All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] linux-user/elfload: fix address calculation in fallback scenario
@ 2021-01-31  6:19 Vincent Fazio
  2021-02-13 21:48 ` Laurent Vivier
  2021-03-09 21:23 ` Laurent Vivier
  0 siblings, 2 replies; 4+ messages in thread
From: Vincent Fazio @ 2021-01-31  6:19 UTC (permalink / raw)
  To: qemu-devel; +Cc: qemu-trivial, laurent, Vincent Fazio

From: Vincent Fazio <vfazio@gmail.com>

Previously, guest_loaddr was not taken into account when returning an
address from pgb_find_hole when /proc/self/maps was unavailable which
caused an improper guest_base address to be calculated.

This could cause a SIGSEGV later in load_elf_image -> target_mmap for
ET_EXEC type images since the mmap MAP_FIXED flag is specified which
could clobber existing mappings at the address returnd by g2h().

  mmap(0xd87000, 16846912, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_NORESERVE|0x100000, -1, 0) = 0xd87000
  munmap(0xd87000, 16846912)              = 0
  write(2, "Locating guest address space @ 0"..., 40Locating guest address space @ 0xd87000) = 40
  mmap(0x1187000, 16850944, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS|MAP_NORESERVE, -1, 0) = 0x1187000
  --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_ACCERR, si_addr=0x2188310} ---
  +++ killed by SIGSEGV +++

Now, pgd_find_hole accounts for guest_loaddr in this scenario.

Fixes: ad592e37dfcc ("linux-user: provide fallback pgd_find_hole for bare chroots")
Signed-off-by: Vincent Fazio <vfazio@gmail.com>
---
 linux-user/elfload.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/linux-user/elfload.c b/linux-user/elfload.c
index 8d425f9ed0..6d606b9442 100644
--- a/linux-user/elfload.c
+++ b/linux-user/elfload.c
@@ -2242,7 +2242,8 @@ static uintptr_t pgb_find_hole(uintptr_t guest_loaddr, uintptr_t guest_size,
     brk = (uintptr_t)sbrk(0);
 
     if (!maps) {
-        return pgd_find_hole_fallback(guest_size, brk, align, offset);
+        ret = pgd_find_hole_fallback(guest_size, brk, align, offset);
+        return ret - guest_loaddr;
     }
 
     /* The first hole is before the first map entry. */
-- 
2.30.0



^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH] linux-user/elfload: fix address calculation in fallback scenario
  2021-01-31  6:19 [PATCH] linux-user/elfload: fix address calculation in fallback scenario Vincent Fazio
@ 2021-02-13 21:48 ` Laurent Vivier
  2021-03-09 21:04   ` Vincent Fazio
  2021-03-09 21:23 ` Laurent Vivier
  1 sibling, 1 reply; 4+ messages in thread
From: Laurent Vivier @ 2021-02-13 21:48 UTC (permalink / raw)
  To: Vincent Fazio, qemu-devel; +Cc: qemu-trivial, Alex Bennée, Vincent Fazio

Le 31/01/2021 à 07:19, Vincent Fazio a écrit :
> From: Vincent Fazio <vfazio@gmail.com>
> 
> Previously, guest_loaddr was not taken into account when returning an
> address from pgb_find_hole when /proc/self/maps was unavailable which
> caused an improper guest_base address to be calculated.
> 
> This could cause a SIGSEGV later in load_elf_image -> target_mmap for
> ET_EXEC type images since the mmap MAP_FIXED flag is specified which
> could clobber existing mappings at the address returnd by g2h().
> 
>   mmap(0xd87000, 16846912, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_NORESERVE|0x100000, -1, 0) = 0xd87000
>   munmap(0xd87000, 16846912)              = 0
>   write(2, "Locating guest address space @ 0"..., 40Locating guest address space @ 0xd87000) = 40
>   mmap(0x1187000, 16850944, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS|MAP_NORESERVE, -1, 0) = 0x1187000
>   --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_ACCERR, si_addr=0x2188310} ---
>   +++ killed by SIGSEGV +++
> 
> Now, pgd_find_hole accounts for guest_loaddr in this scenario.
> 
> Fixes: ad592e37dfcc ("linux-user: provide fallback pgd_find_hole for bare chroots")
> Signed-off-by: Vincent Fazio <vfazio@gmail.com>
> ---
>  linux-user/elfload.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/linux-user/elfload.c b/linux-user/elfload.c
> index 8d425f9ed0..6d606b9442 100644
> --- a/linux-user/elfload.c
> +++ b/linux-user/elfload.c
> @@ -2242,7 +2242,8 @@ static uintptr_t pgb_find_hole(uintptr_t guest_loaddr, uintptr_t guest_size,
>      brk = (uintptr_t)sbrk(0);
>  
>      if (!maps) {
> -        return pgd_find_hole_fallback(guest_size, brk, align, offset);
> +        ret = pgd_find_hole_fallback(guest_size, brk, align, offset);
> +        return ret - guest_loaddr;
>      }
>  
>      /* The first hole is before the first map entry. */
> 

Reviewed-by: Laurent Vivier <laurent@vivier.eu>

CC: Alex


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] linux-user/elfload: fix address calculation in fallback scenario
  2021-02-13 21:48 ` Laurent Vivier
@ 2021-03-09 21:04   ` Vincent Fazio
  0 siblings, 0 replies; 4+ messages in thread
From: Vincent Fazio @ 2021-03-09 21:04 UTC (permalink / raw)
  To: Laurent Vivier, qemu-devel; +Cc: qemu-trivial, Alex Bennée, Vincent Fazio

Pinging per Laurent.

