| 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
| ||
|
Message-ID: <20210226175837.GA184397@agluck-desk2.amr.corp.intel.com>
Date: Fri, 26 Feb 2021 09:58:37 -0800
From: "Luck, Tony" <tony.luck@...el.com>
To: Aili Yao <yaoaili@...gsoft.com>
Cc: Oscar Salvador <osalvador@...e.de>,
HORIGUCHI NAOYA(堀口 直也)
<naoya.horiguchi@....com>, "david@...hat.com" <david@...hat.com>,
"akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
"bp@...en8.de" <bp@...en8.de>,
"tglx@...utronix.de" <tglx@...utronix.de>,
"mingo@...hat.com" <mingo@...hat.com>,
"hpa@...or.com" <hpa@...or.com>, "x86@...nel.org" <x86@...nel.org>,
"linux-edac@...r.kernel.org" <linux-edac@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"linux-mm@...ck.org" <linux-mm@...ck.org>,
"yangfeng1@...gsoft.com" <yangfeng1@...gsoft.com>
Subject: Re: [PATCH] mm,hwpoison: return -EBUSY when page already poisoned
On Fri, Feb 26, 2021 at 10:52:50AM +0800, Aili Yao wrote:
> Hi naoya,Oscar,david:
> >
> > > We could use some negative value (error code) to report the reported case,
> > > then as you mentioned above, some callers need change to handle the
> > > new case, and the same is true if you use some positive value.
> > > My preference is -EHWPOISON, but other options are fine if justified well.
> >
> > -EHWPOISON seems like a good fit.
> >
> I am OK with the -EHWPOISON error code, But I have one doubt here:
> When we return this -EHWPOISON error code, Does this means we have to add a new error code
> to error-base.h or errno.h? Is this easy realized?
The page already poisoned isn't really an error though. Just the result
of a race condition. What if we added an extra argument to memory_failure()
so it can tell the caller that the specific reason for the early successful
return is that the page was already poisoned?
Something like this (untested - patch against v5.11):
-Tony
---
diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c
index e133ce1e562b..0e32c4d879fb 100644
--- a/arch/x86/kernel/cpu/mce/core.c
+++ b/arch/x86/kernel/cpu/mce/core.c
@@ -637,6 +637,7 @@ static int uc_decode_notifier(struct notifier_block *nb, unsigned long val,
{
struct mce *mce = (struct mce *)data;
unsigned long pfn;
+ int already = 0;
if (!mce || !mce_usable_address(mce))
return NOTIFY_DONE;
@@ -646,8 +647,9 @@ static int uc_decode_notifier(struct notifier_block *nb, unsigned long val,
return NOTIFY_DONE;
pfn = mce->addr >> PAGE_SHIFT;
- if (!memory_failure(pfn, 0)) {
- set_mce_nospec(pfn, whole_page(mce));
+ if (!memory_failure(pfn, 0, &already)) {
+ if (!already)
+ set_mce_nospec(pfn, whole_page(mce));
mce->kflags |= MCE_HANDLED_UC;
}
@@ -1245,15 +1247,19 @@ static void kill_me_maybe(struct callback_head *cb)
{
struct task_struct *p = container_of(cb, struct task_struct, mce_kill_me);
int flags = MF_ACTION_REQUIRED;
+ int already = 0;
pr_err("Uncorrected hardware memory error in user-access at %llx", p->mce_addr);
if (!p->mce_ripv)
flags |= MF_MUST_KILL;
- if (!memory_failure(p->mce_addr >> PAGE_SHIFT, flags) &&
+ if (!memory_failure(p->mce_addr >> PAGE_SHIFT, flags, &already) &&
!(p->mce_kflags & MCE_IN_KERNEL_COPYIN)) {
- set_mce_nospec(p->mce_addr >> PAGE_SHIFT, p->mce_whole_page);
+ if (already)
+ force_sig(SIGBUS); // BETTER CODE NEEDED HERE!!!
