Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Andrzej Hajda <andrzej.hajda@intel.com>
To: Matthew Auld <matthew.auld@intel.com>, intel-xe@lists.freedesktop.org
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Subject: Re: [PATCH 17/20] drm/xe: reset mmio mappings with devm
Date: Mon, 13 May 2024 10:12:14 +0200	[thread overview]
Message-ID: <bde72c53-0cef-4fc0-948e-7964801872ec@intel.com> (raw)
In-Reply-To: <20240510181212.264622-39-matthew.auld@intel.com>

On 10.05.2024 20:12, Matthew Auld wrote:
> Set our various mmio mappings to NULL. This should make it easier to
> catch something rogue trying to mess with mmio after device removal. For
> example, we might unmap everything and then start hitting some mmio
> address which has already been unmamped by us and then remapped by
> something else, causing all kinds of carnage.
> 
> Signed-off-by: Matthew Auld <matthew.auld@intel.com>
> Cc: Andrzej Hajda <andrzej.hajda@intel.com>
> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>

Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>

Regards
Andrzej
> ---
>   drivers/gpu/drm/xe/xe_device.c |  4 +++-
>   drivers/gpu/drm/xe/xe_mmio.c   | 35 ++++++++++++++++++++++++++++------
>   drivers/gpu/drm/xe/xe_mmio.h   |  2 +-
>   3 files changed, 33 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c
> index 40983ad66e3c..ce8bccbb06ba 100644
> --- a/drivers/gpu/drm/xe/xe_device.c
> +++ b/drivers/gpu/drm/xe/xe_device.c
> @@ -554,7 +554,9 @@ int xe_device_probe(struct xe_device *xe)
>   	if (err)
>   		return err;
>   
> -	xe_mmio_probe_tiles(xe);
> +	err = xe_mmio_probe_tiles(xe);
> +	if (err)
> +		return err;
>   
>   	xe_ttm_sys_mgr_init(xe);
>   
> diff --git a/drivers/gpu/drm/xe/xe_mmio.c b/drivers/gpu/drm/xe/xe_mmio.c
> index a3094e741db8..0ad840bab8f3 100644
> --- a/drivers/gpu/drm/xe/xe_mmio.c
> +++ b/drivers/gpu/drm/xe/xe_mmio.c
> @@ -255,6 +255,21 @@ static int xe_mmio_tile_vram_size(struct xe_tile *tile, u64 *vram_size,
>   	return xe_force_wake_put(gt_to_fw(gt), XE_FW_GT);
>   }
>   
> +static void vram_fini(void *arg)
> +{
> +	struct xe_device *xe = arg;
> +	struct xe_tile *tile;
> +	int id;
> +
> +	if (xe->mem.vram.mapping)
> +		iounmap(xe->mem.vram.mapping);
> +
> +	xe->mem.vram.mapping = NULL;
> +
> +	for_each_tile(tile, xe, id)
> +		tile->mem.vram.mapping = NULL;
> +}
> +
>   int xe_mmio_probe_vram(struct xe_device *xe)
>   {
>   	struct xe_tile *tile;
> @@ -331,10 +346,20 @@ int xe_mmio_probe_vram(struct xe_device *xe)
>   	drm_info(&xe->drm, "Available VRAM: %pa, %pa\n", &xe->mem.vram.io_start,
>   		 &available_size);
>   
> -	return 0;
> +	return devm_add_action_or_reset(xe->drm.dev, vram_fini, xe);
>   }
>   
> -void xe_mmio_probe_tiles(struct xe_device *xe)
> +static void tiles_fini(void *arg)
> +{
> +	struct xe_device *xe = arg;
> +	struct xe_tile *tile;
> +	int id;
> +
> +	for_each_tile(tile, xe, id)
> +		tile->mmio.regs = NULL;
> +}
> +
> +int xe_mmio_probe_tiles(struct xe_device *xe)
>   {
>   	size_t tile_mmio_size = SZ_16M, tile_mmio_ext_size = xe->info.tile_mmio_ext_size;
>   	u8 id, tile_count = xe->info.tile_count;
> @@ -385,6 +410,8 @@ void xe_mmio_probe_tiles(struct xe_device *xe)
>   			regs += tile_mmio_ext_size;
>   		}
>   	}
> +
> +	return devm_add_action_or_reset(xe->drm.dev, tiles_fini, xe);
>   }
>   
>   static void mmio_fini(void *arg)
> @@ -392,10 +419,6 @@ static void mmio_fini(void *arg)
>   	struct xe_device *xe = arg;
>   
>   	pci_iounmap(to_pci_dev(xe->drm.dev), xe->mmio.regs);
> -	if (xe->mem.vram.mapping)
> -		iounmap(xe->mem.vram.mapping);
> -
> -	xe->mem.vram.mapping = NULL;
>   	xe->mmio.regs = NULL;
>   }
>   
> diff --git a/drivers/gpu/drm/xe/xe_mmio.h b/drivers/gpu/drm/xe/xe_mmio.h
> index 445ec6a0753e..6488705c1ffe 100644
> --- a/drivers/gpu/drm/xe/xe_mmio.h
> +++ b/drivers/gpu/drm/xe/xe_mmio.h
> @@ -21,7 +21,7 @@ struct xe_device;
>   #define LMEM_BAR		2
>   
>   int xe_mmio_init(struct xe_device *xe);
> -void xe_mmio_probe_tiles(struct xe_device *xe);
> +int xe_mmio_probe_tiles(struct xe_device *xe);
>   
>   u8 xe_mmio_read8(struct xe_gt *gt, struct xe_reg reg);
>   u16 xe_mmio_read16(struct xe_gt *gt, struct xe_reg reg);


