,there is a read/write data-race because m->private
> > > is accessible to multiple CPUs. To fix this, instead of allocating the heap
> > > in proc_vmalloc_init() and passing the heap address to m->private,
> > > vmalloc_info_show() should allocate the heap.
> > >
> > > Fixes: a47a126ad5ea ("vmallocinfo: add NUMA information")
> > > Suggested-by: Eric Dumazet <edumazet@...gle.com>
> > > Suggested-by: Uladzislau Rezki (Sony) <urezki@...il.com>
> > > Suggested-by: Andrew Morton <akpm@...ux-foundation.org>
> > > Signed-off-by: Jeongjun Park <aha310510@...il.com>
> > > ---
> > > v5: Change heap to be allocated only when CONFIG_NUMA is enabled
> > > - Link to v4: https://lore.kernel.org/all/20250508065558.149091-1-aha310510@gmail.com/
> > > v4: Change the way counters array heap is allocated, per Andrew Morton's suggestion.
> > > And fix it to call smp_rmb() in the correct location.
> > > - Link to v3: https://lore.kernel.org/all/20250507142552.9446-1-aha310510@gmail.com/
> > > v3: Following Uladzislau Rezki's suggestion, we check v->flags beforehand
> > > to avoid printing uninitialized members of vm_struct.
> > > - Link to v2: https://lore.kernel.org/all/20250506082520.84153-1-aha310510@gmail.com/
> > > v2: Refactoring some functions and fix patch as per Eric Dumazet suggestion
> > > - Link to v1: https://lore.kernel.org/all/20250505171948.24410-1-aha310510@gmail.com/
> > > ---
> > > mm/vmalloc.c | 62 ++++++++++++++++++++++++++++------------------------
> > > 1 file changed, 34 insertions(+), 28 deletions(-)
> > >
> > > diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> > > index 3ed720a787ec..866f18766dfc 100644
> > > --- a/mm/vmalloc.c
> > > +++ b/mm/vmalloc.c
> > > @@ -3100,7 +3100,7 @@ static void clear_vm_uninitialized_flag(struct vm_struct *vm)
> > > /*
> > > * Before removing VM_UNINITIALIZED,
> > > * we should make sure that vm has proper values.
> > > - * Pair with smp_rmb() in show_numa_info().
> > > + * Pair with smp_rmb() in vread_iter() and vmalloc_info_show().
> > > */
> > > smp_wmb();
> > > vm->flags &= ~VM_UNINITIALIZED;
> > > @@ -4914,28 +4914,29 @@ bool vmalloc_dump_obj(void *object)
> > > #endif
> > >
> > > #ifdef CONFIG_PROC_FS
> > > -static void show_numa_info(struct seq_file *m, struct vm_struct *v)
> > > -{
> > > - if (IS_ENABLED(CONFIG_NUMA)) {
> > > - unsigned int nr, *counters = m->private;
> > > - unsigned int step = 1U << vm_area_page_order(v);
> > >
> > > - if (!counters)
> > > - return;
> > > +/*
> > > + * Print number of pages allocated on each memory node.
> > > + *
> > > + * This function can only be called if CONFIG_NUMA is enabled
> > > + * and VM_UNINITIALIZED bit in v->flags is disabled.
> > > + */
> > > +static void show_numa_info(struct seq_file *m, struct vm_struct *v,
> > > + unsigned int *counters)
> > > +{
> > > + unsigned int nr;
> > > + unsigned int step = 1U << vm_area_page_order(v);
> > >
> > > - if (v->flags & VM_UNINITIALIZED)
> > > - return;
> > > - /* Pair with smp_wmb() in clear_vm_uninitialized_flag() */
> > > - smp_rmb();
> > > + if (!counters)
> > > + return;
> > >
> > > - memset(counters, 0, nr_node_ids * sizeof(unsigned int));
> > > + memset(counters, 0, nr_node_ids * sizeof(unsigned int));
> > >
> > > - for (nr = 0; nr < v->nr_pages; nr += step)
> > > - counters[page_to_nid(v->pages[nr])] += step;
> > > - for_each_node_state(nr, N_HIGH_MEMORY)
> > > - if (counters[nr])
> > > - seq_printf(m, " N%u=%u", nr, counters[nr]);
> > > - }
> > > + for (nr = 0; nr < v->nr_pages; nr += step)
> > > + counters[page_to_nid(v->pages[nr])] += step;
> > > + for_each_node_state(nr, N_HIGH_MEMORY)
> > > + if (counters[nr])
> > > + seq_printf(m, " N%u=%u", nr, counters[nr]);
> > > }
> > >
> > > static void show_purge_info(struct seq_file *m)
> > > @@ -4962,8 +4963,12 @@ static int vmalloc_info_show(struct seq_file *m, void *p)
> > > struct vmap_node *vn;
> > > struct vmap_area *va;
> > > struct vm_struct *v;
> > > + unsigned int *counters = NULL;
> > > int i;
> > >
> > > + if (IS_ENABLED(CONFIG_NUMA))
> > > + counters = kmalloc(nr_node_ids * sizeof(unsigned int), GFP_KERNEL);
> > > +
> > > for (i = 0; i < nr_vmap_nodes; i++) {
> > > vn = &vmap_nodes[i];
> > >
> > > @@ -4979,6 +4984,11 @@ static int vmalloc_info_show(struct seq_file *m, void *p)
> > > }
> > >
> > > v = va->vm;
> > > + if (v->flags & VM_UNINITIALIZED)
> > > + continue;
> > > +
> > > + /* Pair with smp_wmb() in clear_vm_uninitialized_flag() */
> > > + smp_rmb();
> > >
> > > seq_printf(m, "0x%pK-0x%pK %7ld",
> > > v->addr, v->addr + v->size, v->size);
> > > @@ -5013,7 +5023,9 @@ static int vmalloc_info_show(struct seq_file *m, void *p)
> > > if (is_vmalloc_addr(v->pages))
> > > seq_puts(m, " vpages");
> > >
> > > - show_numa_info(m, v);
> > > + if (counters)
> > > + show_numa_info(m, v, counters);
> > > +
> > Let's execute it for NUMA only.
> >
> > > seq_putc(m, '\n');
> > > }
> > > spin_unlock(&vn->busy.lock);
> > > @@ -5023,19 +5035,13 @@ static int vmalloc_info_show(struct seq_file *m, void *p)
> > > * As a final step, dump "unpurged" areas.
> > > */
> > > show_purge_info(m);
> > > + kfree(counters);
> > Let's execute it for NUMA only.
> >
> > > return 0;
> > > }
> > >
> > > static int __init proc_vmalloc_init(void)
> > > {
> > > - void *priv_data = NULL;
> > > -
> > > - if (IS_ENABLED(CONFIG_NUMA))
> > > - priv_data = kmalloc(nr_node_ids * sizeof(unsigned int), GFP_KERNEL);
> > > -
> > > - proc_create_single_data("vmallocinfo",
> > > - 0400, NULL, vmalloc_info_show, priv_data);
> > > -
> > > + proc_create_single("vmallocinfo", 0400, NULL, vmalloc_info_show);
> > > return 0;
> > > }
> > > module_init(proc_vmalloc_init);
> > > --
> >
> > You are so fast :)
> >
>
> oops sorry.
>
> I'm not sure why you'd want to avoid initializing "counters pointer" to
> NULL, but it's a simple change so I'll write a patch for it right away
> and send it out as a v6 patch.
>
Because i would like to avoid of having such rule if someone decides to
drop your initialization later of after a code refactoring. This is a reason.
--
Uladzislau Rezki
Powered by blists - more mailing lists