On 2/13/21 3:48 PM, Laurent Vivier wrote:
> Le 31/01/2021 à 07:19, Vincent Fazio a écrit :
>> From: Vincent Fazio <vfazio@gmail.com>
>>
>> Previously, guest_loaddr was not taken into account when returning an
>> address from pgb_find_hole when /proc/self/maps was unavailable which
>> caused an improper guest_base address to be calculated.
>>
>> This could cause a SIGSEGV later in load_elf_image -> target_mmap for
>> ET_EXEC type images since the mmap MAP_FIXED flag is specified which
>> could clobber existing mappings at the address returnd by g2h().
>>
>>    mmap(0xd87000, 16846912, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_NORESERVE|0x100000, -1, 0) = 0xd87000
>>    munmap(0xd87000, 16846912)              = 0
>>    write(2, "Locating guest address space @ 0"..., 40Locating guest address space @ 0xd87000) = 40
>>    mmap(0x1187000, 16850944, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS|MAP_NORESERVE, -1, 0) = 0x1187000
>>    --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_ACCERR, si_addr=0x2188310} ---
>>    +++ killed by SIGSEGV +++
>>
>> Now, pgd_find_hole accounts for guest_loaddr in this scenario.
>>
>> Fixes: ad592e37dfcc ("linux-user: provide fallback pgd_find_hole for bare chroots")
>> Signed-off-by: Vincent Fazio <vfazio@gmail.com>
>> ---
>>   linux-user/elfload.c | 3 ++-
>>   1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/linux-user/elfload.c b/linux-user/elfload.c
>> index 8d425f9ed0..6d606b9442 100644
>> --- a/linux-user/elfload.c
>> +++ b/linux-user/elfload.c
>> @@ -2242,7 +2242,8 @@ static uintptr_t pgb_find_hole(uintptr_t guest_loaddr, uintptr_t guest_size,
>>       brk = (uintptr_t)sbrk(0);
>>   
>>       if (!maps) {
>> -        return pgd_find_hole_fallback(guest_size, brk, align, offset);
>> +        ret = pgd_find_hole_fallback(guest_size, brk, align, offset);
>> +        return ret - guest_loaddr;
>>       }
>>   
>>       /* The first hole is before the first map entry. */
>>
> 
> Reviewed-by: Laurent Vivier <laurent@vivier.eu>
> 
> CC: Alex
> 


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] linux-user/elfload: fix address calculation in fallback scenario
  2021-01-31  6:19 [PATCH] linux-user/elfload: fix address calculation in fallback scenario Vincent Fazio
  2021-02-13 21:48 ` Laurent Vivier
@ 2021-03-09 21:23 ` Laurent Vivier
  1 sibling, 0 replies; 4+ messages in thread
From: Laurent Vivier @ 2021-03-09 21:23 UTC (permalink / raw)
  To: Vincent Fazio, qemu-devel; +Cc: qemu-trivial, Vincent Fazio

Le 31/01/2021 à 07:19, Vincent Fazio a écrit :
> From: Vincent Fazio <vfazio@gmail.com>
> 
> Previously, guest_loaddr was not taken into account when returning an
> address from pgb_find_hole when /proc/self/maps was unavailable which
> caused an improper guest_base address to be calculated.
> 
> This could cause a SIGSEGV later in load_elf_image -> target_mmap for
> ET_EXEC type images since the mmap MAP_FIXED flag is specified which
> could clobber existing mappings at the address returnd by g2h().
> 
>   mmap(0xd87000, 16846912, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_NORESERVE|0x100000, -1, 0) = 0xd87000
>   munmap(0xd87000, 16846912)              = 0
>   write(2, "Locating guest address space @ 0"..., 40Locating guest address space @ 0xd87000) = 40
>   mmap(0x1187000, 16850944, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS|MAP_NORESERVE, -1, 0) = 0x1187000
>   --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_ACCERR, si_addr=0x2188310} ---
>   +++ killed by SIGSEGV +++
> 
> Now, pgd_find_hole accounts for guest_loaddr in this scenario.
> 
> Fixes: ad592e37dfcc ("linux-user: provide fallback pgd_find_hole for bare chroots")
> Signed-off-by: Vincent Fazio <vfazio@gmail.com>
> ---
>  linux-user/elfload.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/linux-user/elfload.c b/linux-user/elfload.c
> index 8d425f9ed0..6d606b9442 100644
> --- a/linux-user/elfload.c
> +++ b/linux-user/elfload.c
> @@ -2242,7 +2242,8 @@ static uintptr_t pgb_find_hole(uintptr_t guest_loaddr, uintptr_t guest_size,
>      brk = (uintptr_t)sbrk(0);
>  
>      if (!maps) {
> -        return pgd_find_hole_fallback(guest_size, brk, align, offset);
> +        ret = pgd_find_hole_fallback(guest_size, brk, align, offset);
> +        return ret - guest_loaddr;
>      }
>  
>      /* The first hole is before the first map entry. */
> 

Applied to my linux-user-for-6.0 branch.

Thanks,
Laurent



^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2021-03-09 21:34 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-31  6:19 [PATCH] linux-user/elfload: fix address calculation in fallback scenario Vincent Fazio
2021-02-13 21:48 ` Laurent Vivier
2021-03-09 21:04   ` Vincent Fazio
2021-03-09 21:23 ` Laurent Vivier

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.