Skip to contentSkip to conversation

m68k: allow ColdFire m5441x parts to run with MMU enabled

PATCH Angelo Dureghello 3 participants 29 messages Jul 14, 2017 – Sep 8, 2017

Subsystems (7)
lore
Share & tools
lore
#0 lore raw

Hi Greg,

The Freescale ColdFire M5441x system-on-chip parts have full paged MMU hardware support. So far though we have only allowed them to be configured for use in non-MMU mode.

All required kernel changes to support operation of the M5441x parts with MMU enabled have been pushed into the kernel, so now we can allow it to be configured and used with the MMU enabled.

I don't actually have any M5441x based hardware so I can't do any real testing on this. The changes up to now were based on Yannick Gicquel initial patches to support this. Is anybody out there able to test this properly? I won't push this any further until we have some confirmation that it all works as expected.

---
 arch/m68k/Kconfig.cpu | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/m68k/Kconfig.cpu +1 -1
index d2219f30..4dc51c0 100644
@@ -283,7 +283,7 @@ config M548x
 
 config M5441x
 	bool "MCF5441x"
-	depends on !MMU
+	select MMU_COLDFIRE if MMU
 	select GENERIC_CLOCKEVENTS
 	select HAVE_CACHE_CB
 	help
Signature · 44 lines
1.9.1


I tested the patch on my mcf54415 based board, get silence after u-boot at kernel execution


U-Boot 2017.05-00709-g9d9f074dfe-dirty (Jul 14 2017 - 01:11:57 +0200)

CPU:   Freescale MCF54410 (Mask:9f Version:2)
       CPU CLK 240 MHz BUS CLK 120 MHz FLB CLK 60 MHz
       INP CLK 30 MHz VCO CLK 480 MHz
SPI:   ready
DRAM:  128 MiB
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
device 0 offset 0x100000, size 0x1d9720
SF: 1939232 bytes @ 0x100000 Read: OK
## Booting kernel from Legacy Image at 40001000 ...
   Image Name:   mainline kernel
   Created:      2017-07-14  22:46:07 UTC
   Image Type:   M68K Linux Kernel Image (uncompressed)
   Data Size:    1939168 Bytes = 1.8 MiB
   Load Address: 40001000
   Entry Point:  40001000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK

--- silence, kernel execution hangs here ---


Regards,
Angelo Dureghello



--
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
[prev in list] [next in list] [prev in thread] [next in thread]
#1 lore raw

Hi Greg,

Quoted 26 lines · Angelo Dureghello

Hi Greg,

The Freescale ColdFire M5441x system-on-chip parts have full paged MMU hardware support. So far though we have only allowed them to be configured for use in non-MMU mode.

All required kernel changes to support operation of the M5441x parts with MMU enabled have been pushed into the kernel, so now we can allow it to be configured and used with the MMU enabled.

I don't actually have any M5441x based hardware so I can't do any real testing on this. The changes up to now were based on Yannick Gicquel initial patches to support this. Is anybody out there able to test this properly? I won't push this any further until we have some confirmation that it all works as expected.

---
  arch/m68k/Kconfig.cpu | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
arch/m68k/Kconfig.cpu +0 -0
index d2219f30..4dc51c0 100644

arch/m68k/Kconfig.cpu config M548x · line 289

Quoted 8 lines
arch/m68k/Kconfig.cpu +1 -1
@@ -283,7 +283,7 @@ config M548x
 
  config M5441x
      bool "MCF5441x"
-    depends on !MMU
+    select MMU_COLDFIRE if MMU
      select GENERIC_CLOCKEVENTS
      select HAVE_CACHE_CB
      help

forgot some useful info on my test:

- i built a 4.12 kernel from mainline (just pulled git master) and tested to boot properly as no-mmu

- i have those entries in my defconfig:
....
CONFIG_BINFMT_FLAT=y
CONFIG_BINFMT_ZFLAT=y
CONFIG_BINFMT_SHARED_FLAT=y
CONFIG_BINFMT_MISC=y
....

I can help on debug when back from holidays, in 3 weeks.

Regards,
Angelo Dureghello

#2 lore raw

Hi Angelo,

Quoted 48 lines · Angelo Dureghello
Quoted 24 lines · Angelo Dureghello

The Freescale ColdFire M5441x system-on-chip parts have full paged MMU hardware support. So far though we have only allowed them to be configured for use in non-MMU mode.

All required kernel changes to support operation of the M5441x parts with MMU enabled have been pushed into the kernel, so now we can allow it to be configured and used with the MMU enabled.

I don't actually have any M5441x based hardware so I can't do any real testing on this. The changes up to now were based on Yannick Gicquel initial patches to support this. Is anybody out there able to test this properly? I won't push this any further until we have some confirmation that it all works as expected.

---
  arch/m68k/Kconfig.cpu | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
arch/m68k/Kconfig.cpu +0 -0
index d2219f30..4dc51c0 100644

arch/m68k/Kconfig.cpu config M548x · line 289

Quoted 8 lines
arch/m68k/Kconfig.cpu +1 -1
@@ -283,7 +283,7 @@ config M548x
 
  config M5441x
      bool "MCF5441x"
-    depends on !MMU
+    select MMU_COLDFIRE if MMU
      select GENERIC_CLOCKEVENTS
      select HAVE_CACHE_CB
      help

forgot some useful info on my test:

- i built a 4.12 kernel from mainline (just pulled git master) and tested to boot properly as no-mmu

- i have those entries in my defconfig:
....
CONFIG_BINFMT_FLAT=y
CONFIG_BINFMT_ZFLAT=y
CONFIG_BINFMT_SHARED_FLAT=y
CONFIG_BINFMT_MISC=y
....

Quoted 1 line

I can help on debug when back from holidays, in 3 weeks.

Thanks for trying this out.

Let me know when you have some time to debug further. First up I would like to know what your RAM setup looks like. Is it mapped from a base address of 0?
What is the RAM size?

Regards
Greg

#3 lore raw

Hi Greg,

Quoted 51 lines · Greg Ungerer

Hi Angelo,

Quoted 48 lines · Angelo Dureghello
Quoted 24 lines · Angelo Dureghello

The Freescale ColdFire M5441x system-on-chip parts have full paged MMU hardware support. So far though we have only allowed them to be configured for use in non-MMU mode.

All required kernel changes to support operation of the M5441x parts with MMU enabled have been pushed into the kernel, so now we can allow it to be configured and used with the MMU enabled.

I don't actually have any M5441x based hardware so I can't do any real testing on this. The changes up to now were based on Yannick Gicquel initial patches to support this. Is anybody out there able to test this properly? I won't push this any further until we have some confirmation that it all works as expected.

---
  arch/m68k/Kconfig.cpu | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
arch/m68k/Kconfig.cpu +0 -0
index d2219f30..4dc51c0 100644

arch/m68k/Kconfig.cpu config M548x · line 289

Quoted 8 lines
arch/m68k/Kconfig.cpu +1 -1
@@ -283,7 +283,7 @@ config M548x
 
  config M5441x
      bool "MCF5441x"
-    depends on !MMU
+    select MMU_COLDFIRE if MMU
      select GENERIC_CLOCKEVENTS
      select HAVE_CACHE_CB
      help

forgot some useful info on my test:

- i built a 4.12 kernel from mainline (just pulled git master) and tested to boot properly as no-mmu

- i have those entries in my defconfig:
....
CONFIG_BINFMT_FLAT=y
CONFIG_BINFMT_ZFLAT=y
CONFIG_BINFMT_SHARED_FLAT=y
CONFIG_BINFMT_MISC=y
....

Quoted 9 lines
Quoted 1 line

I can help on debug when back from holidays, in 3 weeks.

Thanks for trying this out.

Let me know when you have some time to debug further. First up I would like to know what your RAM setup looks like. Is it mapped from a base address of 0?
What is the RAM size?

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"
CONFIG_DEFAULT_HOSTNAME="stmark2"
CONFIG_SYSVIPC=y
# CONFIG_FHANDLE is not set
CONFIG_LOG_BUF_SHIFT=14
CONFIG_NAMESPACES=y
CONFIG_BLK_DEV_INITRD=y
CONFIG_INITRAMFS_SOURCE="../uClinux-dist/romfs"
# CONFIG_RD_BZIP2 is not set
# CONFIG_RD_LZMA is not set
# CONFIG_RD_XZ is not set
# CONFIG_RD_LZO is not set
# CONFIG_RD_LZ4 is not set
CONFIG_CC_OPTIMIZE_FOR_SIZE=y
# CONFIG_AIO is not set
# CONFIG_ADVISE_SYSCALLS is not set
# CONFIG_MEMBARRIER is not set
CONFIG_EMBEDDED=y
# CONFIG_VM_EVENT_COUNTERS is not set
# CONFIG_COMPAT_BRK is not set
# CONFIG_LBDAF is not set
# CONFIG_BLK_DEV_BSG is not set
CONFIG_BLK_CMDLINE_PARSER=y
CONFIG_COLDFIRE=y
CONFIG_M5441x=y
CONFIG_CLOCK_FREQ=240000000
CONFIG_STMARK2=y
CONFIG_RAMBASE=0x40000000
CONFIG_RAMSIZE=0x8000000
CONFIG_VECTORBASE=0x40000000
CONFIG_KERNELBASE=0x40001000
CONFIG_BINFMT_FLAT=y
CONFIG_BINFMT_ZFLAT=y
CONFIG_BINFMT_SHARED_FLAT=y
CONFIG_BINFMT_MISC=y
# CONFIG_UEVENT_HELPER is not set
CONFIG_DEVTMPFS=y
CONFIG_DEVTMPFS_MOUNT=y
CONFIG_FW_LOADER_USER_HELPER_FALLBACK=y
# CONFIG_ALLOW_DEV_COREDUMP is not set
CONFIG_MTD=y
CONFIG_MTD_CMDLINE_PARTS=y
CONFIG_MTD_BLOCK=y
CONFIG_FTL=y
CONFIG_MTD_CFI=y
CONFIG_MTD_JEDECPROBE=y
CONFIG_MTD_CFI_ADV_OPTIONS=y
CONFIG_MTD_CFI_LE_BYTE_SWAP=y
CONFIG_MTD_CFI_GEOMETRY=y
# CONFIG_MTD_CFI_I2 is not set
CONFIG_MTD_CFI_AMDSTD=y
CONFIG_MTD_CFI_STAA=y
CONFIG_MTD_ROM=y
CONFIG_MTD_COMPLEX_MAPPINGS=y
CONFIG_MTD_PHYSMAP=y
CONFIG_MTD_PLATRAM=y
CONFIG_MTD_SPI_NOR=y
# CONFIG_INPUT_KEYBOARD is not set
# CONFIG_INPUT_MOUSE is not set
CONFIG_SERIO_LIBPS2=y
# CONFIG_UNIX98_PTYS is not set
# CONFIG_DEVMEM is not set
CONFIG_SERIAL_MCF=y
CONFIG_SERIAL_MCF_BAUDRATE=115200
CONFIG_SERIAL_MCF_CONSOLE=y
# CONFIG_HW_RANDOM is not set
CONFIG_SPI=y
# CONFIG_HWMON is not set
# CONFIG_HID is not set
# CONFIG_USB_SUPPORT is not set
# CONFIG_FILE_LOCKING is not set
# CONFIG_DNOTIFY is not set
# CONFIG_INOTIFY_USER is not set
CONFIG_FSCACHE=y
# CONFIG_PROC_SYSCTL is not set
CONFIG_PRINTK_TIME=y
# CONFIG_ENABLE_WARN_DEPRECATED is not set
# CONFIG_SECTION_MISMATCH_WARN_ONLY is not set
CONFIG_PANIC_ON_OOPS=y
# CONFIG_SCHED_DEBUG is not set
# CONFIG_DEBUG_BUGVERBOSE is not set
CONFIG_BOOTPARAM=y
CONFIG_BOOTPARAM_STRING="console=ttyS0,115200 root=/dev/ram0 rw rootfstype=ramfs rdinit=/bin/init devtmpfs.mount=1" CONFIG_CRYPTO=y
# CONFIG_CRYPTO_ECHAINIV is not set
CONFIG_CRYPTO_ANSI_CPRNG=y
# CONFIG_CRYPTO_HW is not set
CONFIG_CRC16=y

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

Regards
Greg

Signature · 3 lines
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Regards,
Angelo

#4 lore raw

Hi Angelo,

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

Regards
Greg

drivers/tty/serial/mcf.c +24 -0
index 02eb322..58b854e 100644
@@ -568,6 +568,30 @@ static int __init mcf_console_setup(struct console *co, char *options)
 
 /****************************************************************************/
 
+static void putrawch(char c)
+{
+	unsigned int membase = MCFUART_BASE0;
+	int i;
+
+	for (i = 0; (i < 0x10000); i++) {
+		if (readb(membase + MCFUART_USR) & MCFUART_USR_TXREADY)
+			break;
+	}
+	writeb(c, membase + MCFUART_UTB);
+	for (i = 0; (i < 0x10000); i++) {
+		if (readb(membase + MCFUART_USR) & MCFUART_USR_TXREADY)
+			break;
+	}
+}
+
+void printraw(const char *s)
+{
+	for (; *s != 0; s++)
+		putrawch(*s);
+}
+
+/****************************************************************************/
+
 static struct uart_driver mcf_driver;
 
 static struct console mcf_console = {
kernel/printk/printk.c +2 -0
index 2984fb0..39e62d1 100644
@@ -1721,6 +1721,8 @@ asmlinkage int vprintk_emit(int facility, int level,
 	 */
 	text_len = vscnprintf(text, sizeof(textbuf), fmt, args);
 
+	{ extern void printraw(const char *); printraw(text); }
+
 	/* mark and strip a trailing newline */
 	if (text_len && text[text_len-1] == '\n') {
 		text_len--;
#5 lore raw

Hi Greg,

Quoted 24 lines · Greg Ungerer

Hi Angelo,

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

But i have seen that it is not a problem, after MMU is enabled in head.S the jump

movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)
...
setup_arch(&command_line); setup_mm.c

      ...
      paging_init();               mm/mcfmmu.c
         ...
         empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
         ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
	cf_bootmem_alloc();
	mmu_context_init();

#endif

Do also m5441x.c maybe need this calls ?

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ? Waiting your suggestion on how to proceed.

Regards,
Angelo Dureghello

Quoted 3 lines

Regards
Greg

#6 lore raw

Hi Angelo,

Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)
...
setup_arch(&command_line); setup_mm.c

      ...
      paging_init();               mm/mcfmmu.c
         ...
         empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
         ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 8 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
    cf_bootmem_alloc();
    mmu_context_init();

#endif

Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
# 

Regards
Greg

arch/m68k/coldfire/m54xx.c +0 -4
index c552851..704efea 100644
@@ -95,10 +95,6 @@ static void mcf54xx_reset(void)
 
 void __init config_BSP(char *commandp, int size)
 {
-#ifdef CONFIG_MMU
-	cf_bootmem_alloc();
-	mmu_context_init();
-#endif
 	mach_reset = mcf54xx_reset;
 	mach_sched_init = hw_timer_init;
 	m54xx_uarts_init();
arch/m68k/include/asm/mmu_context.h +0 -1
index 4a6ae6d..00c28b1 100644
@@ -91,7 +91,6 @@ static inline void activate_mm(struct mm_struct *active_mm,
 
 #define deactivate_mm(tsk, mm) do { } while (0)
 
-extern void mmu_context_init(void);
 #define prepare_arch_switch(next) load_ksp_mmu(next)
 
 static inline void load_ksp_mmu(struct task_struct *task)
arch/m68k/kernel/setup_mm.c +1 -0
index 7a2c212..5632c48 100644
@@ -343,6 +343,7 @@ void __init setup_arch(char **cmdline_p)
 #ifdef CONFIG_COLDFIRE
 	case MACH_M54XX:
 	case MACH_M5441X:
+		cf_bootmem_alloc();
 		config_BSP(NULL, 0);
 		break;
 #endif
arch/m68k/mm/mcfmmu.c +19 -16
index 87131cd..d09a317 100644
@@ -150,6 +150,24 @@ int cf_tlb_miss(struct pt_regs *regs, int write, int dtlb, int extension_word)
 	return 0;
 }
 
+/*
+ * Initialize the context management stuff.
+ * The following was taken from arch/ppc/mmu_context.c
+ */
+static void __init mmu_context_init(void)
+{
+	/*
+	 * Some processors have too few contexts to reserve one for
+	 * init_mm, and require using context 0 for a normal task.
+	 * Other processors reserve the use of context zero for the kernel.
+	 * This code assumes FIRST_CONTEXT < 32.
+	 */
+	context_map[0] = (1 << FIRST_CONTEXT) - 1;
+	next_mmu_context = FIRST_CONTEXT;
+	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+}
+
+
 void __init cf_bootmem_alloc(void)
 {
 	unsigned long start_pfn;
@@ -177,23 +195,8 @@ void __init cf_bootmem_alloc(void)
 	memstart += init_bootmem_node(NODE_DATA(0), start_pfn,
 		min_low_pfn, max_low_pfn);
 	free_bootmem_node(NODE_DATA(0), memstart, _ramend - memstart);
-}
 
-/*
- * Initialize the context management stuff.
- * The following was taken from arch/ppc/mmu_context.c
- */
-void __init mmu_context_init(void)
-{
-	/*
-	 * Some processors have too few contexts to reserve one for
-	 * init_mm, and require using context 0 for a normal task.
-	 * Other processors reserve the use of context zero for the kernel.
-	 * This code assumes FIRST_CONTEXT < 32.
-	 */
-	context_map[0] = (1 << FIRST_CONTEXT) - 1;
-	next_mmu_context = FIRST_CONTEXT;
-	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+	mmu_context_init();
 }
 
 /*
#7 lore raw

Hi Greg,

Quoted 114 lines · Greg Ungerer

Hi Angelo,

Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

    ...
    setup_arch(&command_line);      setup_mm.c
       ...
       paging_init();               mm/mcfmmu.c
          ...
          empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
          ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 8 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
     cf_bootmem_alloc();
     mmu_context_init();

#endif

Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()
__init_single_page()

     set_page_links()       include/linux/mm.h
        set_page_zone()
          kernel hangs silently on this line
          page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

Regards,
Angelo

#8 lore raw

Hi Angelo,

Quoted 133 lines · Angelo Dureghello
Quoted 112 lines · Greg Ungerer
Quoted 43 lines · Angelo Dureghello
Quoted 23 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 12 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano
2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

    ...
    setup_arch(&command_line);      setup_mm.c
       ...
       paging_init();               mm/mcfmmu.c
          ...
          empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
          ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
     cf_bootmem_alloc();
     mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()
__init_single_page()

      set_page_links()       include/linux/mm.h
         set_page_zone()
           kernel hangs silently on this line
           page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Regards
Greg

#9 lore raw

Hi Greg,

Quoted 157 lines · Greg Ungerer

Hi Angelo,

Quoted 132 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

    ...
    setup_arch(&command_line);      setup_mm.c
       ...
       paging_init();               mm/mcfmmu.c
          ...
          empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
          ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
     cf_bootmem_alloc();
     mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()
__init_single_page()

      set_page_links()       include/linux/mm.h
         set_page_zone()
           kernel hangs silently on this line
           page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

Quoted 4 lines

Regards
Greg

Reagrds
Angelo

Quoted 3 lines
#10 lore raw

Hi Angelo,

Quoted 122 lines · Angelo Dureghello
Quoted 121 lines · Greg Ungerer
Quoted 120 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

    ...
    setup_arch(&command_line);      setup_mm.c
       ...
       paging_init();               mm/mcfmmu.c
          ...
          empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
          ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
     cf_bootmem_alloc();
     mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()

Quoted 8 lines
Quoted 8 lines

__init_single_page()

      set_page_links()       include/linux/mm.h
         set_page_zone()
           kernel hangs silently on this line
           page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

Can you run your current code with the console debug code I sent a little while back?

