From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org,
Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>,
Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>,
Dave Martin <Dave.Martin@arm.com>,
Julien Grall <julien.grall@arm.com>,
Will Deacon <will@kernel.org>
Subject: [PATCH 4.19 73/81] arm64/sve: Fix wrong free for task->thread.sve_state
Date: Wed, 16 Oct 2019 14:51:24 -0700 [thread overview]
Message-ID: <20191016214847.124955413@linuxfoundation.org> (raw)
In-Reply-To: <20191016214805.727399379@linuxfoundation.org>
From: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>
commit 4585fc59c0e813188d6a4c5de1f6976fce461fc2 upstream.
The system which has SVE feature crashed because of
the memory pointed by task->thread.sve_state was destroyed
by someone.
That is because sve_state is freed while the forking the
child process. The child process has the pointer of sve_state
which is same as the parent's because the child's task_struct
is copied from the parent's one. If the copy_process()
fails as an error on somewhere, for example, copy_creds(),
then the sve_state is freed even if the parent is alive.
The flow is as follows.
copy_process
p = dup_task_struct
=> arch_dup_task_struct
*dst = *src; // copy the entire region.
:
retval = copy_creds
if (retval < 0)
goto bad_fork_free;
:
bad_fork_free:
...
delayed_free_task(p);
=> free_task
=> arch_release_task_struct
=> fpsimd_release_task
=> __sve_free
=> kfree(task->thread.sve_state);
// free the parent's sve_state
Move child's sve_state = NULL and clearing TIF_SVE flag
to arch_dup_task_struct() so that the child doesn't free the
parent's one.
There is no need to wait until copy_process() to clear TIF_SVE for
dst, because the thread flags for dst are initialized already by
copying the src task_struct.
This change simplifies the code, so get rid of comments that are no
longer needed.
As a note, arm64 used to have thread_info on the stack. So it
would not be possible to clear TIF_SVE until the stack is initialized.
>From commit c02433dd6de3 ("arm64: split thread_info from task stack"),
the thread_info is part of the task, so it should be valid to modify
the flag from arch_dup_task_struct().
Cc: stable@vger.kernel.org # 4.15.x-
Fixes: bc0ee4760364 ("arm64/sve: Core task context handling")
Signed-off-by: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>
Reported-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Suggested-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Tested-by: Julien Grall <julien.grall@arm.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
arch/arm64/kernel/process.c | 32 +++++++++++++++-----------------
1 file changed, 15 insertions(+), 17 deletions(-)
--- a/arch/arm64/kernel/process.c
+++ b/arch/arm64/kernel/process.c
@@ -285,22 +285,27 @@ void arch_release_task_struct(struct tas
fpsimd_release_task(tsk);
}
-/*
- * src and dst may temporarily have aliased sve_state after task_struct
- * is copied. We cannot fix this properly here, because src may have
- * live SVE state and dst's thread_info may not exist yet, so tweaking
- * either src's or dst's TIF_SVE is not safe.
- *
- * The unaliasing is done in copy_thread() instead. This works because
- * dst is not schedulable or traceable until both of these functions
- * have been called.
- */
int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
{
if (current->mm)
fpsimd_preserve_current_state();
*dst = *src;
+ /* We rely on the above assignment to initialize dst's thread_flags: */
+ BUILD_BUG_ON(!IS_ENABLED(CONFIG_THREAD_INFO_IN_TASK));
+
+ /*
+ * Detach src's sve_state (if any) from dst so that it does not
+ * get erroneously used or freed prematurely. dst's sve_state
+ * will be allocated on demand later on if dst uses SVE.
+ * For consistency, also clear TIF_SVE here: this could be done
+ * later in copy_process(), but to avoid tripping up future
+ * maintainers it is best not to leave TIF_SVE and sve_state in
+ * an inconsistent state, even temporarily.
+ */
+ dst->thread.sve_state = NULL;
+ clear_tsk_thread_flag(dst, TIF_SVE);
+
return 0;
}
@@ -314,13 +319,6 @@ int copy_thread(unsigned long clone_flag
memset(&p->thread.cpu_context, 0, sizeof(struct cpu_context));
/*
- * Unalias p->thread.sve_state (if any) from the parent task
- * and disable discard SVE state for p:
- */
- clear_tsk_thread_flag(p, TIF_SVE);
- p->thread.sve_state = NULL;
-
- /*
* In case p was allocated the same task_struct pointer as some
* other recently-exited task, make sure p is disassociated from
* any cpu that may have run that now-exited task recently.
next prev parent reply other threads:[~2019-10-16 22:08 UTC|newest]
Thread overview: 104+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-16 21:50 [PATCH 4.19 00/81] 4.19.80-stable review Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 01/81] panic: ensure preemption is disabled during panic() Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 02/81] f2fs: use EINVAL for superblock with invalid magic Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 03/81] USB: rio500: Remove Rio 500 kernel driver Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 04/81] USB: yurex: Dont retry on unexpected errors Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 05/81] USB: yurex: fix NULL-derefs on disconnect Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 06/81] USB: usb-skeleton: fix runtime PM after driver unbind Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 07/81] USB: usb-skeleton: fix NULL-deref on disconnect Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 08/81] xhci: Fix false warning message about wrong bounce buffer write length Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 09/81] xhci: Prevent device initiated U1/U2 link pm if exit latency is too long Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 10/81] xhci: Check all endpoints for LPM timeout Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 11/81] xhci: Fix USB 3.1 capability detection on early xHCI 1.1 spec based hosts Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 12/81] usb: xhci: wait for CNR controller not ready bit in xhci resume Greg Kroah-Hartman
2019-10-18 17:28 ` Pavel Machek
2019-10-16 21:50 ` [PATCH 4.19 13/81] xhci: Prevent deadlock when xhci adapter breaks during init Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 14/81] xhci: Increase STS_SAVE timeout in xhci_suspend() Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 15/81] USB: adutux: fix use-after-free on disconnect Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 16/81] USB: adutux: fix NULL-derefs " Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 17/81] USB: adutux: fix use-after-free on release Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 18/81] USB: iowarrior: fix use-after-free on disconnect Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 19/81] USB: iowarrior: fix use-after-free on release Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 20/81] USB: iowarrior: fix use-after-free after driver unbind Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 21/81] USB: usblp: fix runtime PM " Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 22/81] USB: chaoskey: fix use-after-free on release Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 23/81] USB: ldusb: fix NULL-derefs on driver unbind Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 24/81] serial: uartlite: fix exit path null pointer Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 25/81] USB: serial: keyspan: fix NULL-derefs on open() and write() Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 26/81] USB: serial: ftdi_sio: add device IDs for Sienna and Echelon PL-20 Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 27/81] USB: serial: option: add Telit FN980 compositions Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 28/81] USB: serial: option: add support for Cinterion CLS8 devices Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 29/81] USB: serial: fix runtime PM after driver unbind Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 30/81] USB: usblcd: fix I/O after disconnect Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 31/81] USB: microtek: fix info-leak at probe Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 32/81] USB: dummy-hcd: fix power budget for SuperSpeed mode Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 33/81] usb: renesas_usbhs: gadget: Do not discard queues in usb_ep_set_{halt,wedge}() Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 34/81] usb: renesas_usbhs: gadget: Fix usb_ep_set_{halt,wedge}() behavior Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 35/81] USB: legousbtower: fix slab info leak at probe Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 36/81] USB: legousbtower: fix deadlock on disconnect Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 37/81] USB: legousbtower: fix potential NULL-deref " Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 38/81] USB: legousbtower: fix open after failed reset request Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 39/81] USB: legousbtower: fix use-after-free on release Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 40/81] mei: me: add comet point (lake) LP device ids Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 41/81] mei: avoid FW version request on Ibex Peak and earlier Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 42/81] gpio: eic: sprd: Fix the incorrect EIC offset when toggling Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 43/81] Staging: fbtft: fix memory leak in fbtft_framebuffer_alloc Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 44/81] staging: vt6655: Fix memory leak in vt6655_probe Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 45/81] iio: adc: hx711: fix bug in sampling of data Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 46/81] iio: adc: ad799x: fix probe error handling Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 47/81] iio: adc: axp288: Override TS pin bias current for some models Greg Kroah-Hartman
