qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[PATCH 0/2] Fix integer overflows in loading of large images


From: Jamie Iles
Subject: [PATCH 0/2] Fix integer overflows in loading of large images
Date: Thu, 11 Nov 2021 14:11:39 +0000

Most of the loader code currently uses a ssize_t or 64 bit integer type  
to store image lengths, but many functions that handle loading return an 
int with a negative value on error or length on success.  Once an image 
exceeds 2GB this will cause an integer overflow and so can end up 
loading truncated images, silently failing to load an image (a 4GB image 
would be interpreted as 0 bytes long).

This is unlikely to affect many deployments, but can manifest when 
preloading RAM disks for example.

This builds upon 8975eb891fb6 ("hw/elf_ops.h: switch to ssize_t for elf 
loader return type") to cover more of the generic loader.

Jamie Iles (2):
  hw/core/loader: return image sizes as ssize_t
  hw/core/loader: workaround read() size limit.

 hw/arm/armv7m.c          |   2 +-
 hw/arm/boot.c            |   8 +--
 hw/core/generic-loader.c |   2 +-
 hw/core/loader.c         | 121 ++++++++++++++++++++++++---------------
 hw/i386/x86.c            |   2 +-
 hw/riscv/boot.c          |   5 +-
 include/hw/loader.h      |  55 +++++++++---------
 7 files changed, 114 insertions(+), 81 deletions(-)

-- 
2.30.2




reply via email to

[Prev in Thread] Current Thread [Next in Thread]