I ask because I suspect it should give something based on your debug above. I played around a little trying to fake out my configuration to make it look like the RAM was non-zero based. I couldn't get a fail, but I would like to add some more debug to see what is going on with the page pointers from your debug.

Can you apply the attached patch and get any extra debug?

Quoted 30 lines
Quoted 22 lines
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

Quoted 7 lines

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

IIRC the 5475 was the first ColdFire with MMU, it is pretty old. Pretty sure the 54411 came later. Not sure what the thinking was on the different default memory layout though.

Regards
Greg

linux/mm/page_alloc.c +1 -0
index a2019fc..38bafc9 100644
@@ -1178,6 +1178,7 @@ static void free_one_page(struct zone *zone,
 static void __meminit __init_single_page(struct page *page, unsigned long pfn,
 				unsigned long zone, int nid)
 {
+	printk("%s(%d): page=%p pfn=%ld\n", __FILE__, __LINE__, page, pfn);
 	set_page_links(page, zone, nid, pfn);
 	init_page_count(page);
 	page_mapcount_reset(page);
#11 lore raw

Hi Greg,

Quoted 141 lines · Greg Ungerer

Hi Angelo,

Quoted 122 lines · Angelo Dureghello
Quoted 121 lines · Greg Ungerer
Quoted 120 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

     ...
     setup_arch(&command_line);      setup_mm.c
        ...
        paging_init();               mm/mcfmmu.c
           ...
           empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
           ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
      cf_bootmem_alloc();
      mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()

Quoted 8 lines
Quoted 8 lines
   __init_single_page()
       set_page_links()       include/linux/mm.h
          set_page_zone()
            kernel hangs silently on this line
            page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

Can you run your current code with the console debug code I sent a little while back?

Quoted 8 lines

I ask because I suspect it should give something based on your debug above. I played around a little trying to fake out my configuration to make it look like the RAM was non-zero based. I couldn't get a fail, but I would like to add some more debug to see what is going on with the page pointers from your debug.

Can you apply the attached patch and get any extra debug?

Sure. Done as you required, removed all my debug parts and just added yours.

Not sure why, but seems there is now a kind of rollback, as the error now comes again before reaching your printk line. Also some strange ctrl char is printed at each line begin, in the debug printf.

This is the output:

U-Boot 2017.09-rc2-00136-gf069d2fd3e (Aug 20 2017 - 00:08:11 +0200)

CPU: Freescale MCF54410 (Mask:9f Version:2)
CPU CLK 240 MHz BUS CLK 120 MHz FLB CLK 60 MHz INP CLK 30 MHz VCO CLK 480 MHz

SPI:   ready
DRAM:  128 MiB

Using default environment

In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
device 0 offset 0x100000, size 0x700000
SF: 7340032 bytes @ 0x100000 Read: OK
## Booting kernel from Legacy Image at 40001000 ...
   Image Name:   mainline kernel
   Created:      2017-08-21  14:18:52 UTC
   Image Type:   M68K Linux Kernel Image (uncompressed)
   Data Size:    1939176 Bytes = 1.8 MiB

Load Address: 40001000
Entry Point: 40001000
Verifying Checksum ... OK
Loading Kernel Image ... OK
5Linux version 4.12.0stmark2-001-11691-g571d81b2b55f-dirty (angelo@jerusalem) (gcc version 4.9.0 (crosstools-sysam-2016.04.16)) #177 Mon Aug 21 16:18:51 CEST 2017
1bootmem alloc of 8192 bytes failed!
0Kernel panic - not syncing: Out of memory
0---[ end Kernel panic - not syncing: Out of memory

Or better, https://snag.gy/X8y4JM.jpg

I attach my whole current patch.

Quoted 35 lines
Quoted 35 lines
Quoted 23 lines
Quoted 9 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

Quoted 8 lines
Quoted 3 lines

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

IIRC the 5475 was the first ColdFire with MMU, it is pretty old. Pretty sure the 54411 came later. Not sure what the thinking was on the different default memory layout though.

Mmm, i was just curiously noticing that 5475 reaches and higher clock speed as it could be a newer technology. But not important probably.

Quoted 4 lines

Regards
Greg

Will also look into this later.

Regards,
Angelo

arch/m68k/Kconfig.cpu +1 -1
index d2219f30b78f..4dc51c090a45 100644
@@ -283,7 +283,7 @@ config M548x
 
 config M5441x
 	bool "MCF5441x"
-	depends on !MMU
+	select MMU_COLDFIRE if MMU
 	select GENERIC_CLOCKEVENTS
 	select HAVE_CACHE_CB
 	help
arch/m68k/coldfire/m54xx.c +0 -4
index c552851ec617..704efeaeab2d 100644
@@ -95,10 +95,6 @@ static void mcf54xx_reset(void)
 
 void __init config_BSP(char *commandp, int size)
 {
-#ifdef CONFIG_MMU
-	cf_bootmem_alloc();
-	mmu_context_init();
-#endif
 	mach_reset = mcf54xx_reset;
 	mach_sched_init = hw_timer_init;
 	m54xx_uarts_init();
arch/m68k/configs/stmark2_defconfig +3 -2
index ea78b8f6a68e..b886f303cc3f 100644
@@ -21,7 +21,7 @@ CONFIG_EMBEDDED=y
 # CONFIG_LBDAF is not set
 # CONFIG_BLK_DEV_BSG is not set
 CONFIG_BLK_CMDLINE_PARSER=y
-# CONFIG_MMU is not set
+CONFIG_COLDFIRE=y
 CONFIG_M5441x=y
 CONFIG_CLOCK_FREQ=240000000
 CONFIG_STMARK2=y
@@ -30,6 +30,8 @@ CONFIG_RAMSIZE=0x8000000
 CONFIG_VECTORBASE=0x40000000
 CONFIG_KERNELBASE=0x40001000
 CONFIG_BINFMT_FLAT=y
+CONFIG_BINFMT_ZFLAT=y
+CONFIG_BINFMT_SHARED_FLAT=y
 CONFIG_BINFMT_MISC=y
 # CONFIG_UEVENT_HELPER is not set
 CONFIG_DEVTMPFS=y
@@ -53,7 +55,6 @@ CONFIG_MTD_COMPLEX_MAPPINGS=y
 CONFIG_MTD_PHYSMAP=y
 CONFIG_MTD_PLATRAM=y
 CONFIG_MTD_SPI_NOR=y
-# CONFIG_INPUT_MOUSEDEV is not set
 # CONFIG_INPUT_KEYBOARD is not set
 # CONFIG_INPUT_MOUSE is not set
 CONFIG_SERIO_LIBPS2=y
arch/m68k/include/asm/mmu_context.h +0 -1
index 4a6ae6dffa34..00c28b1dc47b 100644
@@ -91,7 +91,6 @@ static inline void activate_mm(struct mm_struct *active_mm,
 
 #define deactivate_mm(tsk, mm) do { } while (0)
 
-extern void mmu_context_init(void);
 #define prepare_arch_switch(next) load_ksp_mmu(next)
 
 static inline void load_ksp_mmu(struct task_struct *task)
arch/m68k/mm/mcfmmu.c +18 -17
index 87131cd3bc8f..c7efdf8e8eae 100644
@@ -150,6 +150,23 @@ int cf_tlb_miss(struct pt_regs *regs, int write, int dtlb, int extension_word)
 	return 0;
 }
 
+/*
+ * Initialize the context management stuff.
+ * The following was taken from arch/ppc/mmu_context.c
+ */
+static void __init mmu_context_init(void)
+{
+	/*
+	 * Some processors have too few contexts to reserve one for
+	 * init_mm, and require using context 0 for a normal task.
+	 * Other processors reserve the use of context zero for the kernel.
+	 * This code assumes FIRST_CONTEXT < 32.
+	 */
+	context_map[0] = (1 << FIRST_CONTEXT) - 1;
+	next_mmu_context = FIRST_CONTEXT;
+	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+}
+
 void __init cf_bootmem_alloc(void)
 {
 	unsigned long start_pfn;
@@ -177,23 +194,7 @@ void __init cf_bootmem_alloc(void)
 	memstart += init_bootmem_node(NODE_DATA(0), start_pfn,
 		min_low_pfn, max_low_pfn);
 	free_bootmem_node(NODE_DATA(0), memstart, _ramend - memstart);
-}
-
-/*
- * Initialize the context management stuff.
- * The following was taken from arch/ppc/mmu_context.c
- */
-void __init mmu_context_init(void)
-{
-	/*
-	 * Some processors have too few contexts to reserve one for
-	 * init_mm, and require using context 0 for a normal task.
-	 * Other processors reserve the use of context zero for the kernel.
-	 * This code assumes FIRST_CONTEXT < 32.
-	 */
-	context_map[0] = (1 << FIRST_CONTEXT) - 1;
-	next_mmu_context = FIRST_CONTEXT;
-	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+	mmu_context_init();
 }
 
 /*
drivers/tty/serial/mcf.c +24 -0
index 02eb32217685..58b854e0713f 100644
@@ -568,6 +568,30 @@ static int __init mcf_console_setup(struct console *co, char *options)
 
 /****************************************************************************/
 
+static void putrawch(char c)
+{
+	unsigned int membase = MCFUART_BASE0;
+	int i;
+
+	for (i = 0; (i < 0x10000); i++) {
+		if (readb(membase + MCFUART_USR) & MCFUART_USR_TXREADY)
+			break;
+	}
+	writeb(c, membase + MCFUART_UTB);
+	for (i = 0; (i < 0x10000); i++) {
+		if (readb(membase + MCFUART_USR) & MCFUART_USR_TXREADY)
+			break;
+	}
+}
+
+void printraw(const char *s)
+{
+	for (; *s != 0; s++)
+		putrawch(*s);
+}
+
+/****************************************************************************/
+
 static struct uart_driver mcf_driver;
 
 static struct console mcf_console = {
kernel/printk/printk.c +2 -0
index fc47863f629c..880a36d05abe 100644
@@ -1720,6 +1720,8 @@ asmlinkage int vprintk_emit(int facility, int level,
 	 */
 	text_len = vscnprintf(text, sizeof(textbuf), fmt, args);
 
+	{ extern void printraw(const char *); printraw(text); }
+
 	/* mark and strip a trailing newline */
 	if (text_len && text[text_len-1] == '\n') {
 		text_len--;
mm/page_alloc.c +1 -0
index 6d30e914afb6..4a6bdaa118a4 100644
@@ -1168,6 +1168,7 @@ static void free_one_page(struct zone *zone,
 static void __meminit __init_single_page(struct page *page, unsigned long pfn,
 				unsigned long zone, int nid)
 {
+	printk("%s(%d): page=%p pfn=%ld\n", __FILE__, __LINE__, page, pfn);
 	set_page_links(page, zone, nid, pfn);
 	init_page_count(page);
 	page_mapcount_reset(page);
#12 lore raw

Hi Angelo,

Quoted 9 lines · Angelo Dureghello
Quoted 1 line · Greg Ungerer

Can you apply the attached patch and get any extra debug?

Sure. Done as you required, removed all my debug parts and just added yours.

Not sure why, but seems there is now a kind of rollback, as the error now comes again before reaching your printk line. Also some strange ctrl char is printed at each line begin, in the debug printf.

See include/linux/kern_levels.h.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

#13 lore raw

Hi Geert,

nice to hear you :)

Quoted 6 lines · Geert Uytterhoeven

Hi Angelo,

Quoted 3 lines · Angelo Dureghello
Quoted 1 line · Greg Ungerer

Can you apply the attached patch and get any extra debug?

Quoted 8 lines
Quoted 6 lines

Sure. Done as you required, removed all my debug parts and just added yours.

Not sure why, but seems there is now a kind of rollback, as the error now comes again before reaching your printk line. Also some strange ctrl char is printed at each line begin, in the debug printf.

See include/linux/kern_levels.h.

many thanks, i fixed a bit the debug routine now it is perfect.

Quoted 5 lines

Gr{oetje,eeting}s,

Geert

Regards,
Angelo

Quoted 6 lines

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

Quoted 1 line

In personal conversations with technical people, I call myself a hacker. But

when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

#14 lore raw
In reply to Greg Ungerer

Hi Greg,

Quoted 184 lines · Greg Ungerer

Hi Angelo,

Quoted 122 lines · Angelo Dureghello
Quoted 121 lines · Greg Ungerer
Quoted 120 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

     ...
     setup_arch(&command_line);      setup_mm.c
        ...
        paging_init();               mm/mcfmmu.c
           ...
           empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
           ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
      cf_bootmem_alloc();
      mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()

Quoted 8 lines
Quoted 8 lines
   __init_single_page()
       set_page_links()       include/linux/mm.h
          set_page_zone()
            kernel hangs silently on this line
            page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

Can you run your current code with the console debug code I sent a little while back?

I ask because I suspect it should give something based on your debug above. I played around a little trying to fake out my configuration to make it look like the RAM was non-zero based. I couldn't get a fail, but I would like to add some more debug to see what is going on with the page pointers from your debug.

Can you apply the attached patch and get any extra debug?

Quoted 34 lines
Quoted 22 lines
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

Quoted 8 lines
Quoted 3 lines

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

IIRC the 5475 was the first ColdFire with MMU, it is pretty old. Pretty sure the 54411 came later. Not sure what the thinking was on the different default memory layout though.

Finally, cleaning out my debug lines, i found i removed an important line. So i am back to original "second" error we was trying to understand.

So current more clear status is:

U-Boot 2017.09-rc2-00151-g2d7cb5b426-dirty (Aug 22 2017 - 00:22:46 +0200)

CPU: Freescale MCF54410 (Mask:9f Version:2)
CPU CLK 240 MHz BUS CLK 120 MHz FLB CLK 60 MHz INP CLK 30 MHz VCO CLK 480 MHz

SPI:   ready
DRAM:  128 MiB
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
device 0 offset 0x100000, size 0x1d9728
SF: 1939240 bytes @ 0x100000 Read: OK
## Booting kernel from Legacy Image at 40001000 ...
   Image Name:   mainline kernel
   Created:      2017-08-22   0:07:25 UTC
   Image Type:   M68K Linux Kernel Image (uncompressed)
   Data Size:    1939176 Bytes = 1.8 MiB

Load Address: 40001000
Entry Point: 40001000
Verifying Checksum ... OK
Loading Kernel Image ... OK
Linux version 4.12.0stmark2-001-11691-g571d81b2b55f-dirty (angelo@jerusalem) (gcc version 4.9.0 (crosstools-sysam-2016.04.16)) #182 Tue Aug 22 02:07:24 CEST 2017

------------[ cut here ]------------
WARNING: CPU: 0 PID: 0 at mm/page_alloc.c:6219 free_area_init_node+0x2f4/0x2fa
CPU: 0 PID: 0 Comm: swapper Not tainted 4.12.0stmark2-001-11691-g571d81b2b55f-dirty #182

Stack from 4017deec:

4017deec
4017b3dd
40007972
00000000
00000000
47d9f62c
00020000
00000000

00000000
4017df9c
40007a14
4016dd8e
0000184b
4019caca
00000009
00000000

00000000
4019caca
4016dd8e
0000184b
48000000
40204000
47d9f62c
40001000

00000000
47d9ef1c
40001480
4013010c
4012cd16
4017dfa8
4019ecc0
00012000

00002000
4019ccb4
00000000
4017df9c
00020000
00000000
4019a3f2
4017df9c

00000001
401da8c0
401da774
4019ebc8
00004000
00000000
00000000
4017dfc8

Call Trace:
 [<40007972>] __warn+0xa4/0xc0
 [<40007a14>] warn_slowpath_null+0x1a/0x22
 [<4019caca>] free_area_init_node+0x2f4/0x2fa
 [<4019caca>] free_area_init_node+0x2f4/0x2fa
 [<40001000>] kernel_pg_dir+0x0/0x1000
 [<40001480>] kernel_pg_dir+0x480/0x1000
 [<4013010c>] memset+0x0/0x80
 [<4012cd16>] strlen+0x0/0x14
 [<4019ecc0>] __alloc_bootmem+0x16/0x3c
 [<4019ccb4>] free_area_init+0x20/0x26
 [<4019a3f2>] paging_init+0xee/0xfa
 [<4019ebc8>] free_bootmem_node+0x0/0x34
 [<40199fbc>] setup_arch+0xcc/0x16e
 [<40024eb2>] printk+0x0/0x18
 [<4019ecaa>] __alloc_bootmem+0x0/0x3c
 [<40198550>] start_kernel+0x68/0x3ae
 [<40001000>] kernel_pg_dir+0x0/0x1000
 [<400020f2>] _exit+0x0/0x6

---[ end trace 0000000000000000 ]---

On node 0 totalpages: 16384
free_area_init_node: node 0, pgdat 401da8c0, node_mem_map a8c0401d
DMA zone: 72 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 16384 pages, LIFO batch:3
/page_alloc.c(1171): page=a8c0401d pfn=131072

Quoted 4 lines

Regards
Greg

Reagrds,
Angelo

#15 lore raw

Hi Angelo,

Quoted 303 lines · Angelo Dureghello
Quoted 182 lines · Greg Ungerer
Quoted 122 lines · Angelo Dureghello
Quoted 121 lines · Greg Ungerer
Quoted 120 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

     ...
     setup_arch(&command_line);      setup_mm.c
        ...
        paging_init();               mm/mcfmmu.c
           ...
           empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
           ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
      cf_bootmem_alloc();
      mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()

Quoted 8 lines
Quoted 8 lines
   __init_single_page()
       set_page_links()       include/linux/mm.h
          set_page_zone()
            kernel hangs silently on this line
            page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

Can you run your current code with the console debug code I sent a little while back?