  reply	other threads:[~2024-05-13  8:12 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-10 18:12 [PATCH 00/20] core_hotunplug improvements Matthew Auld
2024-05-10 18:12 ` [PATCH 01/20] drm/drm_managed: try to improve the drmm DOC Matthew Auld
2024-05-13  7:06   ` Andrzej Hajda
2024-05-13 15:38   ` Rodrigo Vivi
2024-05-10 18:12 ` [PATCH 02/20] drm/amdgpu: don't trample pdev drvdata Matthew Auld
2024-05-13 11:54   ` Christian König
2024-05-10 18:12 ` [PATCH 03/20] drm/xe/pci: remove broken driver_release Matthew Auld
2024-05-13  7:14   ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 04/20] drm/xe: covert sysfs over to devm Matthew Auld
2024-05-10 18:12 ` [PATCH 05/20] drm/xe/ggtt: use drm_dev_enter to mark device section Matthew Auld
2024-05-13  7:16   ` Andrzej Hajda
2024-05-13 19:34   ` Randhawa, Jagmeet
2024-05-10 18:12 ` [PATCH 06/20] drm/xe/guc: move guc_fini over to devm Matthew Auld
2024-05-10 18:12 ` [PATCH 07/20] drm/xe/guc: s/guc_fini/guc_fini_hw/ Matthew Auld
2024-05-13  7:30   ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 08/20] drm/xe/guc_pc: move pc_fini to devm Matthew Auld
2024-05-10 18:12 ` [PATCH 09/20] drm/xe/guc_pc: s/pc_fini/pc_fini_hw/ Matthew Auld
2024-05-11  9:19   ` kernel test robot
2024-05-13  7:37   ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 10/20] drm/xe/irq: move irq_uninstall over to devm Matthew Auld
2024-05-10 18:12 ` [PATCH 11/20] drm/xe/device: move flr " Matthew Auld
2024-05-10 18:12 ` [PATCH 12/20] drm/xe/device: move xe_device_sanitize over " Matthew Auld
2024-05-10 18:12 ` [PATCH 13/20] drm/xe/coredump: move " Matthew Auld
2024-05-13  7:38   ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 14/20] drm/xe/gt: break out gt_fini into sw vs hw state Matthew Auld
2024-05-13  7:50   ` Andrzej Hajda
2024-05-13  8:37     ` Matthew Auld
2024-05-13  9:19       ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 15/20] drm/xe: make gt_remove use devm Matthew Auld
2024-05-13  7:58   ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 16/20] drm/xe/mmio: move mmio_fini over to devm Matthew Auld
2024-05-13  8:00   ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 17/20] drm/xe: reset mmio mappings with devm Matthew Auld
2024-05-13  8:12   ` Andrzej Hajda [this message]
2024-05-10 18:12 ` [PATCH 18/20] drm/xe/display: move display fini stuff to devm Matthew Auld
2024-05-13  8:13   ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 19/20] drm/xe/display: stop calling domains_driver_remove twice Matthew Auld
2024-05-13  8:19   ` Andrzej Hajda
2024-05-13  8:27     ` Matthew Auld
2024-05-13  9:17       ` Andrzej Hajda
2024-05-10 18:12 ` [PATCH 20/20] drm/xe/display: move device_remove over to drmm Matthew Auld
2024-05-11 10:33   ` kernel test robot
2024-05-11 10:43   ` kernel test robot
2024-05-13  8:22   ` Andrzej Hajda
2024-05-10 18:17 ` ✓ CI.Patch_applied: success for core_hotunplug improvements Patchwork
2024-05-10 18:18 ` ✗ CI.checkpatch: warning " Patchwork
2024-05-10 18:18 ` ✗ CI.KUnit: failure " Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bde72c53-0cef-4fc0-948e-7964801872ec@intel.com \
    --to=andrzej.hajda@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=matthew.auld@intel.com \
    --cc=rodrigo.vivi@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox