linux-kernel - Re: [PATCH] mm: correct type for vmalloc vm

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: <3c223f62-4aca-4b7d-82e6-a1fbf758de26@lucifer.local>
Date: Tue, 29 Jul 2025 13:32:10 +0100
From: Lorenzo Stoakes <lorenzo.stoakes@...cle.com>
To: Harry Yoo <harry.yoo@...cle.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>, Mike Rapoport <rppt@...nel.org>,
        David Hildenbrand <david@...hat.com>,
        "Liam R . Howlett" <Liam.Howlett@...cle.com>,
        Vlastimil Babka <vbabka@...e.cz>,
        Suren Baghdasaryan <surenb@...gle.com>, Michal Hocko <mhocko@...e.com>,
        Jann Horn <jannh@...gle.com>, Pedro Falcato <pfalcato@...e.de>,
        linux-mm@...ck.org, linux-kernel@...r.kernel.org,
        Uladzislau Rezki <urezki@...il.com>
Subject: Re: [PATCH] mm: correct type for vmalloc vm_flags fields

On Tue, Jul 29, 2025 at 09:10:50PM +0900, Harry Yoo wrote:
> I see one more thing in patch 3 of the series:

Can we please try to do this review closer to the series in future, and
ideally in one go. Doing things this way isn't helpful.

While you're right (this function is just awful, no wonder - struct
vm_struct vma...!), the vm_flags_t series is in NO WAY broken, since
vm_flags_t == unsigned long.

I expect there to be a few 'stragglers' of one kind or another, when I do
the next stage of the work _all such callers_ will be detected (as the
kernel will not compile otherwise).

Anyway of course we should fix this, the TL; DR is this isn't an urgent
fix.

I'll send a fix-patch.

>
> diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c
> index 8fcf59ba39db..248d96349fd0 100644
> --- a/arch/arm64/mm/mmu.c
> +++ b/arch/arm64/mm/mmu.c
> @@ -720,7 +720,7 @@ void mark_rodata_ro(void)
>
>  static void __init declare_vma(struct vm_struct *vma,

^ this is beyond appalling.

Somebody from arm side should really fix this because this is just insane.

>  			       void *va_start, void *va_end,
> -			       unsigned long vm_flags)
> +			       vm_flags_t vm_flags)
>  {
>  	phys_addr_t pa_start = __pa_symbol(va_start);
>  	unsigned long size = va_end - va_start;