Skip to contentSkip to conversation

Only do century BCD conversion when we know the RTC is BCD

PATCH In review Andi Kleen 3 participants 6 messages Feb 8, 2008 – Feb 9, 2008 x86

lore
Share & tools
lore
author
#1 lore raw

PATCH 2/3 Use 2000 offset for 32bit kernels

In reply to Andi Kleen

We know it is already after 2000.

This extends the effective lifetime of 32bit systems by 8 years:
from 2030 to 2038.

The only drawback is that users cannot set the cmos date to before 2000 now on 32bit with systems that don't support extended century in the RTC clock. 64bit systems had this limitation for some time and nobody complained.

And they can always set it to such a date in Linux only using date -s if they really want.

---
 arch/x86/kernel/rtc.c |   10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)
arch/x86/kernel/rtc.c +3 -7
Index: linux/arch/x86/kernel/rtc.c
===================================================================
@@ -9,7 +9,6 @@
 #include <asm/vsyscall.h>
 
 #ifdef CONFIG_X86_32
-# define CMOS_YEARS_OFFS 1900
 /*
  * This is a special lock that is owned by the CPU and holds the index
  * register we are working with.  It is required for NMI access to the
@@ -17,14 +16,11 @@
  */
 volatile unsigned long cmos_lock = 0;
 EXPORT_SYMBOL(cmos_lock);
-#else
-/*
- * x86-64 systems only exists since 2002.
- * This will work up to Dec 31, 2100
- */
-# define CMOS_YEARS_OFFS 2000
 #endif
 
+/* For two digit years assume time is always after that */
+#define CMOS_YEARS_OFFS 2000
+
 DEFINE_SPINLOCK(rtc_lock);
 EXPORT_SYMBOL(rtc_lock);
 

Andi Kleen

author
#0 lore raw

PATCH 1/3 Only do century BCD conversion when we know the RTC is BCD

Minor logic fix. The century change was previously always BCD, even when the CMOS data would report itself not being BCD.

---
 arch/x86/kernel/rtc.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/rtc.c +1 -1
Index: linux/arch/x86/kernel/rtc.c
===================================================================
@@ -130,10 +130,10 @@ unsigned long mach_get_cmos_time(void)
 		BCD_TO_BIN(day);
 		BCD_TO_BIN(mon);
 		BCD_TO_BIN(year);
+		BCD_TO_BIN(century);
 	}
 
 	if (century) {
-		BCD_TO_BIN(century);
 		year += century * 100;
 		printk(KERN_INFO "Extended CMOS year: %d\n", century * 100);
 	} else {
author
#2 lore raw

PATCH 3/3 Enable ACPI extended century handling for 32bit too

Not that it matters much, see comment in the code.

---
 arch/x86/kernel/rtc.c |    7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
arch/x86/kernel/rtc.c +4 -3
Index: linux/arch/x86/kernel/rtc.c
===================================================================
@@ -112,12 +112,13 @@ unsigned long mach_get_cmos_time(void)
 	mon = CMOS_READ(RTC_MONTH);
 	year = CMOS_READ(RTC_YEAR);
 
-#if defined(CONFIG_ACPI) && defined(CONFIG_X86_64)
-	/* CHECKME: Is this really 64bit only ??? */
+	/*
+	 * On 32bit not strictly needed because the world ends in 2038
+	 * and the code can handle that with the 2 digit heuristics.
+	 */
 	if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID &&
 	    acpi_gbl_FADT.century)
 		century = CMOS_READ(acpi_gbl_FADT.century);
-#endif
 
 	if (RTC_ALWAYS_BCD || !(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY)) {
 		BCD_TO_BIN(sec);
#4 lore raw
In reply to Andi Kleen
Quoted 3 lines · Andi Kleen

Minor logic fix. The century change was previously always BCD, even when the CMOS data would report itself not being BCD.

Where was it previously always BCD ?

The code flow is taken 1:1 from the original x8664 code. Just the BCD_TO_BIN code has been made conditional for i386, which is not chanining the logic at all, because RTC_ALWAYS_BCD is always true on x8664.

While your change does not do any harm due to BCD_TO_BIN(0) = 0, it is just not fixing anything.

Thanks,
tglx

Quoted 23 lines
---
 arch/x86/kernel/rtc.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: linux/arch/x86/kernel/rtc.c
===================================================================

arch/x86/kernel/rtc.c +1 -1
@@ -130,10 +130,10 @@ unsigned long mach_get_cmos_time(void)
 		BCD_TO_BIN(day);
 		BCD_TO_BIN(mon);
 		BCD_TO_BIN(year);
+		BCD_TO_BIN(century);
 	}
 
 	if (century) {
-		BCD_TO_BIN(century);
 		year += century * 100;
 		printk(KERN_INFO "Extended CMOS year: %d\n", century * 100);
 	} else {
#3 lore raw

Re: [PATCH] [3/3] Enable ACPI extended century handling for 32bit too

In reply to Andi Kleen
Quoted 16 lines · Andi Kleen

Not that it matters much, see comment in the code.

---
 arch/x86/kernel/rtc.c |    7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

Index: linux/arch/x86/kernel/rtc.c
===================================================================

arch/x86/kernel/rtc.c +0 -0
@@ -112,12 +112,13 @@ unsigned long mach_get_cmos_time(void)
 	mon = CMOS_READ(RTC_MONTH);
 	year = CMOS_READ(RTC_YEAR);

arch/x86/kernel/rtc.c unsigned long mach_get_cmos_time(void) · line 121

Quoted 10 lines
arch/x86/kernel/rtc.c +4 -2
 
-#if defined(CONFIG_ACPI) && defined(CONFIG_X86_64)
-	/* CHECKME: Is this really 64bit only ??? */
+	/*
+	 * On 32bit not strictly needed because the world ends in 2038
+	 * and the code can handle that with the 2 digit heuristics.
+	 */
 	if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID &&
 	    acpi_gbl_FADT.century)
 		century = CMOS_READ(acpi_gbl_FADT.century);

How does this compile with CONFIG_ACPI=n ?

Thanks,
tglx

#5 lore raw
In reply to Thomas Gleixner
Quoted 6 lines · Thomas Gleixner
Quoted 3 lines · Andi Kleen

Minor logic fix. The century change was previously always BCD, even when the CMOS data would report itself not being BCD.

Quoted 6 lines

Where was it previously always BCD ?

The code flow is taken 1:1 from the original x8664 code. Just the BCD_TO_BIN code has been made conditional for i386, which is not chanining the logic at all, because RTC_ALWAYS_BCD is always true on x8664.

Ah that's true -- i missed that indeed. I don't actually remember why it was hardcoded like this on x86-64. I don't think there is a good reason for it and if there's a bit for this in the CMOS it ought be better checked.

Quoted 3 lines

While your change does not do any harm due to BCD_TO_BIN(0) = 0, it is just not fixing anything.

In theory it would make sense with the followup change to do the extended century on 32bit too, but in practice all these systems should be BCD anyways (I hope)

-Andi