| Message ID | 20190128094301.151252-1-drinkcat@chromium.org (mailing list archive) |
|---|---|
| State | Mainlined, archived |
| Commit | 032ebd8548c9d05e8d2bdc7a7ec2fe29454b0ad0 |
| Headers | show |
| Series | iommu/io-pgtable-arm-v7s: only kmemleak_ignore L2 tables | expand |
On Mon, Jan 28, 2019 at 05:43:01PM +0800, Nicolas Boichat wrote: > L1 tables are allocated with __get_dma_pages, and therefore already > ignored by kmemleak. > > Without this, the kernel would print this error message on boot, > when the first L1 table is allocated: > > [ 2.810533] kmemleak: Trying to color unknown object at 0xffffffd652388000 as Black > [ 2.818190] CPU: 5 PID: 39 Comm: kworker/5:0 Tainted: G S 4.19.16 #8 > [ 2.831227] Workqueue: events deferred_probe_work_func > [ 2.836353] Call trace: > ... > [ 2.852532] paint_ptr+0xa0/0xa8 > [ 2.855750] kmemleak_ignore+0x38/0x6c > [ 2.859490] __arm_v7s_alloc_table+0x168/0x1f4 > [ 2.863922] arm_v7s_alloc_pgtable+0x114/0x17c > [ 2.868354] alloc_io_pgtable_ops+0x3c/0x78 > ... > > Fixes: e5fc9753b1a8314 ("iommu/io-pgtable: Add ARMv7 short descriptor support") > Signed-off-by: Nicolas Boichat <drinkcat@chromium.org> > --- > > I only tested this on top of my other series > (https://patchwork.kernel.org/patch/10720495/), but I think the same fix > applies. I'm still a bit confused as to why this only shows up now, as IIUC, > the kmemleak_ignore call was always wrong with L1 tables. > > drivers/iommu/io-pgtable-arm-v7s.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) <formletter> This is not the correct way to submit patches for inclusion in the stable kernel tree. Please read: https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html for how to do this properly. </formletter>
On Mon, Jan 28, 2019 at 05:43:01PM +0800, Nicolas Boichat wrote: > L1 tables are allocated with __get_dma_pages, and therefore already > ignored by kmemleak. > > Without this, the kernel would print this error message on boot, > when the first L1 table is allocated: > > [ 2.810533] kmemleak: Trying to color unknown object at 0xffffffd652388000 as Black > [ 2.818190] CPU: 5 PID: 39 Comm: kworker/5:0 Tainted: G S 4.19.16 #8 > [ 2.831227] Workqueue: events deferred_probe_work_func > [ 2.836353] Call trace: > ... > [ 2.852532] paint_ptr+0xa0/0xa8 > [ 2.855750] kmemleak_ignore+0x38/0x6c > [ 2.859490] __arm_v7s_alloc_table+0x168/0x1f4 > [ 2.863922] arm_v7s_alloc_pgtable+0x114/0x17c > [ 2.868354] alloc_io_pgtable_ops+0x3c/0x78 > ... > > Fixes: e5fc9753b1a8314 ("iommu/io-pgtable: Add ARMv7 short descriptor support") > Signed-off-by: Nicolas Boichat <drinkcat@chromium.org> > --- > > I only tested this on top of my other series > (https://patchwork.kernel.org/patch/10720495/), but I think the same fix > applies. I'm still a bit confused as to why this only shows up now, as IIUC, > the kmemleak_ignore call was always wrong with L1 tables. Yes, I managed to reproduce this on top of -rc4 (see below). I suspect you /are/ the intersection of people using v7s w/ kmemleak, so this has just lingered and never been hit until now. For the patch (assuming this is going via Joerg): Acked-by: Will Deacon <will.deacon@arm.com> Will --->8 [ 0.124473] kmemleak: Trying to color unknown object at 0xffff8000042d8000 as Black [ 0.125312] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.0.0-rc4-00012-g40b114779944 #1 [ 0.126181] Hardware name: linux,dummy-virt (DT) [ 0.126680] Call trace: [ 0.126950] dump_backtrace+0x0/0x140 [ 0.127346] show_stack+0x14/0x20 [ 0.127706] dump_stack+0x90/0xb4 [ 0.128066] paint_ptr+0x94/0xa8 [ 0.128417] kmemleak_ignore+0x54/0x60 [ 0.128991] __arm_v7s_alloc_table+0x6c/0x240 [ 0.129661] arm_v7s_alloc_pgtable+0x10c/0x188 [ 0.130359] alloc_io_pgtable_ops+0x44/0xb0 [ 0.131006] arm_v7s_do_selftests+0x84/0x4bc [ 0.131663] do_one_initcall+0x74/0x178 [ 0.132253] kernel_init_freeable+0x188/0x220 [ 0.132923] kernel_init+0x10/0x100 [ 0.133460] ret_from_fork+0x10/0x18 [ 0.142102] arm-v7s io-pgtable: self test ok
Joerg: Just to make sure, is this patch in your queue? Thanks. On Thu, Jan 31, 2019 at 2:21 AM Will Deacon <will.deacon@arm.com> wrote: > > On Mon, Jan 28, 2019 at 05:43:01PM +0800, Nicolas Boichat wrote: > > L1 tables are allocated with __get_dma_pages, and therefore already > > ignored by kmemleak. > > > > Without this, the kernel would print this error message on boot, > > when the first L1 table is allocated: > > > > [ 2.810533] kmemleak: Trying to color unknown object at 0xffffffd652388000 as Black > > [ 2.818190] CPU: 5 PID: 39 Comm: kworker/5:0 Tainted: G S 4.19.16 #8 > > [ 2.831227] Workqueue: events deferred_probe_work_func > > [ 2.836353] Call trace: > > ... > > [ 2.852532] paint_ptr+0xa0/0xa8 > > [ 2.855750] kmemleak_ignore+0x38/0x6c > > [ 2.859490] __arm_v7s_alloc_table+0x168/0x1f4 > > [ 2.863922] arm_v7s_alloc_pgtable+0x114/0x17c > > [ 2.868354] alloc_io_pgtable_ops+0x3c/0x78 > > ... > > > > Fixes: e5fc9753b1a8314 ("iommu/io-pgtable: Add ARMv7 short descriptor support") > > Signed-off-by: Nicolas Boichat <drinkcat@chromium.org> > > --- > > > > I only tested this on top of my other series > > (https://patchwork.kernel.org/patch/10720495/), but I think the same fix > > applies. I'm still a bit confused as to why this only shows up now, as IIUC, > > the kmemleak_ignore call was always wrong with L1 tables. > > Yes, I managed to reproduce this on top of -rc4 (see below). I suspect you > /are/ the intersection of people using v7s w/ kmemleak, so this has just > lingered and never been hit until now. > > For the patch (assuming this is going via Joerg): > > Acked-by: Will Deacon <will.deacon@arm.com> > > Will > > --->8 > > [ 0.124473] kmemleak: Trying to color unknown object at 0xffff8000042d8000 as Black > [ 0.125312] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.0.0-rc4-00012-g40b114779944 #1 > [ 0.126181] Hardware name: linux,dummy-virt (DT) > [ 0.126680] Call trace: > [ 0.126950] dump_backtrace+0x0/0x140 > [ 0.127346] show_stack+0x14/0x20 > [ 0.127706] dump_stack+0x90/0xb4 > [ 0.128066] paint_ptr+0x94/0xa8 > [ 0.128417] kmemleak_ignore+0x54/0x60 > [ 0.128991] __arm_v7s_alloc_table+0x6c/0x240 > [ 0.129661] arm_v7s_alloc_pgtable+0x10c/0x188 > [ 0.130359] alloc_io_pgtable_ops+0x44/0xb0 > [ 0.131006] arm_v7s_do_selftests+0x84/0x4bc > [ 0.131663] do_one_initcall+0x74/0x178 > [ 0.132253] kernel_init_freeable+0x188/0x220 > [ 0.132923] kernel_init+0x10/0x100 > [ 0.133460] ret_from_fork+0x10/0x18 > [ 0.142102] arm-v7s io-pgtable: self test ok
On Mon, Feb 25, 2019 at 08:21:46AM +0800, Nicolas Boichat wrote:
> Joerg: Just to make sure, is this patch in your queue? Thanks.
It is now, thanks.
diff --git a/drivers/iommu/io-pgtable-arm-v7s.c b/drivers/iommu/io-pgtable-arm-v7s.c index cec29bf45c9bd7..98a4a4a0dfb0c1 100644 --- a/drivers/iommu/io-pgtable-arm-v7s.c +++ b/drivers/iommu/io-pgtable-arm-v7s.c @@ -217,7 +217,8 @@ static void *__arm_v7s_alloc_table(int lvl, gfp_t gfp, if (dma != phys) goto out_unmap; } - kmemleak_ignore(table); + if (lvl == 2) + kmemleak_ignore(table); return table; out_unmap:
L1 tables are allocated with __get_dma_pages, and therefore already ignored by kmemleak. Without this, the kernel would print this error message on boot, when the first L1 table is allocated: [ 2.810533] kmemleak: Trying to color unknown object at 0xffffffd652388000 as Black [ 2.818190] CPU: 5 PID: 39 Comm: kworker/5:0 Tainted: G S 4.19.16 #8 [ 2.831227] Workqueue: events deferred_probe_work_func [ 2.836353] Call trace: ... [ 2.852532] paint_ptr+0xa0/0xa8 [ 2.855750] kmemleak_ignore+0x38/0x6c [ 2.859490] __arm_v7s_alloc_table+0x168/0x1f4 [ 2.863922] arm_v7s_alloc_pgtable+0x114/0x17c [ 2.868354] alloc_io_pgtable_ops+0x3c/0x78 ... Fixes: e5fc9753b1a8314 ("iommu/io-pgtable: Add ARMv7 short descriptor support") Signed-off-by: Nicolas Boichat <drinkcat@chromium.org> --- I only tested this on top of my other series (https://patchwork.kernel.org/patch/10720495/), but I think the same fix applies. I'm still a bit confused as to why this only shows up now, as IIUC, the kmemleak_ignore call was always wrong with L1 tables. drivers/iommu/io-pgtable-arm-v7s.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)