Linux KVM/arm64 development list
 help / color / mirror / Atom feed
From: Oliver Upton <oliver.upton@linux.dev>
To: Dan Carpenter <dan.carpenter@linaro.org>
Cc: kvmarm@lists.linux.dev
Subject: Re: [bug report] KVM: arm64: Always invalidate TLB for stage-2 permission faults
Date: Thu, 1 Feb 2024 17:59:40 +0000	[thread overview]
Message-ID: <ZbvcDMPTWQIf5Vj8@linux.dev> (raw)
In-Reply-To: <1479cd6b-a10c-4459-8b34-318865cb807a@moroto.mountain>

On Thu, Feb 01, 2024 at 01:17:15PM +0300, Dan Carpenter wrote:
> Hello Oliver Upton,
> 
> The patch be097997a273: "KVM: arm64: Always invalidate TLB for
> stage-2 permission faults" from Sep 22, 2023 (linux-next), leads to
> the following Smatch static checker warning:
> 
> 	arch/arm64/kvm/hyp/nvhe/../pgtable.c:1342 kvm_pgtable_stage2_relax_perms()
> 	error: uninitialized symbol 'level'.
> 
> arch/arm64/kvm/hyp/nvhe/../pgtable.c
>     1319 int kvm_pgtable_stage2_relax_perms(struct kvm_pgtable *pgt, u64 addr,
>     1320                                    enum kvm_pgtable_prot prot)
>     1321 {
>     1322         int ret;
>     1323         s8 level;
>     1324         kvm_pte_t set = 0, clr = 0;
>     1325 
>     1326         if (prot & KVM_PTE_LEAF_ATTR_HI_SW)
>     1327                 return -EINVAL;
>     1328 
>     1329         if (prot & KVM_PGTABLE_PROT_R)
>     1330                 set |= KVM_PTE_LEAF_ATTR_LO_S2_S2AP_R;
>     1331 
>     1332         if (prot & KVM_PGTABLE_PROT_W)
>     1333                 set |= KVM_PTE_LEAF_ATTR_LO_S2_S2AP_W;
>     1334 
>     1335         if (prot & KVM_PGTABLE_PROT_X)
>     1336                 clr |= KVM_PTE_LEAF_ATTR_HI_S2_XN;
>     1337 
>     1338         ret = stage2_update_leaf_attrs(pgt, addr, 1, set, clr, NULL, &level,
>     1339                                        KVM_PGTABLE_WALK_HANDLE_FAULT |
>     1340                                        KVM_PGTABLE_WALK_SHARED);
>     1341         if (!ret || ret == -EAGAIN)
> --> 1342                 kvm_call_hyp(__kvm_tlb_flush_vmid_ipa_nsh, pgt->mmu, addr, level);
>                                                                                     ^^^^^
> level is uninitialized if ret is -EAGAIN

Hmm, race detection can happen at any level in the walk and that may not
be a leaf level. We can initialize level to -1 which would provide no
TTL hint for this case.

-- 
Thanks,
Oliver

      reply	other threads:[~2024-02-01 17:59 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-01 10:17 [bug report] KVM: arm64: Always invalidate TLB for stage-2 permission faults Dan Carpenter
2024-02-01 17:59 ` Oliver Upton [this message]

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=ZbvcDMPTWQIf5Vj8@linux.dev \
    --to=oliver.upton@linux.dev \
    --cc=dan.carpenter@linaro.org \
    --cc=kvmarm@lists.linux.dev \
    /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