From: will.deacon@arm.com (Will Deacon)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH] arm64: mm: unaligned access by user-land should be received as SIGBUS
Date: Mon, 3 Apr 2017 10:24:31 +0100 [thread overview]
Message-ID: <20170403092430.GA5706@arm.com> (raw)
In-Reply-To: <1491198314-17025-2-git-send-email-kamensky@cisco.com>
Hi Victor,
Thanks for reporting this.
On Sun, Apr 02, 2017 at 10:45:14PM -0700, Victor Kamensky wrote:
> After 52d7523 (arm64: mm: allow the kernel to handle alignment faults on
> user accesses) commit user-land accesses that produce unaligned exceptions
> like in case of aarch32 ldm/stm/ldrd/strd instructions operating on
> unaligned memory received by user-land as SIGSEGV. It is wrong, it should
> be reported as SIGBUS as it was before 52d7523 commit.
>
> Changed do_bad_area function to take signal and code parameters, so caller
> can pass them down properly depending on fault type, as SIGSEGV in case of
> do_translation_fault and SIGBUS in case of do_alignment_fault.
>
> Signed-off-by: Victor Kamensky <kamensky@cisco.com>
> Cc: xe-linux-external at cisco.com
> Fixes: 52d7523 (arm64: mm: allow the kernel to handle alignment faults on user accesses)
> ---
> arch/arm64/mm/fault.c | 9 +++++----
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
> index 4bf899f..204eb58 100644
> --- a/arch/arm64/mm/fault.c
> +++ b/arch/arm64/mm/fault.c
> @@ -215,7 +215,8 @@ static void __do_user_fault(struct task_struct *tsk, unsigned long addr,
> force_sig_info(sig, &si, tsk);
> }
>
> -static void do_bad_area(unsigned long addr, unsigned int esr, struct pt_regs *regs)
> +static void do_bad_area(unsigned long addr, unsigned int esr,
> + struct pt_regs *regs, int sig, int code)
> {
> struct task_struct *tsk = current;
> struct mm_struct *mm = tsk->active_mm;
> @@ -225,7 +226,7 @@ static void do_bad_area(unsigned long addr, unsigned int esr, struct pt_regs *re
> * handle this fault with.
> */
> if (user_mode(regs))
> - __do_user_fault(tsk, addr, esr, SIGSEGV, SEGV_MAPERR, regs);
> + __do_user_fault(tsk, addr, esr, sig, code, regs);
> else
> __do_kernel_fault(mm, addr, esr, regs);
> }
> @@ -469,14 +470,14 @@ static int __kprobes do_translation_fault(unsigned long addr,
> if (addr < TASK_SIZE)
> return do_page_fault(addr, esr, regs);
>
> - do_bad_area(addr, esr, regs);
> + do_bad_area(addr, esr, regs, SIGSEGV, SEGV_MAPERR);
> return 0;
> }
>
> static int do_alignment_fault(unsigned long addr, unsigned int esr,
> struct pt_regs *regs)
> {
> - do_bad_area(addr, esr, regs);
> + do_bad_area(addr, esr, regs, SIGBUS, BUS_ADRALN);
Can we not just extract the signal number and code from the fault info
table?
E.g. leave the type signature of do_bad_area like it is, but do:
const struct fault_info *inf = fault_info + (esr & 63);
__do_user_fault(tsk, addr, esr, inf->sig, inf->code, regs);
The '& 63' is ugly as hell, so maybe wrap that up in a esr_to_fault_info
function and kill the fault_name thing we have at the moment.
Will
next prev parent reply other threads:[~2017-04-03 9:24 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-04-03 5:45 [PATCH] fix armv8 kernel generation of SIGSEGV upon unaligned access Victor Kamensky
2017-04-03 5:45 ` [PATCH] arm64: mm: unaligned access by user-land should be received as SIGBUS Victor Kamensky
2017-04-03 9:24 ` Will Deacon [this message]
2017-04-06 6:01 ` Victor Kamensky
2017-04-06 8:44 ` Will Deacon
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170403092430.GA5706@arm.com \
--to=will.deacon@arm.com \
--cc=linux-arm-kernel@lists.infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).