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: <ZHb3tVHVV2l1BJBl@MiWiFi-R3L-srv>
Date:   Wed, 31 May 2023 15:31:01 +0800
From:   Baoquan He <bhe@...hat.com>
To:     "Leizhen (ThunderTown)" <thunder.leizhen@...wei.com>
Cc:     Eric Biederman <ebiederm@...ssion.com>, kexec@...ts.infradead.org,
        linux-kernel@...r.kernel.org,
        Michael Holzheu <holzheu@...ux.vnet.ibm.com>,
        Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [PATCH 1/6] kexec: fix a memory leak in crash_shrink_memory()

On 05/31/23 at 09:16am, Leizhen (ThunderTown) wrote:
> 
> 
> On 2023/5/31 8:13, Baoquan He wrote:
> > On 05/27/23 at 08:34pm, Zhen Lei wrote:
> >> If the value of parameter 'new_size' is in the semi-open and semi-closed
> >> interval (crashk_res.end - KEXEC_CRASH_MEM_ALIGN + 1, crashk_res.end], the
> >> calculation result of ram_res is:
> >> 	ram_res->start = crashk_res.end + 1
> >> 	ram_res->end   = crashk_res.end
> > 
> > If the new_size is smaller than KEXEC_CRASH_MEM_ALIGN, does it make
> > any sense except of testing purpose? Do we need to fail this kind of
> > shrinking, or just shrink all the left crash memory?

OK, I misread your log. You are saying the new_size is close to
crashk_res.end but has a tiny difference in your example, I
thought the new_size is smaller than KEXEC_CRASH_MEM_ALIGN which is just
in the opposite direction.

Yea, it does have the possibility to waste a ram_res but does nothing
even though the chance is very small.

Acked-by: Baoquan He <bhe@...hat.com>

> 
> We can't give a fixed value, that is, how much crash memory is reserved to
> ensure that the capture kernel runs. The size of KEXEC_CRASH_MEM_ALIGN is
> only one page on non-s390 platforms. So, it's better to keep the code simple,
> and let the user(administrator) shrink the crash memory reasonably.
> 
> include/linux/kexec.h
> #define KEXEC_CRASH_MEM_ALIGN	PAGE_SIZE

> 
> > 
> >> The operation of function insert_resource() fails, and ram_res is not
> >> added to iomem_resource. As a result, the memory of the control block
> >> ram_res is leaked.
> >>
> >> In fact, on all architectures, the start address and size of crashk_res
> >> are already aligned by KEXEC_CRASH_MEM_ALIGN. Therefore, we do not need to
> >> round up crashk_res.start again. Instead, we should round up 'new_size'
> >> in advance.
> >>
> >> Fixes: 6480e5a09237 ("kdump: add missing RAM resource in crash_shrink_memory()")
> >> Fixes: 06a7f711246b ("kexec: premit reduction of the reserved memory size")
> >> Signed-off-by: Zhen Lei <thunder.leizhen@...wei.com>
> >> ---
> >>  kernel/kexec_core.c | 5 ++---
> >>  1 file changed, 2 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c
> >> index 3d578c6fefee385..22acee18195a591 100644
> >> --- a/kernel/kexec_core.c
> >> +++ b/kernel/kexec_core.c
> >> @@ -1122,6 +1122,7 @@ int crash_shrink_memory(unsigned long new_size)
> >>  	start = crashk_res.start;
> >>  	end = crashk_res.end;
> >>  	old_size = (end == 0) ? 0 : end - start + 1;
> >> +	new_size = roundup(new_size, KEXEC_CRASH_MEM_ALIGN);
> >>  	if (new_size >= old_size) {
> >>  		ret = (new_size == old_size) ? 0 : -EINVAL;
> >>  		goto unlock;
> >> @@ -1133,9 +1134,7 @@ int crash_shrink_memory(unsigned long new_size)
> >>  		goto unlock;
> >>  	}
> >>  
> >> -	start = roundup(start, KEXEC_CRASH_MEM_ALIGN);
> >> -	end = roundup(start + new_size, KEXEC_CRASH_MEM_ALIGN);
> >> -
> >> +	end = start + new_size;
> >>  	crash_free_reserved_phys_range(end, crashk_res.end);
> >>  
> >>  	if ((start == end) && (crashk_res.parent != NULL))
> >> -- 
> >> 2.25.1
> >>
> > 
> > .
> > 
> 
> -- 
> Regards,
>   Zhen Lei
> 

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