From: Sean Christopherson <sean.j.christopherson@intel.com>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
x86@kernel.org, Andy Lutomirski <luto@kernel.org>,
Vincenzo Frascino <vincenzo.frascino@arm.com>,
Kees Cook <keescook@chromium.org>,
Paul Bolle <pebolle@tiscali.nl>, Will Deacon <will@kernel.org>
Subject: Re: [patch 3/5] lib/vdso/32: Provide legacy syscall fallbacks
Date: Mon, 29 Jul 2019 07:48:31 -0700 [thread overview]
Message-ID: <20190729144831.GA21120@linux.intel.com> (raw)
In-Reply-To: <20190728131648.786513965@linutronix.de>
On Sun, Jul 28, 2019 at 03:12:54PM +0200, Thomas Gleixner wrote:
> To address the regression which causes seccomp to deny applications the
> access to clock_gettime64() and clock_getres64() syscalls because they
> are not enabled in the existing filters.
>
> That trips over the fact that 32bit VDSOs use the new clock_gettime64() and
> clock_getres64() syscalls in the fallback path.
>
> Implement a __cvdso_clock_get*time32() variants which invokes the legacy
> 32bit syscalls when the architecture requests it.
>
> The conditional can go away once all architectures are converted.
>
> Fixes: 00b26474c2f1 ("lib/vdso: Provide generic VDSO implementation")
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> ---
> lib/vdso/gettimeofday.c | 46 +++++++++++++++++++++++++++++++++++++++++++++-
> 1 file changed, 45 insertions(+), 1 deletion(-)
>
> --- a/lib/vdso/gettimeofday.c
> +++ b/lib/vdso/gettimeofday.c
> @@ -117,6 +117,8 @@ static __maybe_unused int
> return 0;
> }
>
> +#ifndef VDSO_HAS_32BIT_FALLBACK
> +
> static __maybe_unused int
> __cvdso_clock_gettime32(clockid_t clock, struct old_timespec32 *res)
> {
> @@ -132,10 +134,29 @@ static __maybe_unused int
> res->tv_sec = ts.tv_sec;
> res->tv_nsec = ts.tv_nsec;
> }
> -
> return ret;
> }
>
> +#else
> +
> +static __maybe_unused int
> +__cvdso_clock_gettime32(clockid_t clock, struct old_timespec32 *res)
> +{
> + struct __kernel_timespec ts;
> + int ret;
> +
> + ret = __cvdso_clock_gettime_common(clock, &ts);
> +
> + if (likely(!ret)) {
> + res->tv_sec = ts.tv_sec;
> + res->tv_nsec = ts.tv_nsec;
> + return 0;
> + }
> + return clock_gettime32_fallback(clock, res);
> +}
> +
> +#endif
> +
> static __maybe_unused int
> __cvdso_gettimeofday(struct __kernel_old_timeval *tv, struct timezone *tz)
> {
> @@ -225,6 +246,8 @@ int __cvdso_clock_getres(clockid_t clock
> return 0;
> }
>
> +#ifndef VDSO_HAS_32BIT_FALLBACK
> +
> static __maybe_unused int
> __cvdso_clock_getres_time32(clockid_t clock, struct old_timespec32 *res)
> {
> @@ -241,4 +264,25 @@ static __maybe_unused int
> }
> return ret;
> }
> +
> +#else
> +
> +static __maybe_unused int
> +__cvdso_clock_getres_time32(clockid_t clock, struct old_timespec32 *res)
> +{
> + struct __kernel_timespec ts;
> + int ret;
> +
> + ret = __cvdso_clock_getres_common(clock, &ts);
> +
> + if (likely(!ret)) {
> + res->tv_sec = ts.tv_sec;
> + res->tv_nsec = ts.tv_nsec;
> + return 0;
> + }
> +
> + return clock_getres32_fallback(clock, res);
> +}
> +#endif
> +
> #endif /* VDSO_HAS_CLOCK_GETRES */
Any reason not to have the #ifndef apply only to the fallback? Wrapping
the entire function and flipping the order of handling 'ret' makes it a bit
difficult to discern that the only difference is the fallback invocation.
