linux-kernel - Re: [PATCH -next 1/2] kho: fix missing early_memunmap() call in kho

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: <2vxzv7g4sof5.fsf@kernel.org>
Date: Tue, 10 Feb 2026 14:45:02 +0100
From: Pratyush Yadav <pratyush@...nel.org>
To: ranxiaokai627@....com
Cc: graf@...zon.com,  rppt@...nel.org,  pasha.tatashin@...een.com,
  pratyush@...nel.org,  akpm@...ux-foundation.org,
  kexec@...ts.infradead.org,  linux-mm@...ck.org,
  linux-kernel@...r.kernel.org,  ran.xiaokai@....com.cn,
  stable@...r.kernel.org
Subject: Re: [PATCH -next 1/2] kho: fix missing early_memunmap() call in
 kho_populate()

Hi Ran,

Thanks for the fix.

On Fri, Feb 06 2026, ranxiaokai627@....com wrote:

> From: Ran Xiaokai <ran.xiaokai@....com.cn>
>
> kho_populate() returns without calling early_memunmap() on success
> path, this will cause early ioremap virtual address space leak.
>
> Fixes: b50634c5e84a ("kho: cleanup error handling in kho_populate()")
> Signed-off-by: Ran Xiaokai <ran.xiaokai@....com.cn>
> ---
>
> b50634c5e84a ("kho: cleanup error handling in kho_populate()")
> has not landed in upstream, so
> Cc: <stable@...r.kernel.org> is unnecessary?
>
>  kernel/liveupdate/kexec_handover.c | 8 +++++---
>  1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/kernel/liveupdate/kexec_handover.c b/kernel/liveupdate/kexec_handover.c
> index fb3a7b67676e..76b714db175d 100644
> --- a/kernel/liveupdate/kexec_handover.c
> +++ b/kernel/liveupdate/kexec_handover.c
> @@ -1463,6 +1463,7 @@ void __init kho_populate(phys_addr_t fdt_phys, u64 fdt_len,
>  	struct kho_scratch *scratch = NULL;
>  	phys_addr_t mem_map_phys;
>  	void *fdt = NULL;
> +	int populated = 0;

Nit: Please use a bool and true/false. I think it reads much nicer.

>  	int err;
>  
>  	/* Validate the input FDT */
> @@ -1529,16 +1530,17 @@ void __init kho_populate(phys_addr_t fdt_phys, u64 fdt_len,
>  	kho_in.scratch_phys = scratch_phys;
>  	kho_in.mem_map_phys = mem_map_phys;
>  	kho_scratch_cnt = scratch_cnt;
> -	pr_info("found kexec handover data.\n");
>  
> -	return;
> +	populated = 1;
> +	pr_info("found kexec handover data.\n");
>  
>  err_unmap_scratch:
>  	early_memunmap(scratch, scratch_len);
>  err_unmap_fdt:
>  	early_memunmap(fdt, fdt_len);
>  err_report:

Nit: now that this code can be reached by non-error paths, we should
re-name the labels. I think dropping the "err_" prefix should be enough.

With these fixed,

Reviewed-by: Pratyush Yadav <pratyush@...nel.org>

> -	pr_warn("disabling KHO revival\n");
> +	if (!populated)
> +		pr_warn("disabling KHO revival\n");
>  }
>  
>  /* Helper functions for kexec_file_load */

-- 
Regards,
Pratyush Yadav