I ask because I suspect it should give something based on your debug above. I played around a little trying to fake out my configuration to make it look like the RAM was non-zero based. I couldn't get a fail, but I would like to add some more debug to see what is going on with the page pointers from your debug.

Can you apply the attached patch and get any extra debug?

Quoted 34 lines
Quoted 22 lines
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

Quoted 8 lines
Quoted 3 lines

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

IIRC the 5475 was the first ColdFire with MMU, it is pretty old. Pretty sure the 54411 came later. Not sure what the thinking was on the different default memory layout though.

Finally, cleaning out my debug lines, i found i removed an important line. So i am back to original "second" error we was trying to understand.

So current more clear status is:

U-Boot 2017.09-rc2-00151-g2d7cb5b426-dirty (Aug 22 2017 - 00:22:46 +0200)

CPU: Freescale MCF54410 (Mask:9f Version:2)
CPU CLK 240 MHz BUS CLK 120 MHz FLB CLK 60 MHz INP CLK 30 MHz VCO CLK 480 MHz

SPI:   ready
DRAM:  128 MiB
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
device 0 offset 0x100000, size 0x1d9728
SF: 1939240 bytes @ 0x100000 Read: OK
## Booting kernel from Legacy Image at 40001000 ...
   Image Name:   mainline kernel
   Created:      2017-08-22   0:07:25 UTC
   Image Type:   M68K Linux Kernel Image (uncompressed)
   Data Size:    1939176 Bytes = 1.8 MiB

Load Address: 40001000
Entry Point: 40001000
Verifying Checksum ... OK
Loading Kernel Image ... OK
Linux version 4.12.0stmark2-001-11691-g571d81b2b55f-dirty (angelo@jerusalem) (gcc version 4.9.0 (crosstools-sysam-2016.04.16)) #182 Tue Aug 22 02:07:24 CEST 2017

------------[ cut here ]------------
WARNING: CPU: 0 PID: 0 at mm/page_alloc.c:6219 free_area_init_node+0x2f4/0x2fa
CPU: 0 PID: 0 Comm: swapper Not tainted 4.12.0stmark2-001-11691-g571d81b2b55f-dirty #182

Stack from 4017deec:

        4017deec

4017b3dd
40007972
00000000
00000000
47d9f62c
00020000
00000000

        00000000

4017df9c
40007a14
4016dd8e
0000184b
4019caca
00000009
00000000

        00000000

4019caca
4016dd8e
0000184b
48000000
40204000
47d9f62c
40001000

        00000000

47d9ef1c
40001480
4013010c
4012cd16
4017dfa8
4019ecc0
00012000

        00002000

4019ccb4
00000000
4017df9c
00020000
00000000
4019a3f2
4017df9c

        00000001

401da8c0
401da774
4019ebc8
00004000
00000000
00000000
4017dfc8

Call Trace:
 [<40007972>] __warn+0xa4/0xc0
 [<40007a14>] warn_slowpath_null+0x1a/0x22
 [<4019caca>] free_area_init_node+0x2f4/0x2fa
 [<4019caca>] free_area_init_node+0x2f4/0x2fa
 [<40001000>] kernel_pg_dir+0x0/0x1000
 [<40001480>] kernel_pg_dir+0x480/0x1000
 [<4013010c>] memset+0x0/0x80
 [<4012cd16>] strlen+0x0/0x14
 [<4019ecc0>] __alloc_bootmem+0x16/0x3c
 [<4019ccb4>] free_area_init+0x20/0x26
 [<4019a3f2>] paging_init+0xee/0xfa
 [<4019ebc8>] free_bootmem_node+0x0/0x34
 [<40199fbc>] setup_arch+0xcc/0x16e
 [<40024eb2>] printk+0x0/0x18
 [<4019ecaa>] __alloc_bootmem+0x0/0x3c
 [<40198550>] start_kernel+0x68/0x3ae
 [<40001000>] kernel_pg_dir+0x0/0x1000
 [<400020f2>] _exit+0x0/0x6

---[ end trace 0000000000000000 ]---
Quoted 2 lines

On node 0 totalpages: 16384
free_area_init_node: node 0, pgdat 401da8c0, node_mem_map a8c0401d

                                                            ^^^^^^^^
Quoted 4 lines

DMA zone: 72 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 16384 pages, LIFO batch:3
/page_alloc.c(1171): page=a8c0401d pfn=131072

                            ^^^^^^^^

Ok, this is getting somewhere. Clearly that page pointer is not valid.

I'll have a dig around and see if I can figure out what might cause that.

Regards
Greg

#16 lore raw
In reply to Angelo Dureghello

Hi Angelo,

Quoted 303 lines · Angelo Dureghello
Quoted 182 lines · Greg Ungerer
Quoted 122 lines · Angelo Dureghello
Quoted 121 lines · Greg Ungerer
Quoted 120 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

     ...
     setup_arch(&command_line);      setup_mm.c
        ...
        paging_init();               mm/mcfmmu.c
           ...
           empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
           ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
      cf_bootmem_alloc();
      mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()

Quoted 8 lines
Quoted 8 lines
   __init_single_page()
       set_page_links()       include/linux/mm.h
          set_page_zone()
            kernel hangs silently on this line
            page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

Can you run your current code with the console debug code I sent a little while back?

I ask because I suspect it should give something based on your debug above. I played around a little trying to fake out my configuration to make it look like the RAM was non-zero based. I couldn't get a fail, but I would like to add some more debug to see what is going on with the page pointers from your debug.

Can you apply the attached patch and get any extra debug?

Quoted 34 lines
Quoted 22 lines
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

Quoted 8 lines
Quoted 3 lines

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

IIRC the 5475 was the first ColdFire with MMU, it is pretty old. Pretty sure the 54411 came later. Not sure what the thinking was on the different default memory layout though.

Finally, cleaning out my debug lines, i found i removed an important line. So i am back to original "second" error we was trying to understand.

So current more clear status is:

U-Boot 2017.09-rc2-00151-g2d7cb5b426-dirty (Aug 22 2017 - 00:22:46 +0200)

CPU: Freescale MCF54410 (Mask:9f Version:2)
CPU CLK 240 MHz BUS CLK 120 MHz FLB CLK 60 MHz INP CLK 30 MHz VCO CLK 480 MHz

SPI:   ready
DRAM:  128 MiB
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
device 0 offset 0x100000, size 0x1d9728
SF: 1939240 bytes @ 0x100000 Read: OK
## Booting kernel from Legacy Image at 40001000 ...
   Image Name:   mainline kernel
   Created:      2017-08-22   0:07:25 UTC
   Image Type:   M68K Linux Kernel Image (uncompressed)
   Data Size:    1939176 Bytes = 1.8 MiB

Load Address: 40001000
Entry Point: 40001000
Verifying Checksum ... OK
Loading Kernel Image ... OK
Linux version 4.12.0stmark2-001-11691-g571d81b2b55f-dirty (angelo@jerusalem) (gcc version 4.9.0 (crosstools-sysam-2016.04.16)) #182 Tue Aug 22 02:07:24 CEST 2017

------------[ cut here ]------------
WARNING: CPU: 0 PID: 0 at mm/page_alloc.c:6219 free_area_init_node+0x2f4/0x2fa
CPU: 0 PID: 0 Comm: swapper Not tainted 4.12.0stmark2-001-11691-g571d81b2b55f-dirty #182

Stack from 4017deec:

        4017deec

4017b3dd
40007972
00000000
00000000
47d9f62c
00020000
00000000

        00000000

4017df9c
40007a14
4016dd8e
0000184b
4019caca
00000009
00000000

        00000000

4019caca
4016dd8e
0000184b
48000000
40204000
47d9f62c
40001000

        00000000

47d9ef1c
40001480
4013010c
4012cd16
4017dfa8
4019ecc0
00012000

        00002000

4019ccb4
00000000
4017df9c
00020000
00000000
4019a3f2
4017df9c

        00000001

401da8c0
401da774
4019ebc8
00004000
00000000
00000000
4017dfc8

Call Trace:
 [<40007972>] __warn+0xa4/0xc0
 [<40007a14>] warn_slowpath_null+0x1a/0x22
 [<4019caca>] free_area_init_node+0x2f4/0x2fa
 [<4019caca>] free_area_init_node+0x2f4/0x2fa
 [<40001000>] kernel_pg_dir+0x0/0x1000
 [<40001480>] kernel_pg_dir+0x480/0x1000
 [<4013010c>] memset+0x0/0x80
 [<4012cd16>] strlen+0x0/0x14
 [<4019ecc0>] __alloc_bootmem+0x16/0x3c
 [<4019ccb4>] free_area_init+0x20/0x26
 [<4019a3f2>] paging_init+0xee/0xfa
 [<4019ebc8>] free_bootmem_node+0x0/0x34
 [<40199fbc>] setup_arch+0xcc/0x16e
 [<40024eb2>] printk+0x0/0x18
 [<4019ecaa>] __alloc_bootmem+0x0/0x3c
 [<40198550>] start_kernel+0x68/0x3ae
 [<40001000>] kernel_pg_dir+0x0/0x1000
 [<400020f2>] _exit+0x0/0x6

---[ end trace 0000000000000000 ]---
Quoted 6 lines

On node 0 totalpages: 16384
free_area_init_node: node 0, pgdat 401da8c0, node_mem_map a8c0401d
DMA zone: 72 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 16384 pages, LIFO batch:3
/page_alloc.c(1171): page=a8c0401d pfn=131072

Another patch attached that digs a little deeper into why that page pointer ends up being invalid. If you could run with this and send the output that would be great.

Regards
Greg

linux/mm/page_alloc.c +7 -0
index a2019fc..ab95a8c 100644
@@ -1178,6 +1178,13 @@ static void free_one_page(struct zone *zone,
 static void __meminit __init_single_page(struct page *page, unsigned long pfn,
 				unsigned long zone, int nid)
 {
+	printk("%s(%d): page=%p pfn=%ld\n", __FILE__, __LINE__, page, pfn);
+	printk("%s(%d): __va(pfn=%ld)=%p\n", __FILE__, __LINE__, pfn, __va(pfn));
+	printk("%s(%d): pfn_to_virt(pfn=%ld)=%p\n", __FILE__, __LINE__, pfn, pfn_to_virt(pfn));
+	printk("%s(%d): __virt_to_node(pfn_to_virt(pfn=%ld))=%p\n", __FILE__, __LINE__, pfn, __virt_to_node((unsigned long)pfn_to_virt(pfn)));
+	printk("%s(%d): pg_data_map[0].node_start_pfn=%ld\n", __FILE__, __LINE__, pg_data_map[0].node_start_pfn);
+	printk("%s(%d): pg_data_map[0].node_present_pages=%ld\n", __FILE__, __LINE__, pg_data_map[0].node_present_pages);
+
 	set_page_links(page, zone, nid, pfn);
 	init_page_count(page);
 	page_mapcount_reset(page);
#17 lore raw

Hi Greg,

sorry for the late reply,
i decided to study inside the mmu code, working on a separate branch, and after a long hard fight i have the prompt.

So good news, seems i probably have mmu working. See below for further details.

Quoted 313 lines · Greg Ungerer

Hi Angelo,

Quoted 303 lines · Angelo Dureghello
Quoted 182 lines · Greg Ungerer
Quoted 122 lines · Angelo Dureghello
Quoted 121 lines · Greg Ungerer
Quoted 120 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

      ...
      setup_arch(&command_line);      setup_mm.c
         ...
         paging_init();               mm/mcfmmu.c
            ...
            empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
            ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
       cf_bootmem_alloc();
       mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()

Quoted 8 lines
Quoted 8 lines
    __init_single_page()
        set_page_links()       include/linux/mm.h
           set_page_zone()
             kernel hangs silently on this line
             page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

Can you run your current code with the console debug code I sent a little while back?

I ask because I suspect it should give something based on your debug above. I played around a little trying to fake out my configuration to make it look like the RAM was non-zero based. I couldn't get a fail, but I would like to add some more debug to see what is going on with the page pointers from your debug.

Can you apply the attached patch and get any extra debug?

Quoted 34 lines
Quoted 22 lines
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

Quoted 8 lines
Quoted 3 lines

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

IIRC the 5475 was the first ColdFire with MMU, it is pretty old. Pretty sure the 54411 came later. Not sure what the thinking was on the different default memory layout though.

Finally, cleaning out my debug lines, i found i removed an important line. So i am back to original "second" error we was trying to understand.

So current more clear status is:

U-Boot 2017.09-rc2-00151-g2d7cb5b426-dirty (Aug 22 2017 - 00:22:46 +0200)

CPU: Freescale MCF54410 (Mask:9f Version:2)
CPU CLK 240 MHz BUS CLK 120 MHz FLB CLK 60 MHz INP CLK 30 MHz VCO CLK 480 MHz

SPI:   ready
DRAM:  128 MiB
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
device 0 offset 0x100000, size 0x1d9728
SF: 1939240 bytes @ 0x100000 Read: OK
## Booting kernel from Legacy Image at 40001000 ...
    Image Name:   mainline kernel
    Created:      2017-08-22   0:07:25 UTC
    Image Type:   M68K Linux Kernel Image (uncompressed)
    Data Size:    1939176 Bytes = 1.8 MiB

Load Address: 40001000
Entry Point: 40001000
Verifying Checksum ... OK
Loading Kernel Image ... OK
Linux version 4.12.0stmark2-001-11691-g571d81b2b55f-dirty (angelo@jerusalem) (gcc version 4.9.0 (crosstools-sysam-2016.04.16)) #182 Tue Aug 22 02:07:24 CEST 2017

------------[ cut here ]------------
WARNING: CPU: 0 PID: 0 at mm/page_alloc.c:6219 free_area_init_node+0x2f4/0x2fa
CPU: 0 PID: 0 Comm: swapper Not tainted 4.12.0stmark2-001-11691-g571d81b2b55f-dirty #182

Stack from 4017deec:

         4017deec

4017b3dd
40007972
00000000
00000000
47d9f62c
00020000
00000000

         00000000

4017df9c
40007a14
4016dd8e
0000184b
4019caca
00000009
00000000

         00000000

4019caca
4016dd8e
0000184b
48000000
40204000
47d9f62c
40001000

         00000000

47d9ef1c
40001480
4013010c
4012cd16
4017dfa8
4019ecc0
00012000

         00002000

4019ccb4
00000000
4017df9c
00020000
00000000
4019a3f2
4017df9c

         00000001

401da8c0
401da774
4019ebc8
00004000
00000000
00000000
4017dfc8

Call Trace:
  [<40007972>] __warn+0xa4/0xc0
  [<40007a14>] warn_slowpath_null+0x1a/0x22
  [<4019caca>] free_area_init_node+0x2f4/0x2fa
  [<4019caca>] free_area_init_node+0x2f4/0x2fa
  [<40001000>] kernel_pg_dir+0x0/0x1000
  [<40001480>] kernel_pg_dir+0x480/0x1000
  [<4013010c>] memset+0x0/0x80
  [<4012cd16>] strlen+0x0/0x14
  [<4019ecc0>] __alloc_bootmem+0x16/0x3c
  [<4019ccb4>] free_area_init+0x20/0x26
  [<4019a3f2>] paging_init+0xee/0xfa
  [<4019ebc8>] free_bootmem_node+0x0/0x34
  [<40199fbc>] setup_arch+0xcc/0x16e
  [<40024eb2>] printk+0x0/0x18
  [<4019ecaa>] __alloc_bootmem+0x0/0x3c
  [<40198550>] start_kernel+0x68/0x3ae
  [<40001000>] kernel_pg_dir+0x0/0x1000
  [<400020f2>] _exit+0x0/0x6

---[ end trace 0000000000000000 ]---
Quoted 6 lines

On node 0 totalpages: 16384
free_area_init_node: node 0, pgdat 401da8c0, node_mem_map a8c0401d
DMA zone: 72 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 16384 pages, LIFO batch:3
/page_alloc.c(1171): page=a8c0401d pfn=131072

Quoted 4 lines

Another patch attached that digs a little deeper into why that page pointer ends up being invalid. If you could run with this and send the output that would be great.

First of all, this is the trace you required me:

mm/page_alloc.c(1171): page=a8d0401d pfn=131072 mm/page_alloc.c(1172): __va(pfn=131072)=00020000 mm/page_alloc.c(1173): pfn_to_virt(pfn=131072)=40000000 mm/page_alloc.c(1174): __virt_to_node(pfn_to_virt(pfn=131072))=401da8d0 mm/page_alloc.c(1175): pg_data_map[0].node_start_pfn=131072 mm/page_alloc.c(1176): pg_data_map[0].node_present_pages=16384

---------------------------------------------------------------------------------

Mainly, as you have suspected initially, the issue is related the the _rambase that has to be 0x40000000 for mcf5441x.

Initially i tried by the m68k_fixup, setting an offset of 0x40000000, with something like

m68k_memoffset = 0x40000000;

but kernel in itial stages seems to need to access memory with logical addresses to be the same as physical, or it hangs at first memset.

So i just found the way to fix the issue tuning the node related pg_data pointer table, it is:

pg_data_t *pg_data_table[65];

This table was accessed at pg_data_table[0x200] for the first node, so the boot was crashing for an out of bounds.
So, since the slot calculation is done by the logical address, i subtracted there the _rambase (0x40000000) and the table is accessed properly now.