2019-10-16 21:50 ` [PATCH 4.19 48/81] iio: light: opt3001: fix mutex unlock race Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 49/81] efivar/ssdt: Dont iterate over EFI vars if no SSDT override was specified Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 50/81] perf llvm: Dont access out-of-scope array Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 51/81] perf inject jit: Fix JIT_CODE_MOVE filename Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 52/81] blk-wbt: fix performance regression in wbt scale_up/scale_down Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 53/81] CIFS: Gracefully handle QueryInfo errors during open Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 54/81] CIFS: Force revalidate inode when dentry is stale Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 55/81] CIFS: Force reval dentry if LOOKUP_REVAL flag is set Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 56/81] kernel/sysctl.c: do not override max_threads provided by userspace Greg Kroah-Hartman
2019-10-17 10:59 ` Pavel Machek
2019-10-17 11:05 ` Michal Hocko
2019-10-17 11:25 ` David Laight
2019-10-17 11:39 ` Michal Hocko
2019-11-18 15:25 ` Pavel Machek
2019-11-18 15:52 ` Michal Hocko
2019-10-16 21:51 ` [PATCH 4.19 57/81] mm/vmpressure.c: fix a signedness bug in vmpressure_register_event() Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 58/81] firmware: google: increment VPD key_len properly Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 59/81] gpiolib: dont clear FLAG_IS_OUT when emulating open-drain/open-source Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 60/81] iio: adc: stm32-adc: move registers definitions Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 61/81] iio: adc: stm32-adc: fix a race when using several adcs with dma and irq Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 62/81] cifs: use cifsInodeInfo->open_file_lock while iterating to avoid a panic Greg Kroah-Hartman
2019-10-17 8:55 ` Pavel Machek
2019-10-17 16:01 ` Greg Kroah-Hartman
2019-10-17 17:19 ` Sasha Levin
2019-10-16 21:51 ` [PATCH 4.19 63/81] btrfs: fix incorrect updating of log root tree Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 64/81] btrfs: fix uninitialized ret in ref-verify Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 65/81] NFS: Fix O_DIRECT accounting of number of bytes read/written Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 66/81] MIPS: Disable Loongson MMI instructions for kernel build Greg Kroah-Hartman
2020-08-26 21:06 ` Guenter Roeck
2020-09-03 9:26 ` Greg Kroah-Hartman
2020-09-07 3:35 ` Philippe Mathieu-Daudé
2020-09-24 13:54 ` Thomas Bogendoerfer
2019-10-16 21:51 ` [PATCH 4.19 67/81] MIPS: elf_hwcap: Export userspace ASEs Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 68/81] ACPICA: ACPI 6.3: PPTT add additional fields in Processor Structure Flags Greg Kroah-Hartman
2019-10-17 8:59 ` Pavel Machek
2019-10-17 15:59 ` Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 69/81] ACPI/PPTT: Add support for ACPI 6.3 thread flag Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 70/81] arm64: topology: Use PPTT to determine if PE is a thread Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 71/81] Fix the locking in dcache_readdir() and friends Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 72/81] media: stkwebcam: fix runtime PM after driver unbind Greg Kroah-Hartman
2019-10-16 21:51 ` Greg Kroah-Hartman [this message]
2019-10-16 21:51 ` [PATCH 4.19 74/81] tracing/hwlat: Report total time spent in all NMIs during the sample Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 75/81] tracing/hwlat: Dont ignore outer-loop duration when calculating max_latency Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 76/81] ftrace: Get a reference counter for the trace_array on filter files Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 77/81] tracing: Get trace_array reference for available_tracers files Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 78/81] hwmon: Fix HWMON_P_MIN_ALARM mask Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 79/81] x86/asm: Fix MWAITX C-state hint value Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 80/81] PCI: vmd: Fix config addressing when using bus offsets Greg Kroah-Hartman
2019-10-16 21:51 ` [PATCH 4.19 81/81] perf/hw_breakpoint: Fix arch_hw_breakpoint use-before-initialization Greg Kroah-Hartman
2019-10-17 4:42 ` [PATCH 4.19 00/81] 4.19.80-stable review kernelci.org bot
2019-10-17 13:50 ` Naresh Kamboju
2019-10-17 15:02 ` shuah
2019-10-17 18:04 ` Guenter Roeck
2019-10-17 18:35 ` Didik Setiawan
2019-10-18 8:00 ` Jon Hunter
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=20191016214847.124955413@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=Dave.Martin@arm.com \
--cc=julien.grall@arm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=m.mizuma@jp.fujitsu.com \
--cc=seto.hidetoshi@jp.fujitsu.com \
--cc=stable@vger.kernel.org \
--cc=will@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
Powered by JetHome