| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Wed, 16 Nov 2016 10:42:05 -0800
From: Joel Fernandes <joelaf@...gle.com>
To: Arnd Bergmann <arnd@...db.de>
Cc: Anton Vorontsov <anton@...msg.org>,
Colin Cross <ccross@...roid.com>,
Kees Cook <keescook@...omium.org>,
Tony Luck <tony.luck@...el.com>,
Namhyung Kim <namhyung@...nel.org>,
Geliang Tang <geliangtang@...il.com>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] pstore: fix warning about conditional locking
Hi Arnd,
On Wed, Nov 16, 2016 at 7:15 AM, Arnd Bergmann <arnd@...db.de> wrote:
> gcc cannot prove that "prz->flags" remains constant over the
> function, so it warns that we might restore flags that have
> never been saved:
>
> fs/pstore/ram_core.c: In function ‘buffer_size_add’:
> include/linux/spinlock.h:246:3: error: ‘flags’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
>
> This reworks the code to avoid the spinlock entirely, by actually
> using the atomic_t for atomic updates. The resulting code is
> more efficient and easier to understand, and it avoids the
> local variable that the compiler did not track correctly.
Kees fixed this in his tree by initializing the flags variable.
atomic_cmpxchg is not an option as the memory may be uncached and
caused issues, it was previously this way and was replaced with
locking.
Thanks,
Joel
>
> Fixes: 95937ddce59a ("pstore: Allow prz to control need for locking")
> Signed-off-by: Arnd Bergmann <arnd@...db.de>
> ---
> fs/pstore/ram.c | 4 +---
> fs/pstore/ram_core.c | 46 ++++++++++++++++++----------------------------
> include/linux/pstore_ram.h | 8 --------
> 3 files changed, 19 insertions(+), 39 deletions(-)
>
> diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c
> index 380222432eff..cb66ba710349 100644
> --- a/fs/pstore/ram.c
> +++ b/fs/pstore/ram.c
> @@ -787,9 +787,7 @@ static int ramoops_probe(struct platform_device *pdev)
> : 1;
> err = ramoops_init_przs("ftrace", dev, cxt, &cxt->fprzs, &paddr,
> cxt->ftrace_size, -1,
> - &cxt->max_ftrace_cnt, LINUX_VERSION_CODE,
> - (cxt->flags & RAMOOPS_FLAG_FTRACE_PER_CPU)
> - ? PRZ_FLAG_NO_LOCK : 0);
> + &cxt->max_ftrace_cnt, LINUX_VERSION_CODE, 0);
> if (err)
> goto fail_init_fprz;
>
> diff --git a/fs/pstore/ram_core.c b/fs/pstore/ram_core.c
> index 0cc23cb18719..87be7bdfaf48 100644
> --- a/fs/pstore/ram_core.c
> +++ b/fs/pstore/ram_core.c
> @@ -51,21 +51,17 @@ static inline size_t buffer_start(struct persistent_ram_zone *prz)
> /* increase and wrap the start pointer, returning the old value */
> static size_t buffer_start_add(struct persistent_ram_zone *prz, size_t a)
> {
> - int old;
> - int new;
> - unsigned long flags;
> -
> - if (!(prz->flags & PRZ_FLAG_NO_LOCK))
> - raw_spin_lock_irqsave(&prz->buffer_lock, flags);
> + int old, new, before;
>
> old = atomic_read(&prz->buffer->start);
> - new = old + a;
> - while (unlikely(new >= prz->buffer_size))
> - new -= prz->buffer_size;
> - atomic_set(&prz->buffer->start, new);
> + do {
> + new = old + a;
> + while (unlikely(new >= prz->buffer_size))
> + new -= prz->buffer_size;
>
> - if (!(prz->flags & PRZ_FLAG_NO_LOCK))
> - raw_spin_unlock_irqrestore(&prz->buffer_lock, flags);
> + before = old;
> + old = atomic_cmpxchg(&prz->buffer->start, before, new);
> + } while (old != before);
>
> return old;
> }
> @@ -73,25 +69,20 @@ static size_t buffer_start_add(struct persistent_ram_zone *prz, size_t a)
> /* increase the size counter until it hits the max size */
> static void buffer_size_add(struct persistent_ram_zone *prz, size_t a)
> {
> - size_t old;
> - size_t new;
> - unsigned long flags;
> -
> - if (!(prz->flags & PRZ_FLAG_NO_LOCK))
> - raw_spin_lock_irqsave(&prz->buffer_lock, flags);
> + size_t old, new, before;
>
> old = atomic_read(&prz->buffer->size);
> - if (old == prz->buffer_size)
> - goto exit;
> + do {
> + if (old == prz->buffer_size)
> + break;
>
> - new = old + a;
> - if (new > prz->buffer_size)
> - new = prz->buffer_size;
> - atomic_set(&prz->buffer->size, new);
> + new = old + a;
> + if (new > prz->buffer_size)
> + new = prz->buffer_size;
>
> -exit:
> - if (!(prz->flags & PRZ_FLAG_NO_LOCK))
> - raw_spin_unlock_irqrestore(&prz->buffer_lock, flags);
> + before = old;
> + old = atomic_cmpxchg(&prz->buffer->size, before, new);
> + } while (old != before);
> }
>
> static void notrace persistent_ram_encode_rs8(struct persistent_ram_zone *prz,
> @@ -496,7 +487,6 @@ static int persistent_ram_post_init(struct persistent_ram_zone *prz, u32 sig,
>
> prz->buffer->sig = sig;
> persistent_ram_zap(prz);
> - prz->buffer_lock = __RAW_SPIN_LOCK_UNLOCKED(buffer_lock);
> prz->flags = flags;
>
> return 0;
> diff --git a/include/linux/pstore_ram.h b/include/linux/pstore_ram.h
> index 9395f06e8372..e755803ba58a 100644
> --- a/include/linux/pstore_ram.h
> +++ b/include/linux/pstore_ram.h
> @@ -24,13 +24,6 @@
> #include <linux/list.h>
> #include <linux/types.h>
>
> -/*
> - * Choose whether access to the RAM zone requires locking or not. If a zone
> - * can be written to from different CPUs like with ftrace for example, then
> - * PRZ_FLAG_NO_LOCK is used. For all other cases, locking is required.
> - */
> -#define PRZ_FLAG_NO_LOCK BIT(0)
> -
> struct persistent_ram_buffer;
> struct rs_control;
>
> @@ -48,7 +41,6 @@ struct persistent_ram_zone {
> struct persistent_ram_buffer *buffer;
> size_t buffer_size;
> u32 flags;
> - raw_spinlock_t buffer_lock;
>
> /* ECC correction */
> char *par_buffer;
> --
> 2.9.0
>
Powered by blists - more mailing lists