Linux version 4.12.0stmark2-001-11692-gc54cfe5cf339-dirty (angelo@jerusalem) (gcc version 5.2.0 (crosstools-sysam-2016.04.16)) #29 Sun Aug 27 02:18:49 CEST 2017 On node 0 totalpages: 16384
free_area_init_node: node 0, pgdat 401d88b0, node_mem_map 40204000
DMA zone: 72 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 16384 pages, LIFO batch:3
pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768

pcpu-alloc:
[0]
0

Built 1 zonelists in Zone order, mobility grouping on. Total pages: 16312
Kernel command line: console=ttyS0,115200 root=/dev/ram0 rw rootfstype=ramfs rdinit=/bin/init devtmpfs.mount=1 PID hash table entries: 512 (order: -2, 2048 bytes)
Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
Sorting __ex_table...
PID hash table entries: 512 (order: -2, 2048 bytes)
Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
Sorting __ex_table...
Memory: 128288K/131072K available (1219K kernel code, 103K rwdata, 288K rodata, 264K init, 79K bss, 2784K reserved, 0K cma-reserved) Virtual kernel memory layout:
vector : 0x40000000 - 0x40000400 ( 1 KiB)
kmap : 0xe0000000 - 0xf0000000 ( 256 MiB)
vmalloc : 0xd0000000 - 0xe0000000 ( 256 MiB)
lowmem : 0x40000000 - 0x48000000 ( 128 MiB)

  .init : 0x40196000 - 0x401d8000   ( 264 KiB)
  .text : 0x40001000 - 0x40131d50   (1220 KiB)
  .data : 0x40131d50 - 0x40193a00   ( 392 KiB)
  .bss  : 0x401d86e0 - 0x401ec674   (  80 KiB)

...

/ # cat /proc/meminfo

MemTotal:         128552 kB
MemFree:          124336 kB
MemAvailable:     122728 kB
Buffers:               0 kB
Cached:             1264 kB
SwapCached:            0 kB
Active:             1440 kB
Inactive:            888 kB
Active(anon):       1064 kB
Inactive(anon):      216 kB
Active(file):        376 kB
Inactive(file):      672 kB
Unevictable:           0 kB
Mlocked:               0 kB
SwapTotal:             0 kB
SwapFree:              0 kB
Dirty:                 0 kB
Writeback:             0 kB
AnonPages:          1080 kB
Mapped:              216 kB
Shmem:               216 kB
Slab:                  0 kB
SReclaimable:          0 kB
SUnreclaim:            0 kB
KernelStack:         144 kB
PageTables:          208 kB
NFS_Unstable:          0 kB
Bounce:                0 kB
WritebackTmp:          0 kB
CommitLimit:       64272 kB
Committed_AS:       1680 kB
VmallocTotal:     262144 kB
VmallocUsed:           0 kB
VmallocChunk:          0 kB

So for now i solved in this way:

arch/m68k/include/asm/page_mm.h +3 -1
index e7a1946455a8..4f3cb6218b8a 100644
@@ -142,7 +142,9 @@ static inline __attribute_const__ int __virt_to_node_shift(void)
        return shift;
 }
 
-#define __virt_to_node(addr)   (pg_data_table[(unsigned long)(addr) >> __virt_to_node_shift()])
+#define __virt_to_node(addr) \
+       pg_data_table[((unsigned long)(addr) - _rambase) \
+       >> __virt_to_node_shift()]
 #endif
 
 #define virt_to_page(addr) ({                                          \
arch/m68k/mm/init.c +11 -2
index a6ffead9bef5..61df1a1c8986 100644
@@ -61,15 +61,24 @@ void __init m68k_setup_node(int node)
 #ifndef CONFIG_SINGLE_MEMORY_CHUNK
        struct m68k_mem_info *info = m68k_memory + node;
        int i, end;
+       unsigned long addr_relative = info->addr - _rambase;
 
-       i = (unsigned long)phys_to_virt(info->addr) >> __virt_to_node_shift();
-       end = (unsigned long)phys_to_virt(info->addr + info->size - 1) >> __virt_to_node_shift();
+       i = (unsigned long)phys_to_virt(addr_relative) >> __virt_to_node_shift();
+       end = (unsigned long)phys_to_virt(addr_relative + info->size - 1) >> __virt_to_node_shift();
        for (; i <= end; i++) {
                if (pg_data_table[i])
                        pr_warn("overlap at %u for chunk %u\n", i, node);
                pg_data_table[i] = pg_data_map + node;
        }
 #endif
+       /*
+        * alloc_node_mem_map() in mm/page_alloc.c will setup
+        * node_mem_map member only if it is set to 0,
+        * otherwise it is considered already set properly
+        * before (i.e. as per ia64).
+        * So we need to zero node data here.
+        */
+       memset(NODE_DATA(node), 0, sizeof(pg_data_t));
        pg_data_map[node].bdata = bootmem_node_data + node;
        node_set_online(node);
 }
arch/m68k/mm/mcfmmu.c +2 -0
index c7efdf8e8eae..79af9a478f35 100644
@@ -176,6 +176,8 @@ void __init cf_bootmem_alloc(void)
        m68k_memory[0].addr = _rambase;
        m68k_memory[0].size = _ramend - _rambase;
 
+       m68k_memoffset = m68k_memory[0].addr - PAGE_OFFSET;
+
        /* compute total pages in system */
        num_pages = PFN_DOWN(_ramend - _rambase);
 

(m68k_memoffset = m68k_memory[0].addr - PAGE_OFFSET;) is there but not needed i guess so i would remove it.

This is not probably the best patch, let me know how do we proceed.

Quoted 4 lines

Regards
Greg

Regards,
Angelo

#18 lore raw

Hi Greg and all,

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.

I attach it for your comments.

Mainly what is not clear is what is the "m68k_fixup" needed. I am interested eventually to know where the idea of that fixup come from. The fixup seems should be module-related only, but actually is called from each __pa() and __va(), so i am bit conufsed, it seems also a kind of fixup (assembly replacement at runtime) that i don't find in other archs.

Let me know if you need some tests, i keep the system running with mmu enabled.
Will move as a next task into spi flash driver likely.

Regards,
Angelo

Quoted 496 lines · Angelo Dureghello

Hi Greg,

sorry for the late reply,
i decided to study inside the mmu code, working on a separate branch, and after a long hard fight i have the prompt.

So good news, seems i probably have mmu working. See below for further details.

Quoted 313 lines · Greg Ungerer

Hi Angelo,

Quoted 303 lines · Angelo Dureghello
Quoted 182 lines · Greg Ungerer
Quoted 122 lines · Angelo Dureghello
Quoted 121 lines · Greg Ungerer
Quoted 120 lines · Angelo Dureghello
Quoted 111 lines · Greg Ungerer
Quoted 42 lines · Angelo Dureghello
Quoted 22 lines · Greg Ungerer

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

Quoted 6 lines

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

Ok many thanks. Btw i used a __square(); function written in asm, so i am sure i see the gpio toggling in very early stages.

Quoted 5 lines

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

This MCU has sdram area physically mapped at 0x4000 0000 so U-Boot, to be able to execute the kernel must load it to that location/area anyway.

Quoted 7 lines

But i have seen that it is not a problem, after MMU is enabled in head.S the jump
movel #_vstart,%a0 /* jump to "virtual" space */
jmp %a0@

works fine. Since that range is not hitting anything that is maintained physical, it can be translated into virtual without any issue.

Yeah, it is not so much the initial start up that I think will be the problem. More the setup of the MMU mapping tables later in boot.

Quoted 19 lines

After some hard debug, i see the execution stops at:

asmlinkage __visible void __init start_kernel(void)

      ...
      setup_arch(&command_line);      setup_mm.c
         ...
         paging_init();               mm/mcfmmu.c
            ...
            empty_zero_page = (void *) alloc_bootmem_pages(PAGE_SIZE);
            ^line 47 mcfmmu.c

Inside alloc_bootmem_pages(), execution seems to end up finally to mm/bootmem.c and likely to alloc_bootmem_bdata().
In case i can still proceed to find the exact place where execution stops, but i suspect in the while(1), line 545.

As a curious thing, i find in a different cf CPU code "m54xx.c" the following:

Quoted 7 lines

void __init config_BSP(char *commandp, int size)
{

#ifdef CONFIG_MMU
       cf_bootmem_alloc();
       mmu_context_init();

#endif
Do also m5441x.c maybe need this calls ?

Yes, you will need this. So that code above is only getting run when configured for a 547x CPU family. Attached is a rework of that code so that it will be run for all ColdFire MMU varients. Can you try that out?

Quoted 2 lines

Would be very nice to have MMU working. Strangely, i don't see any board_config with it enabled. Was it ever tested on some Coldfire ?

Oh, yeah, I run this on a real M5475 EVB board for every kernel mainline release, with and without MMU enabled. See the arch/m68k/configs/m5475evb_defconfig, it will default to having the MMU enabled.

I have todays linux-4.13-rc5 running on it here now:

# cat /proc/version
Linux version 4.13.0-rc5-00001-gb014090-dirty (gerg@goober) (gcc version 5.4.0 (GCC)) #1 Mon Aug 14 10:14:12 AEST 2017
# cat /proc/cpuinfo
CPU:            ColdFire
MMU:            ColdFire
FPU:            ColdFire
Clocking:       264.1MHz
BogoMips:       264.19
Calibration:    1320960 loops
#
Quoted 2 lines

Regards
Greg

Ok, i applied your patch, and still the kernel is hanging silently, so i started up a new debug session again.

What is actually happening (after your patch has been applied) is:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

setup_arch() arch/m68k/kernel/setup_mm.c
paging_init()
memmap_init() mm/page_alloc.c
memmap_init_zone()

Quoted 8 lines
Quoted 8 lines
    __init_single_page()
        set_page_links()       include/linux/mm.h
           set_page_zone()
             kernel hangs silently on this line
             page->flags &= ~(ZONES_MASK << ZONES_PGSHIFT);
Quoted 1 line

Can you run your current code with the console debug code I sent a little while back?

I ask because I suspect it should give something based on your debug above. I played around a little trying to fake out my configuration to make it look like the RAM was non-zero based. I couldn't get a fail, but I would like to add some more debug to see what is going on with the page pointers from your debug.

Can you apply the attached patch and get any extra debug?

Quoted 34 lines
Quoted 22 lines
Quoted 8 lines

I am wondering how mmu works, so at the moment mmu is enabled, in head.S, i would expect that code compiled for 0x40001000 would not run, since jumps would be translated to some different physical addresses, but execution sill works.
At the same, after enabling mmu i would expect .data vars to be invalid, since their address would be translated to a different location, while not, the init values of .data variables are still valid. In case, i am interested to understand this points.

On the ColdFire the kernel relies on all RAM and IO peripheral addresses) to "hit" the ACR registers - and essentially be passed through as an identity physical = virtual mapping. If you look at the operation of the memory address translation when virtual mode is enabled (in the ColdFire MMU sections of the 5475 and 54411 reference manual) you will see that addresses are checked in order to be for the MMUBAR, RAMBAR, ACR, then MMU.

For example a kernel address when in supervisor mode will hit ACR1 or ACR3 the way we set them up in arch/m68k/coldfire/head.S. And that is why you see kernel code and data still being valid after the MMU is enabled in virtual mode. No TLB entries required for this.

Quoted 5 lines

Looking at your call sequence above I can see that the physical RAM start address being non-zero is going to come into play. I'll dig into this a little more tomorrow see if I can figure out what is going on.

Thanks for the kind clarifications.

I'll look in this things too in next days, learning is always nice. Btw, about load/entry address, i have noticed a possible basic difference betweeen mcf5441x and mcf547x series:

Quoted 8 lines
Quoted 3 lines

The second one (your cpu) is v4e and probably more recent i guess, and one major difference from datasheet seems to be that it is Harvard. So probably, for this reason, you can address ram from 0 there.

IIRC the 5475 was the first ColdFire with MMU, it is pretty old. Pretty sure the 54411 came later. Not sure what the thinking was on the different default memory layout though.

Finally, cleaning out my debug lines, i found i removed an important line. So i am back to original "second" error we was trying to understand.

So current more clear status is:

U-Boot 2017.09-rc2-00151-g2d7cb5b426-dirty (Aug 22 2017 - 00:22:46 +0200)

CPU: Freescale MCF54410 (Mask:9f Version:2)
CPU CLK 240 MHz BUS CLK 120 MHz FLB CLK 60 MHz INP CLK 30 MHz VCO CLK 480 MHz

SPI:   ready
DRAM:  128 MiB
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
SF: Detected is25lp128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
device 0 offset 0x100000, size 0x1d9728
SF: 1939240 bytes @ 0x100000 Read: OK
## Booting kernel from Legacy Image at 40001000 ...
    Image Name:   mainline kernel
    Created:      2017-08-22   0:07:25 UTC
    Image Type:   M68K Linux Kernel Image (uncompressed)
    Data Size:    1939176 Bytes = 1.8 MiB

Load Address: 40001000
Entry Point: 40001000
Verifying Checksum ... OK
Loading Kernel Image ... OK
Linux version 4.12.0stmark2-001-11691-g571d81b2b55f-dirty (angelo@jerusalem) (gcc version 4.9.0 (crosstools-sysam-2016.04.16)) #182 Tue Aug 22 02:07:24 CEST 2017

------------[ cut here ]------------
WARNING: CPU: 0 PID: 0 at mm/page_alloc.c:6219 free_area_init_node+0x2f4/0x2fa
CPU: 0 PID: 0 Comm: swapper Not tainted 4.12.0stmark2-001-11691-g571d81b2b55f-dirty #182

Stack from 4017deec:

         4017deec

4017b3dd
40007972
00000000
00000000
47d9f62c
00020000
00000000

         00000000

4017df9c
40007a14
4016dd8e
0000184b
4019caca
00000009
00000000

         00000000

4019caca
4016dd8e
0000184b
48000000
40204000
47d9f62c
40001000

         00000000

47d9ef1c
40001480
4013010c
4012cd16
4017dfa8
4019ecc0
00012000

         00002000

4019ccb4
00000000
4017df9c
00020000
00000000
4019a3f2
4017df9c

         00000001

401da8c0
401da774
4019ebc8
00004000
00000000
00000000
4017dfc8

Call Trace:
  [<40007972>] __warn+0xa4/0xc0
  [<40007a14>] warn_slowpath_null+0x1a/0x22
  [<4019caca>] free_area_init_node+0x2f4/0x2fa
  [<4019caca>] free_area_init_node+0x2f4/0x2fa
  [<40001000>] kernel_pg_dir+0x0/0x1000
  [<40001480>] kernel_pg_dir+0x480/0x1000
  [<4013010c>] memset+0x0/0x80
  [<4012cd16>] strlen+0x0/0x14
  [<4019ecc0>] __alloc_bootmem+0x16/0x3c
  [<4019ccb4>] free_area_init+0x20/0x26
  [<4019a3f2>] paging_init+0xee/0xfa
  [<4019ebc8>] free_bootmem_node+0x0/0x34
  [<40199fbc>] setup_arch+0xcc/0x16e
  [<40024eb2>] printk+0x0/0x18
  [<4019ecaa>] __alloc_bootmem+0x0/0x3c
  [<40198550>] start_kernel+0x68/0x3ae
  [<40001000>] kernel_pg_dir+0x0/0x1000
  [<400020f2>] _exit+0x0/0x6

---[ end trace 0000000000000000 ]---
Quoted 6 lines

On node 0 totalpages: 16384
free_area_init_node: node 0, pgdat 401da8c0, node_mem_map a8c0401d
DMA zone: 72 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 16384 pages, LIFO batch:3
/page_alloc.c(1171): page=a8c0401d pfn=131072

Quoted 4 lines

Another patch attached that digs a little deeper into why that page pointer ends up being invalid. If you could run with this and send the output that would be great.

First of all, this is the trace you required me:

mm/page_alloc.c(1171): page=a8d0401d pfn=131072 mm/page_alloc.c(1172): __va(pfn=131072)=00020000 mm/page_alloc.c(1173): pfn_to_virt(pfn=131072)=40000000 mm/page_alloc.c(1174): __virt_to_node(pfn_to_virt(pfn=131072))=401da8d0 mm/page_alloc.c(1175): pg_data_map[0].node_start_pfn=131072 mm/page_alloc.c(1176): pg_data_map[0].node_present_pages=16384

---------------------------------------------------------------------------------

Mainly, as you have suspected initially, the issue is related the the _rambase that has to be 0x40000000 for mcf5441x.

Initially i tried by the m68k_fixup, setting an offset of 0x40000000, with something like

m68k_memoffset = 0x40000000;

but kernel in itial stages seems to need to access memory with logical addresses to be the same as physical, or it hangs at first memset.

So i just found the way to fix the issue tuning the node related pg_data pointer table, it is:

pg_data_t *pg_data_table[65];
This table was accessed at pg_data_table[0x200] for the first node, so the boot was crashing for an out of bounds.
So, since the slot calculation is done by the logical address, i subtracted there the _rambase (0x40000000) and the table is accessed properly now.

Linux version 4.12.0stmark2-001-11692-gc54cfe5cf339-dirty (angelo@jerusalem) (gcc version 5.2.0 (crosstools-sysam-2016.04.16)) #29 Sun Aug 27 02:18:49 CEST 2017 On node 0 totalpages: 16384
free_area_init_node: node 0, pgdat 401d88b0, node_mem_map 40204000
DMA zone: 72 pages used for memmap
DMA zone: 0 pages reserved
DMA zone: 16384 pages, LIFO batch:3
pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768

pcpu-alloc:
[0]
0

Built 1 zonelists in Zone order, mobility grouping on. Total pages: 16312
Kernel command line: console=ttyS0,115200 root=/dev/ram0 rw rootfstype=ramfs rdinit=/bin/init devtmpfs.mount=1 PID hash table entries: 512 (order: -2, 2048 bytes)
Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
Sorting __ex_table...
PID hash table entries: 512 (order: -2, 2048 bytes)
Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
Sorting __ex_table...
Memory: 128288K/131072K available (1219K kernel code, 103K rwdata, 288K rodata, 264K init, 79K bss, 2784K reserved, 0K cma-reserved) Virtual kernel memory layout:
vector : 0x40000000 - 0x40000400 ( 1 KiB)
kmap : 0xe0000000 - 0xf0000000 ( 256 MiB)
vmalloc : 0xd0000000 - 0xe0000000 ( 256 MiB)
lowmem : 0x40000000 - 0x48000000 ( 128 MiB)

   .init : 0x40196000 - 0x401d8000   ( 264 KiB)
   .text : 0x40001000 - 0x40131d50   (1220 KiB)
   .data : 0x40131d50 - 0x40193a00   ( 392 KiB)
   .bss  : 0x401d86e0 - 0x401ec674   (  80 KiB)

...

