ACK/Cmnt: [[b, d][PATCH 1/1] x86/insn-eval: Fix use-after-free access to LDT entry

Stefan Bader stefan.bader at canonical.com
Thu Jul 18 09:51:07 UTC 2019


On 16.07.19 19:08, Benjamin M Romer wrote:
> From: Jann Horn <jannh at google.com>
> 
> get_desc() computes a pointer into the LDT while holding a lock that
> protects the LDT from being freed, but then drops the lock and returns the
> (now potentially dangling) pointer to its caller.
> 
> Fix it by giving the caller a copy of the LDT entry instead.
> 
> Fixes: 670f928ba09b ("x86/insn-eval: Add utility function to get segment descriptor")
> Cc: stable at vger.kernel.org
> Signed-off-by: Jann Horn <jannh at google.com>
> Signed-off-by: Linus Torvalds <torvalds at linux-foundation.org>
> 
> (cherry picked from commit de9f869616dd95e95c00bdd6b0fcd3421e8a4323)
> CVE-2019-13233
> Signed-off-by: Benjamin M Romer <benjamin.romer at canonical.com>
Acked-by: Stefan Bader <stefan.bader at canonical.com>
> ---

Usually we put the CVE number or BugLink before the cherry pick and sob section
(would propose to do that when applying). NACK for Cosmic as that is EOL.

