Re: [PATCH 1/3] drm/gpusvm, drm/pagemap: Move migration functionality to drm_pagemap

[kernel test robot <lkp@intel.com>]

Hi Thomas,

kernel test robot noticed the following build warnings:

[auto build test WARNING on next-20250417]
[cannot apply to drm-xe/drm-xe-next drm-exynos/exynos-drm-next linus/master drm/drm-next drm-intel/for-linux-next drm-intel/for-linux-next-fixes v6.15-rc2 v6.15-rc1 v6.14 v6.15-rc5]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Thomas-Hellstr-m/drm-gpusvm-drm-pagemap-Move-migration-functionality-to-drm_pagemap/20250417-193413
base:   next-20250417
patch link:    https://lore.kernel.org/r/20250417105113.48831-2-thomas.hellstrom%40linux.intel.com
patch subject: [PATCH 1/3] drm/gpusvm, drm/pagemap: Move migration functionality to drm_pagemap
config: x86_64-allyesconfig (https://download.01.org/0day-ci/archive/20250507/202505072016.4HmSIZT8-lkp@intel.com/config)
compiler: clang version 20.1.2 (https://github.com/llvm/llvm-project 58df0ef89dd64126512e4ee27b4ac3fd8ddf6247)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250507/202505072016.4HmSIZT8-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202505072016.4HmSIZT8-lkp@intel.com/

All warnings (new ones prefixed by >>):

>> drivers/gpu/drm/drm_pagemap.c:653:7: warning: variable 'dev' is used uninitialized whenever 'if' condition is true [-Wsometimes-uninitialized]
     653 |                 if (!page)
         |                     ^~~~~
   drivers/gpu/drm/drm_pagemap.c:683:34: note: uninitialized use occurs here
     683 |         drm_pagemap_migrate_unmap_pages(dev, dma_addr, npages,
         |                                         ^~~
   drivers/gpu/drm/drm_pagemap.c:653:3: note: remove the 'if' if its condition is always false
     653 |                 if (!page)
         |                 ^~~~~~~~~~
     654 |                         goto err_finalize;
         |                         ~~~~~~~~~~~~~~~~~
   drivers/gpu/drm/drm_pagemap.c:602:20: note: initialize the variable 'dev' to silence this warning
     602 |         struct device *dev;
         |                           ^
         |                            = NULL
   1 warning generated.


vim +653 drivers/gpu/drm/drm_pagemap.c

   571	
   572	/**
   573	 * __drm_pagemap_migrate_to_ram() - Migrate a virtual range to RAM (internal)
   574	 * @vas: Pointer to the VM area structure
   575	 * @device_private_page_owner: Device private pages owner
   576	 * @page: Pointer to the page for fault handling (can be NULL)
   577	 * @fault_addr: Fault address
   578	 * @size: Size of migration
   579	 *
   580	 * This internal function performs the migration of the specified GPU SVM range
   581	 * to RAM. It sets up the migration, populates + dma maps RAM PFNs, and
   582	 * invokes the driver-specific operations for migration to RAM.
   583	 *
   584	 * Returns:
   585	 * 0 on success, negative error code on failure.
   586	 */
   587	static int __drm_pagemap_migrate_to_ram(struct vm_area_struct *vas,
   588						void *device_private_page_owner,
   589						struct page *page,
   590						unsigned long fault_addr,
   591						unsigned long size)
   592	{
   593		struct migrate_vma migrate = {
   594			.vma		= vas,
   595			.pgmap_owner	= device_private_page_owner,
   596			.flags		= MIGRATE_VMA_SELECT_DEVICE_PRIVATE |
   597			MIGRATE_VMA_SELECT_DEVICE_COHERENT,
   598			.fault_page	= page,
   599		};
   600		struct drm_pagemap_zdd *zdd;
   601		const struct drm_pagemap_devmem_ops *ops;
   602		struct device *dev;
   603		unsigned long npages, mpages = 0;
   604		struct page **pages;
   605		dma_addr_t *dma_addr;
   606		unsigned long start, end;
   607		void *buf;
   608		int i, err = 0;
   609	
   610		start = ALIGN_DOWN(fault_addr, size);
   611		end = ALIGN(fault_addr + 1, size);
   612	
   613		/* Corner where VMA area struct has been partially unmapped */
   614		if (start < vas->vm_start)
   615			start = vas->vm_start;
   616		if (end > vas->vm_end)
   617			end = vas->vm_end;
   618	
   619		migrate.start = start;
   620		migrate.end = end;
   621		npages = npages_in_range(start, end);
   622	
   623		buf = kvcalloc(npages, 2 * sizeof(*migrate.src) + sizeof(*dma_addr) +
   624			       sizeof(*pages), GFP_KERNEL);
   625		if (!buf) {
   626			err = -ENOMEM;
   627			goto err_out;
   628		}
   629		dma_addr = buf + (2 * sizeof(*migrate.src) * npages);
   630		pages = buf + (2 * sizeof(*migrate.src) + sizeof(*dma_addr)) * npages;
   631	
   632		migrate.vma = vas;
   633		migrate.src = buf;
   634		migrate.dst = migrate.src + npages;
   635	
   636		err = migrate_vma_setup(&migrate);
   637		if (err)
   638			goto err_free;
   639	
   640		/* Raced with another CPU fault, nothing to do */
   641		if (!migrate.cpages)
   642			goto err_free;
   643	
   644		if (!page) {
   645			for (i = 0; i < npages; ++i) {
   646				if (!(migrate.src[i] & MIGRATE_PFN_MIGRATE))
   647					continue;
   648	
   649				page = migrate_pfn_to_page(migrate.src[i]);
   650				break;
   651			}
   652	
 > 653			if (!page)
   654				goto err_finalize;
   655		}
   656		zdd = page->zone_device_data;
   657		ops = zdd->devmem_allocation->ops;
   658		dev = zdd->devmem_allocation->dev;
   659	
   660		err = drm_pagemap_migrate_populate_ram_pfn(vas, page, npages, &mpages,
   661							   migrate.src, migrate.dst,
   662							   start);
   663		if (err)
   664			goto err_finalize;
   665	
   666		err = drm_pagemap_migrate_map_pages(dev, dma_addr, migrate.dst, npages,
   667						    DMA_FROM_DEVICE);
   668		if (err)
   669			goto err_finalize;
   670	
   671		for (i = 0; i < npages; ++i)
   672			pages[i] = migrate_pfn_to_page(migrate.src[i]);
   673	
   674		err = ops->copy_to_ram(pages, dma_addr, npages);
   675		if (err)
   676			goto err_finalize;
   677	
   678	err_finalize:
   679		if (err)
   680			drm_pagemap_migration_unlock_put_pages(npages, migrate.dst);
   681		migrate_vma_pages(&migrate);
   682		migrate_vma_finalize(&migrate);
   683		drm_pagemap_migrate_unmap_pages(dev, dma_addr, npages,
   684						DMA_FROM_DEVICE);
   685	err_free:
   686		kvfree(buf);
   687	err_out:
   688	
   689		return err;
   690	}
   691	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki