arch/s390/kvm/kvm-s390.c:5659 kvm_arch_prepare_memory_region() error: we previously assumed 'new' could be null (see line 5645)

arch/s390/kvm/kvm-s390.c:5659 kvm_arch_prepare_memory_region() error: we previously assumed 'new' could be null (see line 5645)

[kernel test robot]

BCC: lkp@intel.com
CC: oe-kbuild-all@lists.linux.dev
CC: linux-kernel@vger.kernel.org
TO: Claudio Imbrenda <imbrenda@linux.ibm.com>
CC: Steffen Eiden <seiden@linux.ibm.com>

tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
head:   50897c955902c93ae71c38698abb910525ebdc89
commit: 9b8e8aad5896d66005d29920cb1643076a20b172 KVM: s390: ucontrol: Fix memslot handling
date:   5 weeks ago
:::::: branch date: 12 hours ago
:::::: commit date: 5 weeks ago
config: s390-randconfig-r073-20260512 (https://download.01.org/0day-ci/archive/20260512/202605121849.vHgrQKWJ-lkp@intel.com/config)
compiler: s390-linux-gcc (GCC) 8.5.0
smatch: v0.5.0-9065-ge9cc34fd

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
| Fixes: 9b8e8aad5896 ("KVM: s390: ucontrol: Fix memslot handling")
| Reported-by: kernel test robot <lkp@intel.com>
| Reported-by: Dan Carpenter <error27@gmail.com>
| Closes: https://lore.kernel.org/r/202605121849.vHgrQKWJ-lkp@intel.com/

smatch warnings:
arch/s390/kvm/kvm-s390.c:5659 kvm_arch_prepare_memory_region() error: we previously assumed 'new' could be null (see line 5645)

vim +/new +5659 arch/s390/kvm/kvm-s390.c

d663b8a28598607 Paolo Bonzini        2022-11-03  5638  
b0c632db637d68a Heiko Carstens       2008-03-25  5639  /* Section: memory related */
f7784b8ec9b6a04 Marcelo Tosatti      2009-12-23  5640  int kvm_arch_prepare_memory_region(struct kvm *kvm,
537a17b31493009 Sean Christopherson  2021-12-06  5641  				   const struct kvm_memory_slot *old,
537a17b31493009 Sean Christopherson  2021-12-06  5642  				   struct kvm_memory_slot *new,
7b6195a91d60909 Takuya Yoshikawa     2013-02-27  5643  				   enum kvm_mr_change change)
b0c632db637d68a Heiko Carstens       2008-03-25  5644  {
9b8e8aad5896d66 Claudio Imbrenda     2026-04-02 @5645  	if (kvm_is_ucontrol(kvm) && new && new->id < KVM_USER_MEM_SLOTS)
7816e58967d0e6c Christoph Schlameuss 2024-06-24  5646  		return -EINVAL;
7816e58967d0e6c Christoph Schlameuss 2024-06-24  5647  
ec5c86976674d2f Sean Christopherson  2021-12-06  5648  	/* When we are protected, we should not change the memory slots */
ec5c86976674d2f Sean Christopherson  2021-12-06  5649  	if (kvm_s390_pv_get_handle(kvm))
ec5c86976674d2f Sean Christopherson  2021-12-06  5650  		return -EINVAL;
ec5c86976674d2f Sean Christopherson  2021-12-06  5651  
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5652  	if (change != KVM_MR_DELETE && change != KVM_MR_FLAGS_ONLY) {
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5653  		/*
06a20c3ab6042ea Claudio Imbrenda     2026-04-02  5654  		 * A few sanity checks. The memory in userland is ok to be
06a20c3ab6042ea Claudio Imbrenda     2026-04-02  5655  		 * fragmented into various different vmas. It is okay to mmap()
06a20c3ab6042ea Claudio Imbrenda     2026-04-02  5656  		 * and munmap() stuff in this slot after doing this call at any
06a20c3ab6042ea Claudio Imbrenda     2026-04-02  5657  		 * time.
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5658  		 */
06a20c3ab6042ea Claudio Imbrenda     2026-04-02 @5659  		if (new->userspace_addr & ~PAGE_MASK)
b0c632db637d68a Heiko Carstens       2008-03-25  5660  			return -EINVAL;
06a20c3ab6042ea Claudio Imbrenda     2026-04-02  5661  		if ((new->base_gfn + new->npages) * PAGE_SIZE > kvm->arch.mem_limit)
a3a92c31bf0b57a Dominik Dingel       2014-12-01  5662  			return -EINVAL;
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5663  	}
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5664  
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5665  	if (!kvm->arch.migration_mode)
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5666  		return 0;
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5667  
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5668  	/*
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5669  	 * Turn off migration mode when:
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5670  	 * - userspace creates a new memslot with dirty logging off,
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5671  	 * - userspace modifies an existing memslot (MOVE or FLAGS_ONLY) and
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5672  	 *   dirty logging is turned off.
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5673  	 * Migration mode expects dirty page logging being enabled to store
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5674  	 * its dirty bitmap.
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5675  	 */
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5676  	if (change != KVM_MR_DELETE &&
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5677  	    !(new->flags & KVM_MEM_LOG_DIRTY_PAGES))
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5678  		WARN(kvm_s390_vm_stop_migration(kvm),
f2d3155e2a6bac4 Nico Boehr           2023-01-27  5679  		     "Failed to stop migration mode");
a3a92c31bf0b57a Dominik Dingel       2014-12-01  5680  
f7784b8ec9b6a04 Marcelo Tosatti      2009-12-23  5681  	return 0;
f7784b8ec9b6a04 Marcelo Tosatti      2009-12-23  5682  }
f7784b8ec9b6a04 Marcelo Tosatti      2009-12-23  5683  

:::::: The code at line 5659 was first introduced by commit
:::::: 06a20c3ab6042ea7f9927fbeb50aa4e79894c136 KVM: s390: Allow 4k granularity for memslots

:::::: TO: Claudio Imbrenda <imbrenda@linux.ibm.com>
:::::: CC: Claudio Imbrenda <imbrenda@linux.ibm.com>

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