linux-kernel - Re: [PATCH v2] of: reserved_mem: Restructure code to call reserved mem init functions earlier

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1c65ae7a-ac35-4c80-aa06-9ca921a7c919@quicinc.com>
Date: Wed, 26 Jun 2024 08:48:02 -0700
From: Oreoluwa Babatunde <quic_obabatun@...cinc.com>
To: Rob Herring <robh@...nel.org>
CC: <broonie@...nel.org>, <catalin.marinas@....com>, <conor@...nel.org>,
        <devicetree@...r.kernel.org>, <hch@....de>, <iommu@...ts.linux.dev>,
        <kernel@...cinc.com>, <linux-kernel@...r.kernel.org>,
        <m.szyprowski@...sung.com>, <nathan@...nel.org>,
        <oe-kbuild-all@...ts.linux.dev>, <robin.murphy@....com>,
        <saravanak@...gle.com>, <will@...nel.org>
Subject: Re: [PATCH v2] of: reserved_mem: Restructure code to call reserved
 mem init functions earlier


On 6/26/2024 8:16 AM, Rob Herring wrote:
> On Wed, Jun 19, 2024 at 05:10:27PM -0700, Oreoluwa Babatunde wrote:
>> After all the reserved memory regions have been added to the
>> reserved_mem array, a region specific initialization function is called
>> on each of reserved memory regions in a loop to initialize them.
>>
>> With recent changes made to allow the reserved_mem array be dynamically
>> allocated, the cma reserved memory regions are not initialized until
>> after the page tables are setup. This causes the warning seen in the
>> dump stack below:
>>
>> 	WARNING: CPU: 0 PID: 1 at mm/memory.c:2789 __apply_to_page_range+0x360/0x380
>> 	CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.10.0-rc1-00007-ga46cccb0ee2d #1
>> 	Hardware name: Generic DT based system
>> 	Call trace:
>> 	unwind_backtrace from show_stack+0x18/0x1c
>> 	show_stack from dump_stack_lvl+0x54/0x68
>> 	dump_stack_lvl from __warn+0x74/0x114
>> 	__warn from warn_slowpath_fmt+0x13c/0x1c0
>> 	warn_slowpath_fmt from __apply_to_page_range+0x360/0x380
>> 	__apply_to_page_range from apply_to_page_range+0x24/0x2c
>> 	apply_to_page_range from __alloc_from_contiguous+0xc4/0x158
>> 	__alloc_from_contiguous from cma_allocator_alloc+0x3c/0x44
>> 	cma_allocator_alloc from arch_dma_alloc+0x128/0x2b4
>> 	arch_dma_alloc from dma_alloc_attrs+0x90/0x150
>> 	dma_alloc_attrs from drm_gem_dma_create+0xa4/0x13c
>> 	drm_gem_dma_create from drm_gem_dma_create_with_handle+0x24/0xac
>> 	drm_gem_dma_create_with_handle from drm_gem_dma_dumb_create+0x44/0x50
>> 	drm_gem_dma_dumb_create from drm_client_framebuffer_create+0x9c/0x164
>> 	drm_client_framebuffer_create from drm_fbdev_dma_helper_fb_probe+0x84/0x23c
>> 	drm_fbdev_dma_helper_fb_probe from __drm_fb_helper_initial_config_and_unlock+0x2e4/0x4f8
>> 	__drm_fb_helper_initial_config_and_unlock from drm_fbdev_dma_client_hotplug+0x74/0xb8
>> 	drm_fbdev_dma_client_hotplug from drm_client_register+0x5c/0x98
>> 	drm_client_register from aspeed_gfx_probe+0x278/0x3c0
>> 	aspeed_gfx_probe from platform_probe+0x60/0xb8
>> 	platform_probe from really_probe+0xd4/0x3b4
>> 	really_probe from __driver_probe_device+0x90/0x1dc
>> 	__driver_probe_device from driver_probe_device+0x38/0xd0
>> 	driver_probe_device from __driver_attach+0x118/0x1dc
>> 	__driver_attach from bus_for_each_dev+0x84/0xd4
>> 	bus_for_each_dev from bus_add_driver+0xec/0x1f0
>> 	bus_add_driver from driver_register+0x84/0x11c
>> 	driver_register from do_one_initcall+0x84/0x1c8
>> 	do_one_initcall from kernel_init_freeable+0x1a4/0x230
>> 	kernel_init_freeable from kernel_init+0x1c/0x138
>> 	kernel_init from ret_from_fork+0x14/0x28
>> 	Exception stack(0x9f015fb0 to 0x9f015ff8)
>> 	5fa0:                                     00000000 00000000 00000000 00000000
>> 	5fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
>> 	5fe0: 00000000 00000000 00000000 00000000 00000013 00000000
>> 	---[ end trace 0000000000000000 ]---
>> 	aspeed_gfx 1e6e6000.display: [drm] fb0: aspeed-gfx-drmd frame buffer device
>>
>> Hence, restructure the code to initialize the regions as soon as each
>> of them are added to the reserved_mem array.
>>
>> Fixes: a46cccb0ee2d ("of: reserved_mem: Restruture how the reserved memory regions are processed")
>> Signed-off-by: Oreoluwa Babatunde <quic_obabatun@...cinc.com>
>> ---
>> v2:
>> - Fix kernel-doc for of_init_reserved_mem_node() in response to the
>>   below warning from v1:
>>   https://lore.kernel.org/all/202406181626.126X1Nbz-lkp@intel.com/
>>
>> v1:
>>   https://lore.kernel.org/all/20240617193357.3929092-1-quic_obabatun@quicinc.com/
>>
>>  drivers/of/fdt.c             |  2 +-
>>  drivers/of/of_private.h      |  2 +-
>>  drivers/of/of_reserved_mem.c | 83 +++++++++++++++++++++---------------
>>  3 files changed, 50 insertions(+), 37 deletions(-)
> Applied.
>
> In the future, do not send fixes or new versions as replies to previous 
> threads. The default behavior of b4 is to apply v6 of the series because 
> that is the 'newest' version.
>
> Rob
ack.

Thank you!