/ # cat /proc/meminfo

MemTotal:         128552 kB
MemFree:          124336 kB
MemAvailable:     122728 kB
Buffers:               0 kB
Cached:             1264 kB
SwapCached:            0 kB
Active:             1440 kB
Inactive:            888 kB
Active(anon):       1064 kB
Inactive(anon):      216 kB
Active(file):        376 kB
Inactive(file):      672 kB
Unevictable:           0 kB
Mlocked:               0 kB
SwapTotal:             0 kB
SwapFree:              0 kB
Dirty:                 0 kB
Writeback:             0 kB
AnonPages:          1080 kB
Mapped:              216 kB
Shmem:               216 kB
Slab:                  0 kB
SReclaimable:          0 kB
SUnreclaim:            0 kB
KernelStack:         144 kB
PageTables:          208 kB
NFS_Unstable:          0 kB
Bounce:                0 kB
WritebackTmp:          0 kB
CommitLimit:       64272 kB
Committed_AS:       1680 kB
VmallocTotal:     262144 kB
VmallocUsed:           0 kB
VmallocChunk:          0 kB

So for now i solved in this way:

arch/m68k/include/asm/page_mm.h +3 -1
index e7a1946455a8..4f3cb6218b8a 100644
@@ -142,7 +142,9 @@ static inline __attribute_const__ int __virt_to_node_shift(void)
         return shift;
  }
 
-#define __virt_to_node(addr)   (pg_data_table[(unsigned long)(addr) >> __virt_to_node_shift()])
+#define __virt_to_node(addr) \
+       pg_data_table[((unsigned long)(addr) - _rambase) \
+       >> __virt_to_node_shift()]
  #endif
 
  #define virt_to_page(addr) ({                                          \
arch/m68k/mm/init.c +11 -2
index a6ffead9bef5..61df1a1c8986 100644
@@ -61,15 +61,24 @@ void __init m68k_setup_node(int node)
  #ifndef CONFIG_SINGLE_MEMORY_CHUNK
         struct m68k_mem_info *info = m68k_memory + node;
         int i, end;
+       unsigned long addr_relative = info->addr - _rambase;
 
-       i = (unsigned long)phys_to_virt(info->addr) >> __virt_to_node_shift();
-       end = (unsigned long)phys_to_virt(info->addr + info->size - 1) >> __virt_to_node_shift();
+       i = (unsigned long)phys_to_virt(addr_relative) >> __virt_to_node_shift();
+       end = (unsigned long)phys_to_virt(addr_relative + info->size - 1) >> __virt_to_node_shift();
         for (; i <= end; i++) {
                 if (pg_data_table[i])
                         pr_warn("overlap at %u for chunk %u\n", i, node);
                 pg_data_table[i] = pg_data_map + node;
         }
  #endif
+       /*
+        * alloc_node_mem_map() in mm/page_alloc.c will setup
+        * node_mem_map member only if it is set to 0,
+        * otherwise it is considered already set properly
+        * before (i.e. as per ia64).
+        * So we need to zero node data here.
+        */
+       memset(NODE_DATA(node), 0, sizeof(pg_data_t));
         pg_data_map[node].bdata = bootmem_node_data + node;
         node_set_online(node);
  }
arch/m68k/mm/mcfmmu.c +2 -0
index c7efdf8e8eae..79af9a478f35 100644
@@ -176,6 +176,8 @@ void __init cf_bootmem_alloc(void)
         m68k_memory[0].addr = _rambase;
         m68k_memory[0].size = _ramend - _rambase;
 
+       m68k_memoffset = m68k_memory[0].addr - PAGE_OFFSET;
+
         /* compute total pages in system */
         num_pages = PFN_DOWN(_ramend - _rambase);
 

(m68k_memoffset = m68k_memory[0].addr - PAGE_OFFSET;) is there but not needed i guess so i would remove it.

This is not probably the best patch, let me know how do we proceed.

Quoted 7 lines
Quoted 4 lines

Regards
Greg

Regards,
Angelo
From 9f0e06f9566de4d91a234753cd30160ce02fc8a8 Mon Sep 17 00:00:00 2001

From: Angelo Dureghello <[email protected]>
Date: Thu, 24 Aug 2017 00:01:06 +0200
Subject: [PATCH 1/2] Patch to enable mmu, from Greg.

---
 arch/m68k/Kconfig.cpu               |  2 +-
 arch/m68k/coldfire/m54xx.c          |  4 ----
 arch/m68k/configs/stmark2_defconfig |  5 +++--
 arch/m68k/include/asm/mmu_context.h |  1 -
 arch/m68k/kernel/setup_mm.c         |  1 +
 arch/m68k/mm/mcfmmu.c               | 35 ++++++++++++++++++-----------------
 6 files changed, 23 insertions(+), 25 deletions(-)
arch/m68k/Kconfig.cpu +1 -1
index d2219f30b78f..4dc51c090a45 100644
@@ -283,7 +283,7 @@ config M548x
 
 config M5441x
 	bool "MCF5441x"
-	depends on !MMU
+	select MMU_COLDFIRE if MMU
 	select GENERIC_CLOCKEVENTS
 	select HAVE_CACHE_CB
 	help
arch/m68k/coldfire/m54xx.c +0 -4
index c552851ec617..704efeaeab2d 100644
@@ -95,10 +95,6 @@ static void mcf54xx_reset(void)
 
 void __init config_BSP(char *commandp, int size)
 {
-#ifdef CONFIG_MMU
-	cf_bootmem_alloc();
-	mmu_context_init();
-#endif
 	mach_reset = mcf54xx_reset;
 	mach_sched_init = hw_timer_init;
 	m54xx_uarts_init();
arch/m68k/configs/stmark2_defconfig +3 -2
index ea78b8f6a68e..b886f303cc3f 100644
@@ -21,7 +21,7 @@ CONFIG_EMBEDDED=y
 # CONFIG_LBDAF is not set
 # CONFIG_BLK_DEV_BSG is not set
 CONFIG_BLK_CMDLINE_PARSER=y
-# CONFIG_MMU is not set
+CONFIG_COLDFIRE=y
 CONFIG_M5441x=y
 CONFIG_CLOCK_FREQ=240000000
 CONFIG_STMARK2=y
@@ -30,6 +30,8 @@ CONFIG_RAMSIZE=0x8000000
 CONFIG_VECTORBASE=0x40000000
 CONFIG_KERNELBASE=0x40001000
 CONFIG_BINFMT_FLAT=y
+CONFIG_BINFMT_ZFLAT=y
+CONFIG_BINFMT_SHARED_FLAT=y
 CONFIG_BINFMT_MISC=y
 # CONFIG_UEVENT_HELPER is not set
 CONFIG_DEVTMPFS=y
@@ -53,7 +55,6 @@ CONFIG_MTD_COMPLEX_MAPPINGS=y
 CONFIG_MTD_PHYSMAP=y
 CONFIG_MTD_PLATRAM=y
 CONFIG_MTD_SPI_NOR=y
-# CONFIG_INPUT_MOUSEDEV is not set
 # CONFIG_INPUT_KEYBOARD is not set
 # CONFIG_INPUT_MOUSE is not set
 CONFIG_SERIO_LIBPS2=y
arch/m68k/include/asm/mmu_context.h +0 -1
index 4a6ae6dffa34..00c28b1dc47b 100644
@@ -91,7 +91,6 @@ static inline void activate_mm(struct mm_struct *active_mm,
 
 #define deactivate_mm(tsk, mm) do { } while (0)
 
-extern void mmu_context_init(void);
 #define prepare_arch_switch(next) load_ksp_mmu(next)
 
 static inline void load_ksp_mmu(struct task_struct *task)
arch/m68k/kernel/setup_mm.c +1 -0
index 7a2c21212820..5632c48b9f3b 100644
@@ -343,6 +343,7 @@ void __init setup_arch(char **cmdline_p)
 #ifdef CONFIG_COLDFIRE
 	case MACH_M54XX:
 	case MACH_M5441X:
+		cf_bootmem_alloc();
 		config_BSP(NULL, 0);
 		break;
 #endif
arch/m68k/mm/mcfmmu.c +18 -17
index 87131cd3bc8f..c7efdf8e8eae 100644
@@ -150,6 +150,23 @@ int cf_tlb_miss(struct pt_regs *regs, int write, int dtlb, int extension_word)
 	return 0;
 }
 
+/*
+ * Initialize the context management stuff.
+ * The following was taken from arch/ppc/mmu_context.c
+ */
+static void __init mmu_context_init(void)
+{
+	/*
+	 * Some processors have too few contexts to reserve one for
+	 * init_mm, and require using context 0 for a normal task.
+	 * Other processors reserve the use of context zero for the kernel.
+	 * This code assumes FIRST_CONTEXT < 32.
+	 */
+	context_map[0] = (1 << FIRST_CONTEXT) - 1;
+	next_mmu_context = FIRST_CONTEXT;
+	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+}
+
 void __init cf_bootmem_alloc(void)
 {
 	unsigned long start_pfn;
@@ -177,23 +194,7 @@ void __init cf_bootmem_alloc(void)
 	memstart += init_bootmem_node(NODE_DATA(0), start_pfn,
 		min_low_pfn, max_low_pfn);
 	free_bootmem_node(NODE_DATA(0), memstart, _ramend - memstart);
-}
-
-/*
- * Initialize the context management stuff.
- * The following was taken from arch/ppc/mmu_context.c
- */
-void __init mmu_context_init(void)
-{
-	/*
-	 * Some processors have too few contexts to reserve one for
-	 * init_mm, and require using context 0 for a normal task.
-	 * Other processors reserve the use of context zero for the kernel.
-	 * This code assumes FIRST_CONTEXT < 32.
-	 */
-	context_map[0] = (1 << FIRST_CONTEXT) - 1;
-	next_mmu_context = FIRST_CONTEXT;
-	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+	mmu_context_init();
 }
 
 /*
Signature · 1 line
2.14.1
Quoted 1 line

From 96e88773601fff1edd61cec88095c60252be1343 Mon Sep 17 00:00:00 2001

From: Angelo Dureghello <[email protected]>
Date: Sun, 27 Aug 2017 02:42:42 +0200
Subject: [PATCH 2/2] m68k: fix mmu for coldfire mcf5441x

This patch fixes mmu not working for CPU's that has the base of the physical memory mapped at a non-zero address.

---
 arch/m68k/include/asm/page_mm.h |  4 +++-
 arch/m68k/mm/init.c             | 14 ++++++++++++--
 2 files changed, 15 insertions(+), 3 deletions(-)
arch/m68k/include/asm/page_mm.h +3 -1
index e7a1946455a8..6c3b8c5181f9 100644
@@ -142,7 +142,9 @@ static inline __attribute_const__ int __virt_to_node_shift(void)
 	return shift;
 }
 
-#define __virt_to_node(addr)	(pg_data_table[(unsigned long)(addr) >> __virt_to_node_shift()])
+#define __virt_to_node(addr) \
+	pg_data_table[((unsigned long)(addr) - PAGE_OFFSET) \
+	>> __virt_to_node_shift()]
 #endif
 
 #define virt_to_page(addr) ({						\
arch/m68k/mm/init.c +12 -2
index a6ffead9bef5..17254e6c3f9a 100644
@@ -62,14 +62,24 @@ void __init m68k_setup_node(int node)
 	struct m68k_mem_info *info = m68k_memory + node;
 	int i, end;
 
-	i = (unsigned long)phys_to_virt(info->addr) >> __virt_to_node_shift();
-	end = (unsigned long)phys_to_virt(info->addr + info->size - 1) >> __virt_to_node_shift();
+	i = (unsigned long)phys_to_virt(info->addr - PAGE_OFFSET)
+		>> __virt_to_node_shift();
+	end = ((unsigned long)phys_to_virt(info->addr - PAGE_OFFSET)
+		+ info->size - 1) >> __virt_to_node_shift();
 	for (; i <= end; i++) {
 		if (pg_data_table[i])
 			pr_warn("overlap at %u for chunk %u\n", i, node);
 		pg_data_table[i] = pg_data_map + node;
 	}
 #endif
+	/*
+	 * alloc_node_mem_map() in mm/page_alloc.c will setup
+	 * node_mem_map member only if it is set to 0,
+	 * otherwise it is considered already set properly
+	 * before (i.e. as per ia64).
+	 * So we need to zero node data here.
+	 */
+	memset(NODE_DATA(node), 0, sizeof(pg_data_t));
 	pg_data_map[node].bdata = bootmem_node_data + node;
 	node_set_online(node);
 }
Signature · 1 line
2.14.1
#19 lore raw

Hi Angelo,

Quoted 8 lines · Angelo Dureghello

Hi Greg and all,

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

    #if defined(CONFIG_RAMBASE)
    #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
    #elif defined(CONFIG_SUN3)
    #define PAGE_OFFSET_RAW         0x0E000000
    #else
    #define PAGE_OFFSET_RAW         0x00000000
    #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero, and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +1 -1
@@ -1,6 +1,6 @@
 /* This handles the memory map.. */
 
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
 #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
 #elif defined(CONFIG_SUN3)
 #define PAGE_OFFSET_RAW                0x0E000000

or

arch/m68k/Kconfig.machine +2 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
 config RAMBASE
        hex "Address of the base of RAM"
+       depends on MMU
        default "0"
        help
          Define the address that RAM starts at. On many platforms this is
@@ -334,6 +335,7 @@ config RAMBASE
 
 config RAMSIZE
        hex "Size of RAM (in bytes), or 0 for automatic"
+       depends on MMU
        default "0x400000"
        help
          Define the size of the system RAM. If you select 0 then the

depending on whether anything else in the Coldfire code needs RAMBASE.

Quoted 6 lines

Mainly what is not clear is what is the "m68k_fixup" needed. I am interested eventually to know where the idea of that fixup come from. The fixup seems should be module-related only, but actually is called from each __pa() and __va(), so i am bit conufsed, it seems also a kind of fixup (assembly replacement at runtime) that i don't find in other archs.

commit fbe9c9612930e0604dc99ef2da7e063fa3278817

Author: Roman Zippel <[email protected]>
Date:   Thu May 31 00:40:50 2007 -0700

m68k: runtime patching infrastructure

Add the basic infrastructure to allow runtime patching of kernel and modules to optimize a few functions with parameters, which are only calculated once during bootup and are otherwise constant. Use this for the
conversion between
virtual and physical addresses.

So it's a micro-optimization for virt/phys conversions, to avoid the extra load of the offset to add/subtract.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

#20 lore raw

Hi Angelo, Geert,

Quoted 59 lines · Geert Uytterhoeven

Hi Angelo,

Quoted 8 lines · Angelo Dureghello

Hi Greg and all,

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

     #if defined(CONFIG_RAMBASE)
     #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
     #elif defined(CONFIG_SUN3)
     #define PAGE_OFFSET_RAW         0x0E000000
     #else
     #define PAGE_OFFSET_RAW         0x00000000
     #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero, and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +1 -1
@@ -1,6 +1,6 @@
  /* This handles the memory map.. */
 
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
  #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
  #elif defined(CONFIG_SUN3)
  #define PAGE_OFFSET_RAW                0x0E000000

or

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 4 lines
arch/m68k/Kconfig.machine +1 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
  config RAMBASE
         hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Regards
Greg

Quoted 51 lines
arch/m68k/Kconfig.machine +1 -0
         default "0"
         help
           Define the address that RAM starts at. On many platforms this is
@@ -334,6 +335,7 @@ config RAMBASE
 
  config RAMSIZE
         hex "Size of RAM (in bytes), or 0 for automatic"
+       depends on MMU
         default "0x400000"
         help
           Define the size of the system RAM. If you select 0 then the

depending on whether anything else in the Coldfire code needs RAMBASE.

Quoted 6 lines

Mainly what is not clear is what is the "m68k_fixup" needed. I am interested eventually to know where the idea of that fixup come from. The fixup seems should be module-related only, but actually is called from each __pa() and __va(), so i am bit conufsed, it seems also a kind of fixup (assembly replacement at runtime) that i don't find in other archs.

commit fbe9c9612930e0604dc99ef2da7e063fa3278817

Author: Roman Zippel <[email protected]>
Date:   Thu May 31 00:40:50 2007 -0700

m68k: runtime patching infrastructure

Add the basic infrastructure to allow runtime patching of kernel and modules to optimize a few functions with parameters, which are only calculated once during bootup and are otherwise constant. Use this for the
conversion between
virtual and physical addresses.

So it's a micro-optimization for virt/phys conversions, to avoid the extra load of the offset to add/subtract.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
--
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#21 lore raw

Hi Greg,

Quoted 60 lines · Greg Ungerer
Quoted 59 lines · Geert Uytterhoeven
Quoted 6 lines · Angelo Dureghello

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

     #if defined(CONFIG_RAMBASE)
     #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
     #elif defined(CONFIG_SUN3)
     #define PAGE_OFFSET_RAW         0x0E000000
     #else
     #define PAGE_OFFSET_RAW         0x00000000
     #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero,
and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +1 -1
@@ -1,6 +1,6 @@
  /* This handles the memory map.. */
 
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
  #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
  #elif defined(CONFIG_SUN3)
  #define PAGE_OFFSET_RAW                0x0E000000

or

arch/m68k/Kconfig.machine +0 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
Quoted 6 lines

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 3 lines
arch/m68k/Kconfig.machine +1 -0
  config RAMBASE
         hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

Sorry, yes, depends on !MMU.

Quoted 6 lines
Quoted 1 line

depending on whether anything else in the Coldfire code needs RAMBASE.

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

OK, as I already feared/expected...

Quoted 2 lines

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Right.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

#22 lore raw

Hi Greg and Geert,

about my patch sent yesterday, unfortunately, i found btw at least 1 issue: the busybox "cd" doesn't change directory :)

/ # cat /proc/vmallocinfo
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc
/ # ls

bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var
/ # cd bin
/ # ls
bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var

/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vmallocinfo
0xd0000000-0xd0006000 24576 unpurged vm_area
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc

And after each "cd" attempt, another "unpurged" message is added. Removing MMU cd works as expected.

Quoted 51 lines · Geert Uytterhoeven

Hi Greg,

Quoted 48 lines · Greg Ungerer
Quoted 47 lines · Geert Uytterhoeven
Quoted 6 lines · Angelo Dureghello

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

      #if defined(CONFIG_RAMBASE)
      #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
      #elif defined(CONFIG_SUN3)
      #define PAGE_OFFSET_RAW         0x0E000000
      #else
      #define PAGE_OFFSET_RAW         0x00000000
      #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero,
and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +0 -0
@@ -1,6 +1,6 @@
   /* This handles the memory map.. */
 
Quoted 6 lines
Quoted 6 lines

arch/m68k/include/asm/page_offset.h line 6

Quoted 6 lines
arch/m68k/include/asm/page_offset.h +1 -1
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
   #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
   #elif defined(CONFIG_SUN3)
   #define PAGE_OFFSET_RAW                0x0E000000

I tested this patch, (removed mine) and kernel hangs somewhere silently at first init, i don't have the debug enabled now, but i suspect it still hangs at some initial "memset" since 0 area for kernel should be allocated before access.

Isn't PAGE_OFFSET the starting area for the virtual kernel address space ? At least so it seems to be for the famous 3G + 1G of x86.

This is what i see at boot with my patch of yesterday:

[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131e70   (1220 KiB)
[    0.000000]       .data : 0x40131e70 - 0x40193900   ( 391 KiB)
[    0.000000]       .bss  : 0x401d86d8 - 0x401ec680   (  80 KiB)
Quoted 36 lines
Quoted 6 lines
Quoted 6 lines

or

arch/m68k/Kconfig.machine +0 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
Quoted 6 lines

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 3 lines
arch/m68k/Kconfig.machine +1 -0
   config RAMBASE
          hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

Sorry, yes, depends on !MMU.

Quoted 6 lines
Quoted 1 line

depending on whether anything else in the Coldfire code needs RAMBASE.

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

OK, as I already feared/expected...

Quoted 2 lines

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Right.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

Quoted 8 lines

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
--
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Regards,
Angelo

#23 lore raw

Hi Angelo,

I have attached a patch with a slightly different approach to fixing this. Can you try this one out?

I have a feel for what is going on, based in particular on the changes you made in 0002-m68k-fix-mmu-for-coldfire-mcf5441x.patch. I see that the virt_node_shift and accessing pg_data_map when not 0 based is going to be problematic with the code as it is.

Quoted 20 lines · Angelo Dureghello

about my patch sent yesterday, unfortunately, i found btw at least 1 issue: the busybox "cd" doesn't change directory :)

/ # cat /proc/vmallocinfo
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc
/ # ls

bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var
/ # cd bin
/ # ls
bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var

/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vmallocinfo
0xd0000000-0xd0006000 24576 unpurged vm_area
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc

Quoted 2 lines

And after each "cd" attempt, another "unpurged" message is added. Removing MMU cd works as expected.

I setup a configuration with my 5475 where I moved the build RAM base to be 32MB - so it was no longer 0 based. Then I could run some tests to at least simulate what you have on the 54411. (The only catch is my code could still access 0 and surrounding memory addresses without faulting...)

Running with my attached patch I didn't see any odd behavior with cd/ls like you see above.

Geert: interested if you have any thoughts on problems around
virt_to_node_shift. Changing CONFIG_RAMBASE I don't think is going to resolve the problem in m68k_setup_node(). We access of the end of the array since it was divided up based on the size of the RAM, but we access it using an index derrived directly from the absolute address.

Quoted 64 lines · Geert Uytterhoeven
Quoted 51 lines

Hi Greg,

Quoted 48 lines · Greg Ungerer
Quoted 47 lines · Geert Uytterhoeven
Quoted 6 lines · Angelo Dureghello

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

      #if defined(CONFIG_RAMBASE)
      #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
      #elif defined(CONFIG_SUN3)
      #define PAGE_OFFSET_RAW         0x0E000000
      #else
      #define PAGE_OFFSET_RAW         0x00000000
      #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero,
and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +0 -0
@@ -1,6 +1,6 @@
   /* This handles the memory map.. */
 
Quoted 6 lines
Quoted 6 lines

arch/m68k/include/asm/page_offset.h line 6

Quoted 6 lines
arch/m68k/include/asm/page_offset.h +1 -1
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
   #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
   #elif defined(CONFIG_SUN3)
   #define PAGE_OFFSET_RAW                0x0E000000

I tested this patch, (removed mine) and kernel hangs somewhere silently at first init, i don't have the debug enabled now, but i suspect it still hangs at some initial "memset" since 0 area for kernel should be allocated before access.

Quoted 2 lines

Isn't PAGE_OFFSET the starting area for the virtual kernel address space ? At least so it seems to be for the famous 3G + 1G of x86.

Well, for the current working ColdFire with MMU that is the case. The kernel virtual addresses start at 0...

Quoted 11 lines

This is what i see at boot with my patch of yesterday:

[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131e70   (1220 KiB)
[    0.000000]       .data : 0x40131e70 - 0x40193900   ( 391 KiB)
[    0.000000]       .bss  : 0x401d86d8 - 0x401ec680   (  80 KiB)

For whatever it is worth this is what I see now on my debug setup:

Virtual kernel memory layout:

    vector  : 0x02000000 - 0x02000400   (   1 KiB)
    kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
    vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
    lowmem  : 0x02000000 - 0x04000000   (  32 MiB)
      .init : 0x02288000 - 0x02296000   (  56 KiB)
      .text : 0x02020000 - 0x0221f270   (2045 KiB)
      .data : 0x0221f270 - 0x02284140   ( 404 KiB)
      .bss  : 0x022967a0 - 0x022ae60c   (  96 KiB)

Regards
Greg

Quoted 45 lines
Quoted 24 lines
Quoted 6 lines
Quoted 6 lines

or

arch/m68k/Kconfig.machine +0 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
Quoted 6 lines

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 3 lines
arch/m68k/Kconfig.machine +1 -0
   config RAMBASE
          hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

Sorry, yes, depends on !MMU.

Quoted 6 lines
Quoted 1 line

depending on whether anything else in the Coldfire code needs RAMBASE.

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

OK, as I already feared/expected...

Quoted 20 lines
Quoted 2 lines

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Right.

Gr{oetje,eeting}s,

Geert

Signature · 9 lines
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                 -- Linus Torvalds
-- 
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Quoted 3 lines

Regards,
Angelo

arch/m68k/mm/mcfmmu.c +1 -1
index 87131cd..6e63d24 100644
@@ -169,7 +187,7 @@ void __init cf_bootmem_alloc(void)
 	max_pfn = max_low_pfn = PFN_DOWN(_ramend);
 	high_memory = (void *)_ramend;
 
-	m68k_virt_to_node_shift = fls(_ramend - _rambase - 1) - 6;
+	m68k_virt_to_node_shift = fls(_ramend - 1) - 6;
 	module_fixup(NULL, __start_fixup, __stop_fixup);
 
 	/* setup bootmem data */
#24 lore raw

Hi Greg,

Quoted 11 lines · Greg Ungerer

Hi Angelo,

I have attached a patch with a slightly different approach to fixing this. Can you try this one out?

I have a feel for what is going on, based in particular on the changes you made in 0002-m68k-fix-mmu-for-coldfire-mcf5441x.patch. I see that the virt_node_shift and accessing pg_data_map when not 0 based is going to be problematic with the code as it is.

Great catch, the patch works,
i maintaind btw the

memset(NODE_DATA(node), 0, sizeof(pg_data_t));

inside void __init m68k_setup_node(int node) since otherwise there was that warning we've seen initially.

About the "cd" issue, it seems to be a hush issue, maybe becouse hush is built as nommu. Re-executing hush, i can now cd to a subfolder, but i get then a SEGV on "cd ..".

This the boot log:

## Booting kernel from Legacy Image at 40001000 ...

   Image Name:   mainline kernel
   Created:      2017-09-04  14:19:47 UTC
   Image Type:   M68K Linux Kernel Image (uncompressed)
   Data Size:    1930976 Bytes = 1.8 MiB
   Load Address: 40001000
   Entry Point:  40001000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK
[    0.000000] Linux version 4.13.0-rc7stmark2-001-00039-g96e88773601f-dirty (angelo@jerusalem) (gcc version 5.2.0 (crosstools-sysam-2016.04.16)) #56 Mon Sep 4 16:19:46 CEST 2017
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 16312
[    0.000000] Kernel command line: console=ttyS0,115200 root=/dev/ram0 rw rootfstype=ramfs rdinit=/bin/init devtmpfs.mount=1
[    0.000000] PID hash table entries: 512 (order: -2, 2048 bytes)
[    0.000000] Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
[    0.000000] Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
[    0.000000] Sorting __ex_table...
[    0.000000] Memory: 128288K/131072K available (1219K kernel code, 103K rwdata, 288K rodata, 264K init, 79K bss, 2784K reserved, 0K cma-reserved)
[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131cb0   (1220 KiB)
[    0.000000]       .data : 0x40131cb0 - 0x40193900   ( 392 KiB)
[    0.000000]       .bss  : 0x401d86e0 - 0x401ec680   (  80 KiB)
[    0.000000] SLUB: HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=8
[    0.000000] NR_IRQS: 256
[    0.000000] clocksource: pit: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 1019338904850 ns
[    0.000000] Console: colour dummy device 80x25
[    0.070000] Calibrating delay loop... 238.38 BogoMIPS (lpj=1191936)
[    0.070000] pid_max: default: 32768 minimum: 301
[    0.070000] Mount-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.070000] Mountpoint-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.080000] devtmpfs: initialized
[    0.090000] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.090000] futex hash table entries: 256 (order: -2, 3072 bytes)
[    0.110000] clocksource: Switched to clocksource pit
[    0.110000] FS-Cache: Loaded
[    0.380000] workingset: timestamp_bits=27 max_order=14 bucket_order=0
[    0.430000] io scheduler noop registered
[    0.430000] io scheduler deadline registered
[    0.430000] io scheduler cfq registered (default)
[    0.430000] io scheduler mq-deadline registered
[    0.430000] io scheduler kyber registered
[    0.920000] ColdFire internal UART serial driver
[    0.920000] mcfuart.0: ttyS0 at MMIO 0xfc060000 (irq = 90, base_baud = 7500000) is a ColdFire UART
[    1.110000] console [ttyS0] enabled
[    1.110000] mcfuart.0: ttyS1 at MMIO 0xfc064000 (irq = 91, base_baud = 7500000) is a ColdFire UART
[    1.120000] mcfuart.0: ttyS2 at MMIO 0xfc068000 (irq = 92, base_baud = 7500000) is a ColdFire UART
[    1.130000] mcfuart.0: ttyS3 at MMIO 0xfc06c000 (irq = 93, base_baud = 7500000) is a ColdFire UART
[    1.150000] random: get_random_bytes called from init_oops_id+0x26/0x46 with crng_init=0
[    1.160000] Freeing unused kernel memory: 264K
[    1.170000] This architecture does not have kernel memory protection.
[    1.410000] random: fast init done

/ #

Quoted 136 lines · Angelo Dureghello
Quoted 20 lines

about my patch sent yesterday, unfortunately, i found btw at least 1 issue: the busybox "cd" doesn't change directory :)

/ # cat /proc/vmallocinfo
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc
/ # ls

bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var
/ # cd bin
/ # ls
bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var

/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vmallocinfo
0xd0000000-0xd0006000 24576 unpurged vm_area
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc

Quoted 2 lines

And after each "cd" attempt, another "unpurged" message is added. Removing MMU cd works as expected.

I setup a configuration with my 5475 where I moved the build RAM base to be 32MB - so it was no longer 0 based. Then I could run some tests to at least simulate what you have on the 54411. (The only catch is my code could still access 0 and surrounding memory addresses without faulting...)

Running with my attached patch I didn't see any odd behavior with cd/ls like you see above.

Geert: interested if you have any thoughts on problems around
virt_to_node_shift. Changing CONFIG_RAMBASE I don't think is going to resolve the problem in m68k_setup_node(). We access of the end of the array since it was divided up based on the size of the RAM, but we access it using an index derrived directly from the absolute address.

Quoted 64 lines · Geert Uytterhoeven
Quoted 51 lines

Hi Greg,

Quoted 48 lines · Greg Ungerer
Quoted 47 lines · Geert Uytterhoeven
Quoted 6 lines · Angelo Dureghello

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

       #if defined(CONFIG_RAMBASE)
       #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
       #elif defined(CONFIG_SUN3)
       #define PAGE_OFFSET_RAW         0x0E000000
       #else
       #define PAGE_OFFSET_RAW         0x00000000
       #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero,
and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +0 -0
@@ -1,6 +1,6 @@
    /* This handles the memory map.. */
 
Quoted 6 lines
Quoted 6 lines

arch/m68k/include/asm/page_offset.h line 6

Quoted 6 lines
arch/m68k/include/asm/page_offset.h +1 -1
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
    #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
    #elif defined(CONFIG_SUN3)
    #define PAGE_OFFSET_RAW                0x0E000000

I tested this patch, (removed mine) and kernel hangs somewhere silently at first init, i don't have the debug enabled now, but i suspect it still hangs at some initial "memset" since 0 area for kernel should be allocated before access.

Quoted 2 lines

Isn't PAGE_OFFSET the starting area for the virtual kernel address space ? At least so it seems to be for the famous 3G + 1G of x86.

Well, for the current working ColdFire with MMU that is the case. The kernel virtual addresses start at 0...

Quoted 11 lines

This is what i see at boot with my patch of yesterday:

[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131e70   (1220 KiB)
[    0.000000]       .data : 0x40131e70 - 0x40193900   ( 391 KiB)
[    0.000000]       .bss  : 0x401d86d8 - 0x401ec680   (  80 KiB)

For whatever it is worth this is what I see now on my debug setup:

Virtual kernel memory layout:

     vector  : 0x02000000 - 0x02000400   (   1 KiB)
     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
     lowmem  : 0x02000000 - 0x04000000   (  32 MiB)
       .init : 0x02288000 - 0x02296000   (  56 KiB)
       .text : 0x02020000 - 0x0221f270   (2045 KiB)
       .data : 0x0221f270 - 0x02284140   ( 404 KiB)
       .bss  : 0x022967a0 - 0x022ae60c   (  96 KiB)
Quoted 54 lines

Regards
Greg

Quoted 48 lines
Quoted 24 lines
Quoted 6 lines
Quoted 6 lines

or

arch/m68k/Kconfig.machine +0 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
Quoted 6 lines

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 3 lines
arch/m68k/Kconfig.machine +1 -0
    config RAMBASE
           hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

Sorry, yes, depends on !MMU.

Quoted 6 lines
Quoted 1 line

depending on whether anything else in the Coldfire code needs RAMBASE.

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

OK, as I already feared/expected...

Quoted 20 lines
Quoted 2 lines

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Right.

Gr{oetje,eeting}s,

Geert

Signature · 9 lines
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                  -- Linus Torvalds
-- 
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Regards,
Angelo

Regards,
Angelo

#25 lore raw

Hi Angelo,

Quoted 11 lines · Angelo Dureghello
Quoted 9 lines · Greg Ungerer

I have attached a patch with a slightly different approach to fixing this. Can you try this one out?

I have a feel for what is going on, based in particular on the changes you made in 0002-m68k-fix-mmu-for-coldfire-mcf5441x.patch. I see that the virt_node_shift and accessing pg_data_map when not 0 based is going to be problematic with the code as it is.

Great catch, the patch works,

Thats great, thanks for trying that out.

Quoted 6 lines

i maintaind btw the

memset(NODE_DATA(node), 0, sizeof(pg_data_t));

inside void __init m68k_setup_node(int node) since otherwise there was that warning we've seen initially.

Ok, I am not quite sure why you would still need that though. Do you mean the warning about overlaping chunks? Can you send the exact warning output again running this now?

Quoted 1 line

From the code I would have expected that array area to be in the

kernel bss and be zeroed out at system startup. (For what it is worth I don't see that warning on my non-zero based test 5475 test config).

Quoted 3 lines

About the "cd" issue, it seems to be a hush issue, maybe becouse hush is built as nommu. Re-executing hush, i can now cd to a subfolder, but i get then a SEGV on "cd ..".

Yeah, maybe that is related to flat format binaries. Normally I run a uClibc ELF based user space with MMU enabled. I'll try with a FLAT format user space and see if I get any problems.

Regards
Greg

Quoted 255 lines

This the boot log:

## Booting kernel from Legacy Image at 40001000 ...

   Image Name:   mainline kernel
   Created:      2017-09-04  14:19:47 UTC
   Image Type:   M68K Linux Kernel Image (uncompressed)
   Data Size:    1930976 Bytes = 1.8 MiB
   Load Address: 40001000
   Entry Point:  40001000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK
[    0.000000] Linux version 4.13.0-rc7stmark2-001-00039-g96e88773601f-dirty (angelo@jerusalem) (gcc version 5.2.0 (crosstools-sysam-2016.04.16)) #56 Mon Sep 4 16:19:46 CEST 2017
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 16312
[    0.000000] Kernel command line: console=ttyS0,115200 root=/dev/ram0 rw rootfstype=ramfs rdinit=/bin/init devtmpfs.mount=1
[    0.000000] PID hash table entries: 512 (order: -2, 2048 bytes)
[    0.000000] Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
[    0.000000] Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
[    0.000000] Sorting __ex_table...
[    0.000000] Memory: 128288K/131072K available (1219K kernel code, 103K rwdata, 288K rodata, 264K init, 79K bss, 2784K reserved, 0K cma-reserved)
[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131cb0   (1220 KiB)
[    0.000000]       .data : 0x40131cb0 - 0x40193900   ( 392 KiB)
[    0.000000]       .bss  : 0x401d86e0 - 0x401ec680   (  80 KiB)
[    0.000000] SLUB: HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=8
[    0.000000] NR_IRQS: 256
[    0.000000] clocksource: pit: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 1019338904850 ns
[    0.000000] Console: colour dummy device 80x25
[    0.070000] Calibrating delay loop... 238.38 BogoMIPS (lpj=1191936)
[    0.070000] pid_max: default: 32768 minimum: 301
[    0.070000] Mount-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.070000] Mountpoint-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.080000] devtmpfs: initialized
[    0.090000] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.090000] futex hash table entries: 256 (order: -2, 3072 bytes)
[    0.110000] clocksource: Switched to clocksource pit
[    0.110000] FS-Cache: Loaded
[    0.380000] workingset: timestamp_bits=27 max_order=14 bucket_order=0
[    0.430000] io scheduler noop registered
[    0.430000] io scheduler deadline registered
[    0.430000] io scheduler cfq registered (default)
[    0.430000] io scheduler mq-deadline registered
[    0.430000] io scheduler kyber registered
[    0.920000] ColdFire internal UART serial driver
[    0.920000] mcfuart.0: ttyS0 at MMIO 0xfc060000 (irq = 90, base_baud = 7500000) is a ColdFire UART
[    1.110000] console [ttyS0] enabled
[    1.110000] mcfuart.0: ttyS1 at MMIO 0xfc064000 (irq = 91, base_baud = 7500000) is a ColdFire UART
[    1.120000] mcfuart.0: ttyS2 at MMIO 0xfc068000 (irq = 92, base_baud = 7500000) is a ColdFire UART
[    1.130000] mcfuart.0: ttyS3 at MMIO 0xfc06c000 (irq = 93, base_baud = 7500000) is a ColdFire UART
[    1.150000] random: get_random_bytes called from init_oops_id+0x26/0x46 with crng_init=0
[    1.160000] Freeing unused kernel memory: 264K
[    1.170000] This architecture does not have kernel memory protection.
[    1.410000] random: fast init done

/ #

Quoted 136 lines · Angelo Dureghello
Quoted 20 lines

about my patch sent yesterday, unfortunately, i found btw at least 1 issue: the busybox "cd" doesn't change directory :)

/ # cat /proc/vmallocinfo
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc
/ # ls

bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var
/ # cd bin
/ # ls
bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var

/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vmallocinfo
0xd0000000-0xd0006000 24576 unpurged vm_area
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc

Quoted 2 lines

And after each "cd" attempt, another "unpurged" message is added. Removing MMU cd works as expected.