static __maybe_unused int
__cvdso_clock_gettime32(clockid_t clock, struct old_timespec32 *res)
{
struct __kernel_timespec ts;
int ret;
ret = __cvdso_clock_gettime_common(clock, &ts);
if (unlikely(ret))
#ifndef VDSO_HAS_32BIT_FALLBACK
ret = clock_gettime_fallback(clock, &ts);
#else
return clock_gettime32_fallback(clock, res);
#endif
if (likely(!ret)) {
res->tv_sec = ts.tv_sec;
res->tv_nsec = ts.tv_nsec;
}
return ret;
}
static __maybe_unused int
__cvdso_clock_getres_time32(clockid_t clock, struct old_timespec32 *res)
{
struct __kernel_timespec ts;
int ret;
ret = __cvdso_clock_getres_common(clock, &ts);
if (unlikely(ret))
#ifndef VDSO_HAS_32BIT_FALLBACK
ret = clock_getres_fallback(clock, &ts);
#else
return clock_getres32_fallback(clock, res);
#endif
if (likely(!ret)) {
res->tv_sec = ts.tv_sec;
res->tv_nsec = ts.tv_nsec;
}
return ret;
}
next prev parent reply other threads:[~2019-07-29 14:48 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-28 13:12 [patch 0/5] lib/vdso, x86/vdso: Fix fallout from generic VDSO conversion Thomas Gleixner
2019-07-28 13:12 ` [patch 1/5] lib/vdso/32: Remove inconsistent NULL pointer checks Thomas Gleixner
2019-07-28 14:33 ` Andy Lutomirski
2019-07-28 14:42 ` Thomas Gleixner
2019-07-28 15:30 ` Vincenzo Frascino
2019-07-29 14:52 ` Sean Christopherson
2019-07-30 22:19 ` [tip:timers/urgent] " tip-bot for Thomas Gleixner
2019-07-28 13:12 ` [patch 2/5] lib/vdso: Move fallback invocation to the callers Thomas Gleixner
2019-07-28 14:37 ` Andy Lutomirski
2019-07-28 15:31 ` Vincenzo Frascino
2019-07-30 22:19 ` [tip:timers/urgent] " tip-bot for Thomas Gleixner
2019-07-28 13:12 ` [patch 3/5] lib/vdso/32: Provide legacy syscall fallbacks Thomas Gleixner
2019-07-28 14:39 ` Andy Lutomirski
2019-07-28 15:33 ` Vincenzo Frascino
2019-07-29 14:48 ` Sean Christopherson [this message]
2019-07-30 9:38 ` [patch V2 " Thomas Gleixner
2019-07-30 14:28 ` Sean Christopherson
2019-07-30 20:09 ` Andy Lutomirski
2019-07-30 20:14 ` Thomas Gleixner
2019-07-30 22:20 ` [tip:timers/urgent] " tip-bot for Thomas Gleixner
2019-07-28 13:12 ` [patch 4/5] x86/vdso/32: Use 32bit syscall fallback Thomas Gleixner
2019-07-28 14:50 ` Andy Lutomirski
2019-07-29 11:23 ` Vincenzo Frascino
2019-07-29 14:51 ` Sean Christopherson
2019-07-30 22:21 ` [tip:timers/urgent] " tip-bot for Thomas Gleixner
2019-07-28 13:12 ` [patch 5/5] arm64: compat: vdso: Use legacy syscalls as fallback Thomas Gleixner
2019-07-28 15:35 ` Vincenzo Frascino
2019-07-30 22:22 ` [tip:timers/urgent] " tip-bot for Thomas Gleixner
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=20190729144831.GA21120@linux.intel.com \
--to=sean.j.christopherson@intel.com \
--cc=keescook@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@kernel.org \
--cc=pebolle@tiscali.nl \
--cc=tglx@linutronix.de \
--cc=vincenzo.frascino@arm.com \
--cc=will@kernel.org \
--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