+ else
+ set_mce_nospec(p->mce_addr >> PAGE_SHIFT, p->mce_whole_page);
sync_core();
return;
}
@@ -1440,7 +1446,7 @@ noinstr void do_machine_check(struct pt_regs *regs)
EXPORT_SYMBOL_GPL(do_machine_check);
#ifndef CONFIG_MEMORY_FAILURE
-int memory_failure(unsigned long pfn, int flags)
+int memory_failure(unsigned long pfn, int flags, int *already)
{
/* mce_severity() should not hand us an ACTION_REQUIRED error */
BUG_ON(flags & MF_ACTION_REQUIRED);
diff --git a/drivers/base/memory.c b/drivers/base/memory.c
index eef4ffb6122c..24c36623e492 100644
--- a/drivers/base/memory.c
+++ b/drivers/base/memory.c
@@ -480,7 +480,7 @@ static ssize_t hard_offline_page_store(struct device *dev,
if (kstrtoull(buf, 0, &pfn) < 0)
return -EINVAL;
pfn >>= PAGE_SHIFT;
- ret = memory_failure(pfn, 0);
+ ret = memory_failure(pfn, 0, NULL);
return ret ? ret : count;
}
diff --git a/include/linux/mm.h b/include/linux/mm.h
index ecdf8a8cd6ae..88b92820465c 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -3045,7 +3045,7 @@ enum mf_flags {
MF_MUST_KILL = 1 << 2,
MF_SOFT_OFFLINE = 1 << 3,
};
-extern int memory_failure(unsigned long pfn, int flags);
+extern int memory_failure(unsigned long pfn, int flags, int *already);
extern void memory_failure_queue(unsigned long pfn, int flags);
extern void memory_failure_queue_kick(int cpu);
extern int unpoison_memory(unsigned long pfn);
diff --git a/mm/hwpoison-inject.c b/mm/hwpoison-inject.c
index 1ae1ebc2b9b1..bfd5151dcd3f 100644
--- a/mm/hwpoison-inject.c
+++ b/mm/hwpoison-inject.c
@@ -48,7 +48,7 @@ static int hwpoison_inject(void *data, u64 val)
inject:
pr_info("Injecting memory failure at pfn %#lx\n", pfn);
- return memory_failure(pfn, 0);
+ return memory_failure(pfn, 0, NULL);
}
static int hwpoison_unpoison(void *data, u64 val)
diff --git a/mm/madvise.c b/mm/madvise.c
index 6a660858784b..ade1956632aa 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -907,7 +907,7 @@ static int madvise_inject_error(int behavior,
} else {
pr_info("Injecting memory failure for pfn %#lx at process virtual address %#lx\n",
pfn, start);
- ret = memory_failure(pfn, MF_COUNT_INCREASED);
+ ret = memory_failure(pfn, MF_COUNT_INCREASED, NULL);
}
if (ret)
diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index e9481632fcd1..e8508e4d70e5 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1388,7 +1388,7 @@ static int memory_failure_dev_pagemap(unsigned long pfn, int flags,
* Must run in process context (e.g. a work queue) with interrupts
* enabled and no spinlocks hold.
*/
-int memory_failure(unsigned long pfn, int flags)
+int memory_failure(unsigned long pfn, int flags, int *already)
{
struct page *p;
struct page *hpage;
@@ -1418,6 +1418,8 @@ int memory_failure(unsigned long pfn, int flags)
if (PageHuge(p))
return memory_failure_hugetlb(pfn, flags);
if (TestSetPageHWPoison(p)) {
+ if (already)
+ *already = 1;
pr_err("Memory failure: %#lx: already hardware poisoned\n",
pfn);
return 0;
@@ -1624,7 +1626,7 @@ static void memory_failure_work_func(struct work_struct *work)
if (entry.flags & MF_SOFT_OFFLINE)
soft_offline_page(entry.pfn, entry.flags);
else
- memory_failure(entry.pfn, entry.flags);
+ memory_failure(entry.pfn, entry.flags, NULL);
}
}
Powered by blists - more mailing lists