I setup a configuration with my 5475 where I moved the build RAM base to be 32MB - so it was no longer 0 based. Then I could run some tests to at least simulate what you have on the 54411. (The only catch is my code could still access 0 and surrounding memory addresses without faulting...)

Running with my attached patch I didn't see any odd behavior with cd/ls like you see above.

Geert: interested if you have any thoughts on problems around
virt_to_node_shift. Changing CONFIG_RAMBASE I don't think is going to resolve the problem in m68k_setup_node(). We access of the end of the array since it was divided up based on the size of the RAM, but we access it using an index derrived directly from the absolute address.

Quoted 64 lines · Geert Uytterhoeven
Quoted 51 lines

Hi Greg,

Quoted 48 lines · Greg Ungerer
Quoted 47 lines · Geert Uytterhoeven
Quoted 6 lines · Angelo Dureghello

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

       #if defined(CONFIG_RAMBASE)
       #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
       #elif defined(CONFIG_SUN3)
       #define PAGE_OFFSET_RAW         0x0E000000
       #else
       #define PAGE_OFFSET_RAW         0x00000000
       #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero,
and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +0 -0
@@ -1,6 +1,6 @@
    /* This handles the memory map.. */
 
Quoted 6 lines
Quoted 6 lines

arch/m68k/include/asm/page_offset.h line 6

Quoted 6 lines
arch/m68k/include/asm/page_offset.h +1 -1
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
    #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
    #elif defined(CONFIG_SUN3)
    #define PAGE_OFFSET_RAW                0x0E000000

I tested this patch, (removed mine) and kernel hangs somewhere silently at first init, i don't have the debug enabled now, but i suspect it still hangs at some initial "memset" since 0 area for kernel should be allocated before access.

Quoted 2 lines

Isn't PAGE_OFFSET the starting area for the virtual kernel address space ? At least so it seems to be for the famous 3G + 1G of x86.

Well, for the current working ColdFire with MMU that is the case. The kernel virtual addresses start at 0...

Quoted 11 lines

This is what i see at boot with my patch of yesterday:

[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131e70   (1220 KiB)
[    0.000000]       .data : 0x40131e70 - 0x40193900   ( 391 KiB)
[    0.000000]       .bss  : 0x401d86d8 - 0x401ec680   (  80 KiB)

For whatever it is worth this is what I see now on my debug setup:

Virtual kernel memory layout:

     vector  : 0x02000000 - 0x02000400   (   1 KiB)
     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
     lowmem  : 0x02000000 - 0x04000000   (  32 MiB)
       .init : 0x02288000 - 0x02296000   (  56 KiB)
       .text : 0x02020000 - 0x0221f270   (2045 KiB)
       .data : 0x0221f270 - 0x02284140   ( 404 KiB)
       .bss  : 0x022967a0 - 0x022ae60c   (  96 KiB)
Quoted 54 lines

Regards
Greg

Quoted 48 lines
Quoted 24 lines
Quoted 6 lines
Quoted 6 lines

or

arch/m68k/Kconfig.machine +0 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
Quoted 6 lines

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 3 lines
arch/m68k/Kconfig.machine +1 -0
    config RAMBASE
           hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

Sorry, yes, depends on !MMU.

Quoted 6 lines
Quoted 1 line

depending on whether anything else in the Coldfire code needs RAMBASE.

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

OK, as I already feared/expected...

Quoted 20 lines
Quoted 2 lines

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Right.

Gr{oetje,eeting}s,

Geert

Signature · 9 lines
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                  -- Linus Torvalds
-- 
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Regards,
Angelo

Regards,
Angelo

#26 lore raw

Hi Greg,

Quoted 25 lines · Greg Ungerer

Hi Angelo,

Quoted 11 lines · Angelo Dureghello
Quoted 9 lines · Greg Ungerer

I have attached a patch with a slightly different approach to fixing this. Can you try this one out?

I have a feel for what is going on, based in particular on the changes you made in 0002-m68k-fix-mmu-for-coldfire-mcf5441x.patch. I see that the virt_node_shift and accessing pg_data_map when not 0 based is going to be problematic with the code as it is.

Great catch, the patch works,

Thats great, thanks for trying that out.

Quoted 6 lines

i maintaind btw the

memset(NODE_DATA(node), 0, sizeof(pg_data_t));

inside void __init m68k_setup_node(int node) since otherwise there was that warning we've seen initially.

Quoted 9 lines

Ok, I am not quite sure why you would still need that though. Do you mean the warning about overlaping chunks? Can you send the exact warning output again running this now?

From the code I would have expected that array area to be in the
kernel bss and be zeroed out at system startup. (For what it is worth I don't see that warning on my non-zero based test 5475 test config).

Sure, my mistake, i just maintained that line and didn't try to remove it after your final 1-line patch.

Just re-tested now, without that memset, all works fine as you was expecting, no warnings.
I attach the 2 progressive patches, the first you sent me initially, and this last one, that actually is just 1 line patch.

Quoted 5 lines
Quoted 4 lines

About the "cd" issue, it seems to be a hush issue, maybe becouse hush is built as nommu. Re-executing hush, i can now cd to a subfolder, but i get then a SEGV on "cd ..".

Quoted 265 lines

Yeah, maybe that is related to flat format binaries. Normally I run a uClibc ELF based user space with MMU enabled. I'll try with a FLAT format user space and see if I get any problems.

Regards
Greg

Quoted 255 lines

This the boot log:

## Booting kernel from Legacy Image at 40001000 ...

    Image Name:   mainline kernel
    Created:      2017-09-04  14:19:47 UTC
    Image Type:   M68K Linux Kernel Image (uncompressed)
    Data Size:    1930976 Bytes = 1.8 MiB
    Load Address: 40001000
    Entry Point:  40001000
    Verifying Checksum ... OK
    Loading Kernel Image ... OK
[    0.000000] Linux version 4.13.0-rc7stmark2-001-00039-g96e88773601f-dirty (angelo@jerusalem) (gcc version 5.2.0 (crosstools-sysam-2016.04.16)) #56 Mon Sep 4 16:19:46 CEST 2017
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 16312
[    0.000000] Kernel command line: console=ttyS0,115200 root=/dev/ram0 rw rootfstype=ramfs rdinit=/bin/init devtmpfs.mount=1
[    0.000000] PID hash table entries: 512 (order: -2, 2048 bytes)
[    0.000000] Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
[    0.000000] Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
[    0.000000] Sorting __ex_table...
[    0.000000] Memory: 128288K/131072K available (1219K kernel code, 103K rwdata, 288K rodata, 264K init, 79K bss, 2784K reserved, 0K cma-reserved)
[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131cb0   (1220 KiB)
[    0.000000]       .data : 0x40131cb0 - 0x40193900   ( 392 KiB)
[    0.000000]       .bss  : 0x401d86e0 - 0x401ec680   (  80 KiB)
[    0.000000] SLUB: HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=8
[    0.000000] NR_IRQS: 256
[    0.000000] clocksource: pit: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 1019338904850 ns
[    0.000000] Console: colour dummy device 80x25
[    0.070000] Calibrating delay loop... 238.38 BogoMIPS (lpj=1191936)
[    0.070000] pid_max: default: 32768 minimum: 301
[    0.070000] Mount-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.070000] Mountpoint-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.080000] devtmpfs: initialized
[    0.090000] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.090000] futex hash table entries: 256 (order: -2, 3072 bytes)
[    0.110000] clocksource: Switched to clocksource pit
[    0.110000] FS-Cache: Loaded
[    0.380000] workingset: timestamp_bits=27 max_order=14 bucket_order=0
[    0.430000] io scheduler noop registered
[    0.430000] io scheduler deadline registered
[    0.430000] io scheduler cfq registered (default)
[    0.430000] io scheduler mq-deadline registered
[    0.430000] io scheduler kyber registered
[    0.920000] ColdFire internal UART serial driver
[    0.920000] mcfuart.0: ttyS0 at MMIO 0xfc060000 (irq = 90, base_baud = 7500000) is a ColdFire UART
[    1.110000] console [ttyS0] enabled
[    1.110000] mcfuart.0: ttyS1 at MMIO 0xfc064000 (irq = 91, base_baud = 7500000) is a ColdFire UART
[    1.120000] mcfuart.0: ttyS2 at MMIO 0xfc068000 (irq = 92, base_baud = 7500000) is a ColdFire UART
[    1.130000] mcfuart.0: ttyS3 at MMIO 0xfc06c000 (irq = 93, base_baud = 7500000) is a ColdFire UART
[    1.150000] random: get_random_bytes called from init_oops_id+0x26/0x46 with crng_init=0
[    1.160000] Freeing unused kernel memory: 264K
[    1.170000] This architecture does not have kernel memory protection.
[    1.410000] random: fast init done

/ #

Quoted 136 lines · Angelo Dureghello
Quoted 20 lines

about my patch sent yesterday, unfortunately, i found btw at least 1 issue: the busybox "cd" doesn't change directory :)

/ # cat /proc/vmallocinfo
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc
/ # ls

bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var
/ # cd bin
/ # ls
bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var

/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vmallocinfo
0xd0000000-0xd0006000 24576 unpurged vm_area
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc

Quoted 2 lines

And after each "cd" attempt, another "unpurged" message is added. Removing MMU cd works as expected.

I setup a configuration with my 5475 where I moved the build RAM base to be 32MB - so it was no longer 0 based. Then I could run some tests to at least simulate what you have on the 54411. (The only catch is my code could still access 0 and surrounding memory addresses without faulting...)

Running with my attached patch I didn't see any odd behavior with cd/ls like you see above.

Geert: interested if you have any thoughts on problems around
virt_to_node_shift. Changing CONFIG_RAMBASE I don't think is going to resolve the problem in m68k_setup_node(). We access of the end of the array since it was divided up based on the size of the RAM, but we access it using an index derrived directly from the absolute address.

Quoted 64 lines · Geert Uytterhoeven
Quoted 51 lines

Hi Greg,

Quoted 48 lines · Greg Ungerer
Quoted 47 lines · Geert Uytterhoeven
Quoted 6 lines · Angelo Dureghello

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

        #if defined(CONFIG_RAMBASE)
        #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
        #elif defined(CONFIG_SUN3)
        #define PAGE_OFFSET_RAW         0x0E000000
        #else
        #define PAGE_OFFSET_RAW         0x00000000
        #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero,
and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +0 -0
@@ -1,6 +1,6 @@
     /* This handles the memory map.. */
 
Quoted 6 lines
Quoted 6 lines

arch/m68k/include/asm/page_offset.h line 6

Quoted 6 lines
arch/m68k/include/asm/page_offset.h +1 -1
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
     #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
     #elif defined(CONFIG_SUN3)
     #define PAGE_OFFSET_RAW                0x0E000000

I tested this patch, (removed mine) and kernel hangs somewhere silently at first init, i don't have the debug enabled now, but i suspect it still hangs at some initial "memset" since 0 area for kernel should be allocated before access.

Quoted 2 lines

Isn't PAGE_OFFSET the starting area for the virtual kernel address space ? At least so it seems to be for the famous 3G + 1G of x86.

Well, for the current working ColdFire with MMU that is the case. The kernel virtual addresses start at 0...

Quoted 11 lines

This is what i see at boot with my patch of yesterday:

[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131e70   (1220 KiB)
[    0.000000]       .data : 0x40131e70 - 0x40193900   ( 391 KiB)
[    0.000000]       .bss  : 0x401d86d8 - 0x401ec680   (  80 KiB)

For whatever it is worth this is what I see now on my debug setup:

Virtual kernel memory layout:

      vector  : 0x02000000 - 0x02000400   (   1 KiB)
      kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
      vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
      lowmem  : 0x02000000 - 0x04000000   (  32 MiB)
        .init : 0x02288000 - 0x02296000   (  56 KiB)
        .text : 0x02020000 - 0x0221f270   (2045 KiB)
        .data : 0x0221f270 - 0x02284140   ( 404 KiB)
        .bss  : 0x022967a0 - 0x022ae60c   (  96 KiB)
Quoted 54 lines

Regards
Greg

Quoted 48 lines
Quoted 24 lines
Quoted 6 lines
Quoted 6 lines

or

arch/m68k/Kconfig.machine +0 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
Quoted 6 lines

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 3 lines
arch/m68k/Kconfig.machine +1 -0
     config RAMBASE
            hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

Sorry, yes, depends on !MMU.

Quoted 6 lines
Quoted 1 line

depending on whether anything else in the Coldfire code needs RAMBASE.

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

OK, as I already feared/expected...

Quoted 20 lines
Quoted 2 lines

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Right.

Gr{oetje,eeting}s,

Geert

Signature · 9 lines
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                   -- Linus Torvalds
-- 
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Regards,
Angelo

Regards,
Angelo

Best regards,
Angelo Dureghello

Quoted 1 line

From 9f0e06f9566de4d91a234753cd30160ce02fc8a8 Mon Sep 17 00:00:00 2001

From: Angelo Dureghello <[email protected]>
Date: Thu, 24 Aug 2017 00:01:06 +0200
Subject: [PATCH 1/2] Patch to enable mmu, from Greg.

---
 arch/m68k/Kconfig.cpu               |  2 +-
 arch/m68k/coldfire/m54xx.c          |  4 ----
 arch/m68k/configs/stmark2_defconfig |  5 +++--
 arch/m68k/include/asm/mmu_context.h |  1 -
 arch/m68k/kernel/setup_mm.c         |  1 +
 arch/m68k/mm/mcfmmu.c               | 35 ++++++++++++++++++-----------------
 6 files changed, 23 insertions(+), 25 deletions(-)
arch/m68k/Kconfig.cpu +1 -1
index d2219f30b78f..4dc51c090a45 100644
@@ -283,7 +283,7 @@ config M548x
 
 config M5441x
 	bool "MCF5441x"
-	depends on !MMU
+	select MMU_COLDFIRE if MMU
 	select GENERIC_CLOCKEVENTS
 	select HAVE_CACHE_CB
 	help
arch/m68k/coldfire/m54xx.c +0 -4
index c552851ec617..704efeaeab2d 100644
@@ -95,10 +95,6 @@ static void mcf54xx_reset(void)
 
 void __init config_BSP(char *commandp, int size)
 {
-#ifdef CONFIG_MMU
-	cf_bootmem_alloc();
-	mmu_context_init();
-#endif
 	mach_reset = mcf54xx_reset;
 	mach_sched_init = hw_timer_init;
 	m54xx_uarts_init();
arch/m68k/configs/stmark2_defconfig +3 -2
index ea78b8f6a68e..b886f303cc3f 100644
@@ -21,7 +21,7 @@ CONFIG_EMBEDDED=y
 # CONFIG_LBDAF is not set
 # CONFIG_BLK_DEV_BSG is not set
 CONFIG_BLK_CMDLINE_PARSER=y
-# CONFIG_MMU is not set
+CONFIG_COLDFIRE=y
 CONFIG_M5441x=y
 CONFIG_CLOCK_FREQ=240000000
 CONFIG_STMARK2=y
@@ -30,6 +30,8 @@ CONFIG_RAMSIZE=0x8000000
 CONFIG_VECTORBASE=0x40000000
 CONFIG_KERNELBASE=0x40001000
 CONFIG_BINFMT_FLAT=y
+CONFIG_BINFMT_ZFLAT=y
+CONFIG_BINFMT_SHARED_FLAT=y
 CONFIG_BINFMT_MISC=y
 # CONFIG_UEVENT_HELPER is not set
 CONFIG_DEVTMPFS=y
@@ -53,7 +55,6 @@ CONFIG_MTD_COMPLEX_MAPPINGS=y
 CONFIG_MTD_PHYSMAP=y
 CONFIG_MTD_PLATRAM=y
 CONFIG_MTD_SPI_NOR=y
-# CONFIG_INPUT_MOUSEDEV is not set
 # CONFIG_INPUT_KEYBOARD is not set
 # CONFIG_INPUT_MOUSE is not set
 CONFIG_SERIO_LIBPS2=y
arch/m68k/include/asm/mmu_context.h +0 -1
index 4a6ae6dffa34..00c28b1dc47b 100644
@@ -91,7 +91,6 @@ static inline void activate_mm(struct mm_struct *active_mm,
 
 #define deactivate_mm(tsk, mm) do { } while (0)
 
-extern void mmu_context_init(void);
 #define prepare_arch_switch(next) load_ksp_mmu(next)
 
 static inline void load_ksp_mmu(struct task_struct *task)
arch/m68k/kernel/setup_mm.c +1 -0
index 7a2c21212820..5632c48b9f3b 100644
@@ -343,6 +343,7 @@ void __init setup_arch(char **cmdline_p)
 #ifdef CONFIG_COLDFIRE
 	case MACH_M54XX:
 	case MACH_M5441X:
+		cf_bootmem_alloc();
 		config_BSP(NULL, 0);
 		break;
 #endif
arch/m68k/mm/mcfmmu.c +18 -17
index 87131cd3bc8f..c7efdf8e8eae 100644
@@ -150,6 +150,23 @@ int cf_tlb_miss(struct pt_regs *regs, int write, int dtlb, int extension_word)
 	return 0;
 }
 
+/*
+ * Initialize the context management stuff.
+ * The following was taken from arch/ppc/mmu_context.c
+ */
+static void __init mmu_context_init(void)
+{
+	/*
+	 * Some processors have too few contexts to reserve one for
+	 * init_mm, and require using context 0 for a normal task.
+	 * Other processors reserve the use of context zero for the kernel.
+	 * This code assumes FIRST_CONTEXT < 32.
+	 */
+	context_map[0] = (1 << FIRST_CONTEXT) - 1;
+	next_mmu_context = FIRST_CONTEXT;
+	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+}
+
 void __init cf_bootmem_alloc(void)
 {
 	unsigned long start_pfn;
@@ -177,23 +194,7 @@ void __init cf_bootmem_alloc(void)
 	memstart += init_bootmem_node(NODE_DATA(0), start_pfn,
 		min_low_pfn, max_low_pfn);
 	free_bootmem_node(NODE_DATA(0), memstart, _ramend - memstart);
-}
-
-/*
- * Initialize the context management stuff.
- * The following was taken from arch/ppc/mmu_context.c
- */
-void __init mmu_context_init(void)
-{
-	/*
-	 * Some processors have too few contexts to reserve one for
-	 * init_mm, and require using context 0 for a normal task.
-	 * Other processors reserve the use of context zero for the kernel.
-	 * This code assumes FIRST_CONTEXT < 32.
-	 */
-	context_map[0] = (1 << FIRST_CONTEXT) - 1;
-	next_mmu_context = FIRST_CONTEXT;
-	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+	mmu_context_init();
 }
 
 /*
Signature · 1 line
2.14.1
Quoted 1 line

From 9f6deae2b7c4291e792e2b341da461091613b790 Mon Sep 17 00:00:00 2001

From: Angelo Dureghello <[email protected]>
Date: Sun, 27 Aug 2017 02:42:42 +0200
Subject: [PATCH 2/2] m68k: fix mmu for coldfire mcf5441x

This patch fixes mmu not working for CPU's that has the base of the physical memory mapped at a non-zero address.

