From: Oleg Nesterov <oleg@redhat.com>
To: Thomas Gleixner <tglx@linutronix.de>,
Steven Rostedt <rostedt@goodmis.org>
Cc: "Gustavo A. R. Silva" <gustavo@embeddedor.com>,
Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
"H. Peter Anvin" <hpa@zytor.com>,
x86@kernel.org, LKML <linux-kernel@vger.kernel.org>,
Dominik Brodowski <linux@dominikbrodowski.net>,
Andy Lutomirski <luto@kernel.org>,
Kees Cook <keescook@chromium.org>,
"Eric W. Biederman" <ebiederm@xmission.com>
Subject: Re: [PATCH v2] x86/syscalls: Mark expected switch fall-throughs
Date: Tue, 26 Mar 2019 16:12:45 +0100 [thread overview]
Message-ID: <20190326151244.GC16837@redhat.com> (raw)
In-Reply-To: <alpine.DEB.2.21.1903231457590.1798@nanos.tec.linutronix.de>
On 03/23, Thomas Gleixner wrote:
>
> On Thu, 28 Feb 2019, Gustavo A. R. Silva wrote:
>
> > arch/x86/include/asm/syscall.h | 28 ++++++++++++++++++++++++++++
> > 1 file changed, 28 insertions(+)
>
> Second thoughts. So this adds 28 /* fall through */ comments. Now I
> appreciate the effort, but can we pretty please look at the code in
> question and figure out whether the implementation makes sense in the first
> place before adding falltrough comments blindly?
>
> The whole exercise can be simplified. Untested patch below.
>
> Looking at that stuff makes me wonder about two things:
>
> 1) The third argument of get/set(), i.e. the argument offset, is 0 on all
> call sites. Do we need it at all?
Probably "maxargs" can be removed too, Steven sent the patches a long ago, see
https://lore.kernel.org/lkml/20161107212634.529267342@goodmis.org/
> 2) syscall_set_arguments() has been introduced in 2008 and we still have
> no caller. Instead of polishing it, can it be removed completely or are
> there plans to actually use it?
I think it can die.
>
> Thanks,
>
> tglx
>
> 8<----------------
>
> arch/x86/include/asm/syscall.h | 174 +++++++++++++++--------------------------
> 1 file changed, 64 insertions(+), 110 deletions(-)
>
> --- a/arch/x86/include/asm/syscall.h
> +++ b/arch/x86/include/asm/syscall.h
> @@ -114,126 +114,80 @@ static inline int syscall_get_arch(void)
>
> #else /* CONFIG_X86_64 */
>
> +static inline unsigned long syscall_get_argreg(struct pt_regs *regs,
> + unsigned int idx)
> +{
> + switch (idx) {
> + case 0: return regs->di;
> + case 1: return regs->si;
> + case 2: return regs->dx;
> + case 3: return regs->r10;
> + case 4: return regs->r8;
> + case 5: return regs->r9;
> +#ifdef CONFIG_IA32_EMULATION
> + case 6: return regs->bx;
> + case 7: return regs->cx;
> + case 8: return regs->dx;
> + case 9: return regs->si;
> + case 10: return regs->di;
> + case 11: return regs->bp;
> +#endif
> + }
> + return 0;
> +}
> +
> static inline void syscall_get_arguments(struct task_struct *task,
> struct pt_regs *regs,
> - unsigned int i, unsigned int n,
> + unsigned int idx, unsigned int cnt,
> unsigned long *args)
> {
> -# ifdef CONFIG_IA32_EMULATION
> - if (task->thread_info.status & TS_COMPAT)
> - switch (i) {
> - case 0:
> - if (!n--) break;
> - *args++ = regs->bx;
> - case 1:
> - if (!n--) break;
> - *args++ = regs->cx;
> - case 2:
> - if (!n--) break;
> - *args++ = regs->dx;
> - case 3:
> - if (!n--) break;
> - *args++ = regs->si;
> - case 4:
> - if (!n--) break;
> - *args++ = regs->di;
> - case 5:
> - if (!n--) break;
> - *args++ = regs->bp;
> - case 6:
> - if (!n--) break;
> - default:
> - BUG();
> - break;
> - }
> - else
> -# endif
> - switch (i) {
> - case 0:
> - if (!n--) break;
> - *args++ = regs->di;
> - case 1:
> - if (!n--) break;
> - *args++ = regs->si;
> - case 2:
> - if (!n--) break;
> - *args++ = regs->dx;
> - case 3:
