From: Anton Blanchard <anton@samba.org>
To: benh@kernel.crashing.org, paulus@samba.org
Cc: linuxppc-dev@ozlabs.org
Subject: [PATCH 7/9] powerpc: Fix corruption when grabbing FWNMI data
Date: Wed, 12 Jan 2011 16:50:51 +1100 [thread overview]
Message-ID: <20110112165051.1a14d652@kryten> (raw)
In-Reply-To: <20110112164318.753a435b@kryten>
The FWNMI code uses a global buffer without any locks to read the RTAS error
information. If two CPUs take a machine check at once then we will corrupt
this buffer.
Since most FWNMI rtas messages are not of the extended type, we can create a
64bit percpu buffer and use it where possible. If we do receive an extended
RTAS log then we fall back to the old behaviour of using the global buffer.
Signed-off-by: Anton Blanchard <anton@samba.org>
---
Index: powerpc.git/arch/powerpc/platforms/pseries/ras.c
===================================================================
--- powerpc.git.orig/arch/powerpc/platforms/pseries/ras.c 2010-10-15 13:23:38.701320228 +1100
+++ powerpc.git/arch/powerpc/platforms/pseries/ras.c 2010-10-15 13:23:39.901333776 +1100
@@ -54,7 +54,8 @@
static unsigned char ras_log_buf[RTAS_ERROR_LOG_MAX];
static DEFINE_SPINLOCK(ras_log_buf_lock);
-static char mce_data_buf[RTAS_ERROR_LOG_MAX];
+static char global_mce_data_buf[RTAS_ERROR_LOG_MAX];
+static DEFINE_PER_CPU(__u64, mce_data_buf);
static int ras_get_sensor_state_token;
static int ras_check_exception_token;
@@ -196,12 +197,24 @@ static irqreturn_t ras_error_interrupt(i
return IRQ_HANDLED;
}
-/* Get the error information for errors coming through the
+/*
+ * Some versions of FWNMI place the buffer inside the 4kB page starting at
+ * 0x7000. Other versions place it inside the rtas buffer. We check both.
+ */
+#define VALID_FWNMI_BUFFER(A) \
+ ((((A) >= 0x7000) && ((A) < 0x7ff0)) || \
+ (((A) >= rtas.base) && ((A) < (rtas.base + rtas.size - 16))))
+
+/*
+ * Get the error information for errors coming through the
* FWNMI vectors. The pt_regs' r3 will be updated to reflect
* the actual r3 if possible, and a ptr to the error log entry
* will be returned if found.
*
- * The mce_data_buf does not have any locks or protection around it,
+ * If the RTAS error is not of the extended type, then we put it in a per
+ * cpu 64bit buffer. If it is the extended type we use global_mce_data_buf.
+ *
+ * The global_mce_data_buf does not have any locks or protection around it,
* if a second machine check comes in, or a system reset is done
* before we have logged the error, then we will get corruption in the
* error log. This is preferable over holding off on calling
@@ -210,20 +223,31 @@ static irqreturn_t ras_error_interrupt(i
*/
static struct rtas_error_log *fwnmi_get_errinfo(struct pt_regs *regs)
{
- unsigned long errdata = regs->gpr[3];
- struct rtas_error_log *errhdr = NULL;
unsigned long *savep;
+ struct rtas_error_log *h, *errhdr = NULL;
- if ((errdata >= 0x7000 && errdata < 0x7fff0) ||
- (errdata >= rtas.base && errdata < rtas.base + rtas.size - 16)) {
- savep = __va(errdata);
- regs->gpr[3] = savep[0]; /* restore original r3 */
- memset(mce_data_buf, 0, RTAS_ERROR_LOG_MAX);
- memcpy(mce_data_buf, (char *)(savep + 1), RTAS_ERROR_LOG_MAX);
- errhdr = (struct rtas_error_log *)mce_data_buf;
+ if (!VALID_FWNMI_BUFFER(regs->gpr[3])) {
+ printk(KERN_ERR "FWNMI: corrupt r3\n");
+ return NULL;
+ }
+
+ savep = __va(regs->gpr[3]);
+ regs->gpr[3] = savep[0]; /* restore original r3 */
+
+ /* If it isn't an extended log we can use the per cpu 64bit buffer */
+ h = (struct rtas_error_log *)&savep[1];
+ if (!h->extended) {
+ memcpy(&__get_cpu_var(mce_data_buf), h, sizeof(__u64));
+ errhdr = (struct rtas_error_log *)&__get_cpu_var(mce_data_buf);
} else {
- printk("FWNMI: corrupt r3\n");
+ int len;
+
+ len = max_t(int, 8+h->extended_log_length, RTAS_ERROR_LOG_MAX);
+ memset(global_mce_data_buf, 0, RTAS_ERROR_LOG_MAX);
+ memcpy(global_mce_data_buf, h, len);
+ errhdr = (struct rtas_error_log *)global_mce_data_buf;
}
+
return errhdr;
}
@@ -235,7 +259,7 @@ static void fwnmi_release_errinfo(void)
{
int ret = rtas_call(rtas_token("ibm,nmi-interlock"), 0, 1, NULL);
if (ret != 0)
- printk("FWNMI: nmi-interlock failed: %d\n", ret);
+ printk(KERN_ERR "FWNMI: nmi-interlock failed: %d\n", ret);
}
int pSeries_system_reset_exception(struct pt_regs *regs)
next prev parent reply other threads:[~2011-01-12 5:50 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-12 5:43 [PATCH 0/9] Machine check handling fixes Anton Blanchard
2011-01-12 5:44 ` [PATCH 1/9] powerpc: Print 32 bits of DSISR in show_regs Anton Blanchard
2011-01-12 5:45 ` [PATCH 2/9] powerpc: Don't force MSR_RI in machine_check_exception Anton Blanchard
2011-01-12 5:46 ` [PATCH 3/9] powerpc: Never halt RTAS error logging after receiving an unrecoverable machine check Anton Blanchard
2011-01-12 5:47 ` [PATCH 4/9] powerpc: Remove duplicate debugger hook in machine_check_exception Anton Blanchard
2011-01-12 5:48 ` [PATCH 5/9] powerpc: Don't silently handle machine checks from userspace Anton Blanchard
2011-01-12 5:49 ` [PATCH 6/9] powerpc: Rework pseries machine check handler Anton Blanchard
2011-01-12 5:50 ` Anton Blanchard [this message]
2011-01-12 5:51 ` [PATCH 8/9] powerpc: Check RTAS extended log flag before checking length Anton Blanchard
2011-01-12 5:52 ` [PATCH 9/9] powerpc: machine_check_generic is wrong on 64bit Anton Blanchard
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=20110112165051.1a14d652@kryten \
--to=anton@samba.org \
--cc=benh@kernel.crashing.org \
--cc=linuxppc-dev@ozlabs.org \
--cc=paulus@samba.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).