---
 arch/m68k/mm/mcfmmu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/m68k/mm/mcfmmu.c +1 -1
index c7efdf8e8eae..a79198a7fcab 100644
@@ -186,7 +186,7 @@ void __init cf_bootmem_alloc(void)
 	max_pfn = max_low_pfn = PFN_DOWN(_ramend);
 	high_memory = (void *)_ramend;
 
-	m68k_virt_to_node_shift = fls(_ramend - _rambase - 1) - 6;
+	m68k_virt_to_node_shift = fls(_ramend - 1) - 6;
 	module_fixup(NULL, __start_fixup, __stop_fixup);
 
 	/* setup bootmem data */
Signature · 1 line
2.14.1
#27 lore raw

Hi Angelo,

Quoted 34 lines · Angelo Dureghello
Quoted 23 lines · Greg Ungerer
Quoted 11 lines · Angelo Dureghello
Quoted 9 lines · Greg Ungerer

I have attached a patch with a slightly different approach to fixing this. Can you try this one out?

I have a feel for what is going on, based in particular on the changes you made in 0002-m68k-fix-mmu-for-coldfire-mcf5441x.patch. I see that the virt_node_shift and accessing pg_data_map when not 0 based is going to be problematic with the code as it is.

Great catch, the patch works,

Thats great, thanks for trying that out.

Quoted 6 lines

i maintaind btw the

memset(NODE_DATA(node), 0, sizeof(pg_data_t));

inside void __init m68k_setup_node(int node) since otherwise there was that warning we've seen initially.

Quoted 9 lines

Ok, I am not quite sure why you would still need that though. Do you mean the warning about overlaping chunks? Can you send the exact warning output again running this now?

From the code I would have expected that array area to be in the
kernel bss and be zeroed out at system startup. (For what it is worth I don't see that warning on my non-zero based test 5475 test config).

Quoted 7 lines

Sure, my mistake, i just maintained that line and didn't try to remove it after your final 1-line patch.

Just re-tested now, without that memset, all works fine as you was expecting, no warnings.
I attach the 2 progressive patches, the first you sent me initially, and this last one, that actually is just 1 line patch.

Great, thanks for your efforts working through this. I'll send out the patches again as a series for final review.

Regards
Greg

Quoted 272 lines
Quoted 4 lines
Quoted 3 lines

About the "cd" issue, it seems to be a hush issue, maybe becouse hush is built as nommu. Re-executing hush, i can now cd to a subfolder, but i get then a SEGV on "cd ..".

Quoted 265 lines

Yeah, maybe that is related to flat format binaries. Normally I run a uClibc ELF based user space with MMU enabled. I'll try with a FLAT format user space and see if I get any problems.

Regards
Greg

Quoted 255 lines

This the boot log:

## Booting kernel from Legacy Image at 40001000 ...

    Image Name:   mainline kernel
    Created:      2017-09-04  14:19:47 UTC
    Image Type:   M68K Linux Kernel Image (uncompressed)
    Data Size:    1930976 Bytes = 1.8 MiB
    Load Address: 40001000
    Entry Point:  40001000
    Verifying Checksum ... OK
    Loading Kernel Image ... OK
[    0.000000] Linux version 4.13.0-rc7stmark2-001-00039-g96e88773601f-dirty (angelo@jerusalem) (gcc version 5.2.0 (crosstools-sysam-2016.04.16)) #56 Mon Sep 4 16:19:46 CEST 2017
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 16312
[    0.000000] Kernel command line: console=ttyS0,115200 root=/dev/ram0 rw rootfstype=ramfs rdinit=/bin/init devtmpfs.mount=1
[    0.000000] PID hash table entries: 512 (order: -2, 2048 bytes)
[    0.000000] Dentry cache hash table entries: 16384 (order: 3, 65536 bytes)
[    0.000000] Inode-cache hash table entries: 8192 (order: 2, 32768 bytes)
[    0.000000] Sorting __ex_table...
[    0.000000] Memory: 128288K/131072K available (1219K kernel code, 103K rwdata, 288K rodata, 264K init, 79K bss, 2784K reserved, 0K cma-reserved)
[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131cb0   (1220 KiB)
[    0.000000]       .data : 0x40131cb0 - 0x40193900   ( 392 KiB)
[    0.000000]       .bss  : 0x401d86e0 - 0x401ec680   (  80 KiB)
[    0.000000] SLUB: HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=8
[    0.000000] NR_IRQS: 256
[    0.000000] clocksource: pit: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 1019338904850 ns
[    0.000000] Console: colour dummy device 80x25
[    0.070000] Calibrating delay loop... 238.38 BogoMIPS (lpj=1191936)
[    0.070000] pid_max: default: 32768 minimum: 301
[    0.070000] Mount-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.070000] Mountpoint-cache hash table entries: 2048 (order: 0, 8192 bytes)
[    0.080000] devtmpfs: initialized
[    0.090000] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.090000] futex hash table entries: 256 (order: -2, 3072 bytes)
[    0.110000] clocksource: Switched to clocksource pit
[    0.110000] FS-Cache: Loaded
[    0.380000] workingset: timestamp_bits=27 max_order=14 bucket_order=0
[    0.430000] io scheduler noop registered
[    0.430000] io scheduler deadline registered
[    0.430000] io scheduler cfq registered (default)
[    0.430000] io scheduler mq-deadline registered
[    0.430000] io scheduler kyber registered
[    0.920000] ColdFire internal UART serial driver
[    0.920000] mcfuart.0: ttyS0 at MMIO 0xfc060000 (irq = 90, base_baud = 7500000) is a ColdFire UART
[    1.110000] console [ttyS0] enabled
[    1.110000] mcfuart.0: ttyS1 at MMIO 0xfc064000 (irq = 91, base_baud = 7500000) is a ColdFire UART
[    1.120000] mcfuart.0: ttyS2 at MMIO 0xfc068000 (irq = 92, base_baud = 7500000) is a ColdFire UART
[    1.130000] mcfuart.0: ttyS3 at MMIO 0xfc06c000 (irq = 93, base_baud = 7500000) is a ColdFire UART
[    1.150000] random: get_random_bytes called from init_oops_id+0x26/0x46 with crng_init=0
[    1.160000] Freeing unused kernel memory: 264K
[    1.170000] This architecture does not have kernel memory protection.
[    1.410000] random: fast init done

/ #

Quoted 136 lines · Angelo Dureghello
Quoted 20 lines

about my patch sent yesterday, unfortunately, i found btw at least 1 issue: the busybox "cd" doesn't change directory :)

/ # cat /proc/vmallocinfo
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc
/ # ls

bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var
/ # cd bin
/ # ls
bin    etc    lib    mnt    proc   sbin   usr
dev    home   media  opt    root   sys    var

/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vm
vmallocinfo vmstat
/ # cat /proc/vmallocinfo
0xd0000000-0xd0006000 24576 unpurged vm_area
0xd0006000-0xd000c000 24576 n_tty_open+0x16/0xae pages=2 vmalloc

Quoted 2 lines

And after each "cd" attempt, another "unpurged" message is added. Removing MMU cd works as expected.

I setup a configuration with my 5475 where I moved the build RAM base to be 32MB - so it was no longer 0 based. Then I could run some tests to at least simulate what you have on the 54411. (The only catch is my code could still access 0 and surrounding memory addresses without faulting...)

Running with my attached patch I didn't see any odd behavior with cd/ls like you see above.

Geert: interested if you have any thoughts on problems around
virt_to_node_shift. Changing CONFIG_RAMBASE I don't think is going to resolve the problem in m68k_setup_node(). We access of the end of the array since it was divided up based on the size of the RAM, but we access it using an index derrived directly from the absolute address.

Quoted 64 lines · Geert Uytterhoeven
Quoted 51 lines

Hi Greg,

Quoted 48 lines · Greg Ungerer
Quoted 47 lines · Geert Uytterhoeven
Quoted 6 lines · Angelo Dureghello

did some additional study and tests.

Actually i cannot find any simpler patch, i just adjusted it a bit. I believe this patch will work on your cpu with 0-based memoryas well.
I attach it for your comments.

I think your issue is caused by arch/m68k/include/asm/page_offset.h:

        #if defined(CONFIG_RAMBASE)
        #define PAGE_OFFSET_RAW         CONFIG_RAMBASE
        #elif defined(CONFIG_SUN3)
        #define PAGE_OFFSET_RAW         0x0E000000
        #else
        #define PAGE_OFFSET_RAW         0x00000000
        #endif

and arch/m68k/Kconfig.machine:

if !MMU || COLDFIRE

config RAMBASE
hex "Address of the base of RAM"
default "0"

So on MC680[2346]0 with MMU (and ignoring Sun-3, which is special), PAGE_OFFSET == PAGE_OFFSET_RAW == 0.

On Greg's zero-based Coldfire with MMU, CONFIG_RAMBASE is zero, and thus PAGE_OFFSET is also zero.

On your board CONFIG_RAMBASE is non-zero, hence PAGE_OFFSET is also non-zero,
and thus you have to compensate for that, cfr. your second patch.

Does it work if you force PAGE_OFFSET_RAW to zero?

If yes, we either need:

arch/m68k/include/asm/page_offset.h +0 -0
@@ -1,6 +1,6 @@
     /* This handles the memory map.. */
 
Quoted 6 lines
Quoted 6 lines

arch/m68k/include/asm/page_offset.h line 6

Quoted 6 lines
arch/m68k/include/asm/page_offset.h +1 -1
-#if defined(CONFIG_RAMBASE)
+#if !defined(CONFIG_MMU)
     #define PAGE_OFFSET_RAW                CONFIG_RAMBASE
     #elif defined(CONFIG_SUN3)
     #define PAGE_OFFSET_RAW                0x0E000000

I tested this patch, (removed mine) and kernel hangs somewhere silently at first init, i don't have the debug enabled now, but i suspect it still hangs at some initial "memset" since 0 area for kernel should be allocated before access.

Quoted 2 lines

Isn't PAGE_OFFSET the starting area for the virtual kernel address space ? At least so it seems to be for the famous 3G + 1G of x86.

Well, for the current working ColdFire with MMU that is the case. The kernel virtual addresses start at 0...

Quoted 11 lines

This is what i see at boot with my patch of yesterday:

[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0x40000000 - 0x40000400   (   1 KiB)
[    0.000000]     kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
[    0.000000]     vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
[    0.000000]     lowmem  : 0x40000000 - 0x48000000   ( 128 MiB)
[    0.000000]       .init : 0x40196000 - 0x401d8000   ( 264 KiB)
[    0.000000]       .text : 0x40001000 - 0x40131e70   (1220 KiB)
[    0.000000]       .data : 0x40131e70 - 0x40193900   ( 391 KiB)
[    0.000000]       .bss  : 0x401d86d8 - 0x401ec680   (  80 KiB)

For whatever it is worth this is what I see now on my debug setup:

Virtual kernel memory layout:

      vector  : 0x02000000 - 0x02000400   (   1 KiB)
      kmap    : 0xe0000000 - 0xf0000000   ( 256 MiB)
      vmalloc : 0xd0000000 - 0xe0000000   ( 256 MiB)
      lowmem  : 0x02000000 - 0x04000000   (  32 MiB)
        .init : 0x02288000 - 0x02296000   (  56 KiB)
        .text : 0x02020000 - 0x0221f270   (2045 KiB)
        .data : 0x0221f270 - 0x02284140   ( 404 KiB)
        .bss  : 0x022967a0 - 0x022ae60c   (  96 KiB)
Quoted 54 lines

Regards
Greg

Quoted 48 lines
Quoted 24 lines
Quoted 6 lines
Quoted 6 lines

or

arch/m68k/Kconfig.machine +0 -0
@@ -325,6 +325,7 @@ comment "RAM configuration"
 
Quoted 6 lines

arch/m68k/Kconfig.machine comment "RAM configuration" · line 328

Quoted 3 lines
arch/m68k/Kconfig.machine +1 -0
     config RAMBASE
            hex "Address of the base of RAM"
+       depends on MMU

Did you mean "depends on !MMU" here?

Sorry, yes, depends on !MMU.

Quoted 6 lines
Quoted 1 line

depending on whether anything else in the Coldfire code needs RAMBASE.

There are a couple of places we depend on CONFIG_RAMBASE even when running with the MMU enabled. Most importantly in setting the cachable regions in arch/m68k/include/asm/m54xxacr.h. So this is probably not going to work on its own.

OK, as I already feared/expected...

Quoted 20 lines
Quoted 2 lines

But the first patch above should be ok. It should certainly work on my 0 address base 5475 ColdFire setup. Angelo can you try that one?

Right.

Gr{oetje,eeting}s,

Geert

Signature · 9 lines
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                   -- Linus Torvalds
-- 
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Regards,
Angelo

Regards,
Angelo

Best regards,
Angelo Dureghello

#28 lore raw
In reply to Greg Ungerer

Hi greg,

Quoted 27 lines · Greg Ungerer

Hi Angelo,

On 10/08/17 01:32, Angelo Dureghello wrote:
[snip]

Quoted 11 lines

sure, on this board http://sysam.it/cff_stmark2.html
there are 128MB of ddr2.

External SDRAM is accessible, at least without any mmc support enabled, from 0x40000000.

I have following test config:

GNU nano 2.8.6 File: arch/m68k/configs/stmark2_defconfig

CONFIG_LOCALVERSION="stmark2-001"

[snip]

Quoted 7 lines

I tried still yesterday a bit, but seems there is no much support for earlyprintk / low level debug for this architecture.

In case i can try with a gpio toggling routine, at least to find where kernel stops.

The attached patch, is a quick and dirty early console output method. It works for me on the m5475, should work for you "as is" on the 5441x too.

Quoted 8 lines

It is kind of an early printk. Of course it still needs the early kernel boot to have succeeded before you will get anything much coming out. But it is worth trying.

I am wondering if the non-0 base RAM may be a problem. I have only run the MMU enabled code on platforms with 0 based RAM so far. But lets see if the early console trace attached gives us anything before digging into that.

I have some news, found the exact point where the kernel stops, it is a " panic("Out of memory");" but you can check the comments in the diff below to see what exactly happen:

iff --git a/mm/bootmem.c b/mm/bootmem.c
index 9fedb27c6451..2bf51bf1bf28 100644

mm/bootmem.c +23 -0
@@ -20,6 +20,8 @@
 
 #include "internal.h"
 
+extern void __square(void);
+
 #ifndef CONFIG_NEED_MULTIPLE_NODES
 struct pglist_data __refdata contig_page_data = {
        .bdata = &bootmem_node_data[0]
@@ -501,6 +503,8 @@ static void * __init alloc_bootmem_bdata(struct bootmem_data *bdata,
        unsigned long fallback = 0;
        unsigned long min, max, start, sidx, midx, step;
 
+       /* we don't even enter here */
+
        bdebug("nid=%td size=%lx [%lu pages] align=%lx goal=%lx limit=%lx\n",
                bdata - bootmem_node_data, size, PAGE_ALIGN(size) >> PAGE_SHIFT,
                align, goal, limit);
@@ -539,6 +543,7 @@ static void * __init alloc_bootmem_bdata(struct bootmem_data *bdata,
                 * catch the fallback below.
                 */
                fallback = sidx + 1;
+
                sidx = align_idx(bdata, bdata->hint_idx, step);
        }
 
@@ -609,12 +614,18 @@ static void * __init alloc_bootmem_core(unsigned long size,
        bootmem_data_t *bdata;
        void *region;
 
+       /* we pass from here */
        if (WARN_ON_ONCE(slab_is_available()))
                return kzalloc(size, GFP_NOWAIT);
 
+       /* we pass from here */
        list_for_each_entry(bdata, &bdata_list, list) {
+
+               /* we never enter here */
+
                if (goal && bdata->node_low_pfn <= PFN_DOWN(goal))
                        continue;
+
                if (limit && bdata->node_min_pfn >= PFN_DOWN(limit))
                        break;
 
@@ -623,6 +634,8 @@ static void * __init alloc_bootmem_core(unsigned long size,
                        return region;
        }
 
+       /* we pass from here */
+
        return NULL;
 }
 
@@ -637,11 +650,18 @@ static void * __init ___alloc_bootmem_nopanic(unsigned long size,
        ptr = alloc_bootmem_core(size, align, goal, limit);
        if (ptr)
                return ptr;
+
+       /* alloc_bootmem_core returns null, ptr = 0 */
+
        if (goal) {
                goal = 0;
                goto restart;
+
+               /* we pass from here yes */
        }
 
+       /* we pass from here yes */
+
        return NULL;
 }
 
@@ -677,7 +697,10 @@ static void * __init ___alloc_bootmem(unsigned long size, unsigned long align,
         * Whoops, we cannot satisfy the allocation request.
         */
        pr_alert("bootmem alloc of %lu bytes failed!\n", size);
+
+       /* we finally stop here */
        panic("Out of memory");
+
        return NULL;
 }
 

So trying now to understand why that

list_for_each_entry(bdata, &bdata_list, list)

looks as an emtty list.

Regards,
Angelo

Quoted 3 lines

Regards
Greg