-Stefan

>  arch/x86/lib/insn-eval.c | 47 ++++++++++++++++++++--------------------
>  1 file changed, 24 insertions(+), 23 deletions(-)
> 
> diff --git a/arch/x86/lib/insn-eval.c b/arch/x86/lib/insn-eval.c
> index 9119d8e41f1f..87dcba101e56 100644
> --- a/arch/x86/lib/insn-eval.c
> +++ b/arch/x86/lib/insn-eval.c
> @@ -555,7 +555,8 @@ static int get_reg_offset_16(struct insn *insn, struct pt_regs *regs,
>  }
>  
>  /**
> - * get_desc() - Obtain pointer to a segment descriptor
> + * get_desc() - Obtain contents of a segment descriptor
> + * @out:	Segment descriptor contents on success
>   * @sel:	Segment selector
>   *
>   * Given a segment selector, obtain a pointer to the segment descriptor.
> @@ -563,18 +564,18 @@ static int get_reg_offset_16(struct insn *insn, struct pt_regs *regs,
>   *
>   * Returns:
>   *
> - * Pointer to segment descriptor on success.
> + * True on success, false on failure.
>   *
>   * NULL on error.
>   */
> -static struct desc_struct *get_desc(unsigned short sel)
> +static bool get_desc(struct desc_struct *out, unsigned short sel)
>  {
>  	struct desc_ptr gdt_desc = {0, 0};
>  	unsigned long desc_base;
>  
>  #ifdef CONFIG_MODIFY_LDT_SYSCALL
>  	if ((sel & SEGMENT_TI_MASK) == SEGMENT_LDT) {
> -		struct desc_struct *desc = NULL;
> +		bool success = false;
>  		struct ldt_struct *ldt;
>  
>  		/* Bits [15:3] contain the index of the desired entry. */
> @@ -582,12 +583,14 @@ static struct desc_struct *get_desc(unsigned short sel)
>  
>  		mutex_lock(&current->active_mm->context.lock);
>  		ldt = current->active_mm->context.ldt;
> -		if (ldt && sel < ldt->nr_entries)
> -			desc = &ldt->entries[sel];
> +		if (ldt && sel < ldt->nr_entries) {
> +			*out = ldt->entries[sel];
> +			success = true;
> +		}
>  
>  		mutex_unlock(&current->active_mm->context.lock);
>  
> -		return desc;
> +		return success;
>  	}
>  #endif
>  	native_store_gdt(&gdt_desc);
> @@ -602,9 +605,10 @@ static struct desc_struct *get_desc(unsigned short sel)
>  	desc_base = sel & ~(SEGMENT_RPL_MASK | SEGMENT_TI_MASK);
>  
>  	if (desc_base > gdt_desc.size)
> -		return NULL;
> +		return false;
>  
> -	return (struct desc_struct *)(gdt_desc.address + desc_base);
> +	*out = *(struct desc_struct *)(gdt_desc.address + desc_base);
> +	return true;
>  }
>  
>  /**
> @@ -626,7 +630,7 @@ static struct desc_struct *get_desc(unsigned short sel)
>   */
>  unsigned long insn_get_seg_base(struct pt_regs *regs, int seg_reg_idx)
>  {
> -	struct desc_struct *desc;
> +	struct desc_struct desc;
>  	short sel;
>  
>  	sel = get_segment_selector(regs, seg_reg_idx);
> @@ -664,11 +668,10 @@ unsigned long insn_get_seg_base(struct pt_regs *regs, int seg_reg_idx)
>  	if (!sel)
>  		return -1L;
>  
> -	desc = get_desc(sel);
> -	if (!desc)
> +	if (!get_desc(&desc, sel))
>  		return -1L;
>  
> -	return get_desc_base(desc);
> +	return get_desc_base(&desc);
>  }
>  
>  /**
> @@ -690,7 +693,7 @@ unsigned long insn_get_seg_base(struct pt_regs *regs, int seg_reg_idx)
>   */
>  static unsigned long get_seg_limit(struct pt_regs *regs, int seg_reg_idx)
>  {
> -	struct desc_struct *desc;
> +	struct desc_struct desc;
>  	unsigned long limit;
>  	short sel;
>  
> @@ -704,8 +707,7 @@ static unsigned long get_seg_limit(struct pt_regs *regs, int seg_reg_idx)
>  	if (!sel)
>  		return 0;
>  
> -	desc = get_desc(sel);
> -	if (!desc)
> +	if (!get_desc(&desc, sel))
>  		return 0;
>  
>  	/*
> @@ -714,8 +716,8 @@ static unsigned long get_seg_limit(struct pt_regs *regs, int seg_reg_idx)
>  	 * not tested when checking the segment limits. In practice,
>  	 * this means that the segment ends in (limit << 12) + 0xfff.
>  	 */
> -	limit = get_desc_limit(desc);
> -	if (desc->g)
> +	limit = get_desc_limit(&desc);
> +	if (desc.g)
>  		limit = (limit << 12) + 0xfff;
>  
>  	return limit;
> @@ -739,7 +741,7 @@ static unsigned long get_seg_limit(struct pt_regs *regs, int seg_reg_idx)
>   */
>  int insn_get_code_seg_params(struct pt_regs *regs)
>  {
> -	struct desc_struct *desc;
> +	struct desc_struct desc;
>  	short sel;
>  
>  	if (v8086_mode(regs))
> @@ -750,8 +752,7 @@ int insn_get_code_seg_params(struct pt_regs *regs)
>  	if (sel < 0)
>  		return sel;
>  
> -	desc = get_desc(sel);
> -	if (!desc)
> +	if (!get_desc(&desc, sel))
>  		return -EINVAL;
>  
>  	/*
> @@ -759,10 +760,10 @@ int insn_get_code_seg_params(struct pt_regs *regs)
>  	 * determines whether a segment contains data or code. If this is a data
>  	 * segment, return error.
>  	 */
> -	if (!(desc->type & BIT(3)))
> +	if (!(desc.type & BIT(3)))
>  		return -EINVAL;
>  
> -	switch ((desc->l << 1) | desc->d) {
> +	switch ((desc.l << 1) | desc.d) {
>  	case 0: /*
>  		 * Legacy mode. CS.L=0, CS.D=0. Address and operand size are
>  		 * both 16-bit.
> 


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.ubuntu.com/archives/kernel-team/attachments/20190718/942f9947/attachment.sig>


More information about the kernel-team mailing list