> - if (!n--) break;
> - *args++ = regs->r10;
> - case 4:
> - if (!n--) break;
> - *args++ = regs->r8;
> - case 5:
> - if (!n--) break;
> - *args++ = regs->r9;
> - case 6:
> - if (!n--) break;
> - default:
> - BUG();
> - break;
> - }
> + if (WARN_ON((idx + cnt) > 6))
> + return;
> +
> + if (IS_ENABLED(CONFIG_IA32_EMULATION) &&
> + task->thread_info.status & TS_COMPAT)
> + idx += 6;
> +
> + for (; cnt > 0; cnt--)
> + *args++ = syscall_get_argreg(regs, idx++);
> +}
> +
> +static inline void syscall_set_argreg(struct pt_regs *regs,
> + unsigned int idx,
> + unsigned long val)
> +{
> + switch (idx) {
> + case 0: regs->di = val; break;
> + case 1: regs->si = val; break;
> + case 2: regs->dx = val; break;
> + case 3: regs->r10 = val; break;
> + case 4: regs->r8 = val; break;
> + case 5: regs->r9 = val; break;
> +#ifdef CONFIG_IA32_EMULATION
> + case 6: regs->bx = val; break;
> + case 7: regs->cx = val; break;
> + case 8: regs->dx = val; break;
> + case 9: regs->si = val; break;
> + case 10: regs->di = val; break;
> + case 11: regs->bp = val; break;
> +#endif
> + }
> }
>
> static inline void syscall_set_arguments(struct task_struct *task,
> struct pt_regs *regs,
> - unsigned int i, unsigned int n,
> + unsigned int idx, unsigned int cnt,
> const unsigned long *args)
> {
> -# ifdef CONFIG_IA32_EMULATION
> - if (task->thread_info.status & TS_COMPAT)
> - switch (i) {
> - case 0:
> - if (!n--) break;
> - regs->bx = *args++;
> - case 1:
> - if (!n--) break;
> - regs->cx = *args++;
> - case 2:
> - if (!n--) break;
> - regs->dx = *args++;
> - case 3:
> - if (!n--) break;
> - regs->si = *args++;
> - case 4:
> - if (!n--) break;
> - regs->di = *args++;
> - case 5:
> - if (!n--) break;
> - regs->bp = *args++;
> - case 6:
> - if (!n--) break;
> - default:
> - BUG();
> - break;
> - }
> - else
> -# endif
> - switch (i) {
> - case 0:
> - if (!n--) break;
> - regs->di = *args++;
> - case 1:
> - if (!n--) break;
> - regs->si = *args++;
> - case 2:
> - if (!n--) break;
> - regs->dx = *args++;
> - case 3:
> - if (!n--) break;
> - regs->r10 = *args++;
> - case 4:
> - if (!n--) break;
> - regs->r8 = *args++;
> - case 5:
> - if (!n--) break;
> - regs->r9 = *args++;
> - case 6:
> - if (!n--) break;
> - default:
> - BUG();
> - break;
> - }
> + if (WARN_ON((idx + cnt) > 6))
> + return;
> +
> + if (IS_ENABLED(CONFIG_IA32_EMULATION) &&
> + task->thread_info.status & TS_COMPAT)
> + idx += 6;
> +
> + for (; cnt > 0; cnt--)
> + syscall_set_argreg(regs, idx++, *args++);
> }
>
> static inline int syscall_get_arch(void)
next prev parent reply other threads:[~2019-03-26 15:13 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-02-28 19:27 [PATCH v2] x86/syscalls: Mark expected switch fall-throughs Gustavo A. R. Silva
2019-03-23 16:14 ` Thomas Gleixner
2019-03-26 15:12 ` Oleg Nesterov [this message]
2019-03-26 16:09 ` Thomas Gleixner
2019-03-26 16:16 ` Steven Rostedt
2019-03-26 16:17 ` Thomas Gleixner
2019-03-27 1:26 ` Dmitry V. Levin
2019-03-27 14:29 ` Thomas Gleixner
2019-03-27 22:20 ` Dmitry V. Levin
2019-03-27 22:52 ` Thomas Gleixner
2019-03-27 23:03 ` Steven Rostedt
2019-03-27 23:12 ` Dmitry V. Levin
2019-03-27 23:15 ` Steven Rostedt
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=20190326151244.GC16837@redhat.com \
--to=oleg@redhat.com \
--cc=bp@alien8.de \
--cc=ebiederm@xmission.com \
--cc=gustavo@embeddedor.com \
--cc=hpa@zytor.com \
--cc=keescook@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@dominikbrodowski.net \
--cc=luto@kernel.org \
--cc=mingo@redhat.com \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=x86@kernel.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).