[patch 6/7] s390: ipl device.
From: Heiko Carstens <[email protected]>
Export the ipl device settings to userspace via the sysfs:
* /sys/kernel/ipl_device
Contains a string in on of the following formats:
1) "ccw <bus_id>", or 2) "fcp <bus_id>,<wwpn>,<lun>".
* /sys/kernel/ipl_parameter
is a binary interface that exports the ipl parameter block for
scsi ipl. For non-scsi ipl the ipl_paramter is irrelevant.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
diffstat:
arch/s390/kernel/head.S | 72 ++++++++++++++++++++++++++++++++++++++++++----
arch/s390/kernel/head64.S | 66 ++++++++++++++++++++++++++++++++++++++++--
arch/s390/kernel/setup.c | 67 ++++++++++++++++++++++++++++++++++++++++++
include/asm-s390/setup.h | 44 ++++++++++++++++++++++++++++
4 files changed, 242 insertions(+), 7 deletions(-)
diff -urpN linux-2.6/arch/s390/kernel/head64.S linux-2.6-patched/arch/s390/kernel/head64.S
--- linux-2.6/arch/s390/kernel/head64.S 2005-09-14 16:47:51.000000000 +0200
+++ linux-2.6-patched/arch/s390/kernel/head64.S 2005-09-14 16:48:18.000000000 +0200
@@ -484,6 +484,8 @@ start:
startup:basr %r13,0 # get base
.LPG1: sll %r13,1 # remove high order bit
srl %r13,1
+ l %r1,.Lget_ipl_device_addr-.LPG1(%r13)
+ basr %r14,%r1
lhi %r1,1 # mode 1 = esame
slr %r0,%r0 # set cpuid to zero
sigp %r1,%r0,0x12 # switch to esame mode
@@ -556,6 +558,9 @@ startup:basr %r13,0
mlgr %r2,%r1 # mem size in bytes in %r3
b .Lfchunk-.LPG1(%r13)
+ .align 4
+.Lget_ipl_device_addr:
+ .long .Lget_ipl_device
.Lpmask:
.byte 0
.align 8
@@ -746,6 +751,63 @@ _pstart:
.global _pend
_pend:
+.Lget_ipl_device:
+ basr %r12,0
+.LPG2: l %r1,0xb8 # get sid
+ sll %r1,15 # test if subchannel is enabled
+ srl %r1,31
+ ltr %r1,%r1
+ bz 0(%r14) # subchannel disabled
+ l %r1,0xb8
+ la %r5,.Lipl_schib-.LPG2(%r12)
+ stsch 0(%r5) # get schib of subchannel
+ bnz 0(%r14) # schib not available
+ tm 5(%r5),0x01 # devno valid?
+ bno 0(%r14)
+ la %r6,ipl_parameter_flags-.LPG2(%r12)
+ oi 3(%r6),0x01 # set flag
+ la %r2,ipl_devno-.LPG2(%r12)
+ mvc 0(2,%r2),6(%r5) # store devno
+ tm 4(%r5),0x80 # qdio capable device?
+ bno 0(%r14)
+ oi 3(%r6),0x02 # set flag
+
+ # copy ipl parameters
+
+ lhi %r0,4096
+ l %r2,20(%r0) # get address of parameter list
+ lhi %r3,IPL_PARMBLOCK_ORIGIN
+ st %r3,20(%r0)
+ lhi %r4,1
+ cr %r2,%r3 # start parameters < destination ?
+ jl 0f
+ lhi %r1,1 # copy direction is upwards
+ j 1f
+0: lhi %r1,-1 # copy direction is downwards
+ ar %r2,%r0
+ ar %r3,%r0
+ ar %r2,%r1
+ ar %r3,%r1
+1: mvc 0(1,%r3),0(%r2) # finally copy ipl parameters
+ ar %r3,%r1
+ ar %r2,%r1
+ sr %r0,%r4
+ jne 1b
+ b 0(%r14)
+
+ .align 4
+.Lipl_schib:
+ .rept 13
+ .long 0
+ .endr
+
+ .globl ipl_parameter_flags
+ipl_parameter_flags:
+ .long 0
+ .globl ipl_devno
+ipl_devno:
+ .word 0
+
#ifdef CONFIG_SHARED_KERNEL
.org 0x100000
#endif
@@ -755,7 +817,7 @@ _pend:
#
.globl _stext
_stext: basr %r13,0 # get base
-.LPG2:
+.LPG3:
#
# Setup stack
#
@@ -774,7 +836,7 @@ _stext: basr %r13,0
lctlg %c0,%c15,0(%r15)
#
- lam 0,15,.Laregs-.LPG2(%r13) # load access regs needed by uaccess
+ lam 0,15,.Laregs-.LPG3(%r13) # load access regs needed by uaccess
brasl %r14,start_kernel # go to C code
#
# We returned from start_kernel ?!? PANIK
diff -urpN linux-2.6/arch/s390/kernel/head.S linux-2.6-patched/arch/s390/kernel/head.S
--- linux-2.6/arch/s390/kernel/head.S 2005-09-14 16:47:51.000000000 +0200
+++ linux-2.6-patched/arch/s390/kernel/head.S 2005-09-14 16:48:18.000000000 +0200
@@ -485,7 +485,9 @@ start:
#
.org 0x10000
startup:basr %r13,0 # get base
-.LPG1: lctl %c0,%c15,.Lctl-.LPG1(%r13) # load control registers
+.LPG1: l %r1, .Lget_ipl_device_addr-.LPG1(%r13)
+ basr %r14, %r1
+ lctl %c0,%c15,.Lctl-.LPG1(%r13) # load control registers
la %r12,_pstart-.LPG1(%r13) # pointer to parameter area
# move IPL device to lowcore
mvc __LC_IPLDEV(4),IPL_DEVICE-PARMAREA(%r12)
@@ -560,6 +562,9 @@ startup:basr %r13,0
mr %r2,%r1 # mem size in bytes in %r3
b .Lfchunk-.LPG1(%r13)
+ .align 4
+.Lget_ipl_device_addr:
+ .long .Lget_ipl_device
.Lpmask:
.byte 0
.align 8
@@ -755,6 +760,63 @@ _pstart:
.global _pend
_pend:
+.Lget_ipl_device:
+ basr %r12,0
+.LPG2: l %r1,0xb8 # get sid
+ sll %r1,15 # test if subchannel is enabled
+ srl %r1,31
+ ltr %r1,%r1
+ bz 0(%r14) # subchannel disabled
+ l %r1,0xb8
+ la %r5,.Lipl_schib-.LPG2(%r12)
+ stsch 0(%r5) # get schib of subchannel
+ bnz 0(%r14) # schib not available
+ tm 5(%r5),0x01 # devno valid?
+ bno 0(%r14)
+ la %r6,ipl_parameter_flags-.LPG2(%r12)
+ oi 3(%r6),0x01 # set flag
+ la %r2,ipl_devno-.LPG2(%r12)
+ mvc 0(2,%r2),6(%r5) # store devno
+ tm 4(%r5),0x80 # qdio capable device?
+ bno 0(%r14)
+ oi 3(%r6),0x02 # set flag
+
+ # copy ipl parameters
+
+ lhi %r0,4096
+ l %r2,20(%r0) # get address of parameter list
+ lhi %r3,IPL_PARMBLOCK_ORIGIN
+ st %r3,20(%r0)
+ lhi %r4,1
+ cr %r2,%r3 # start parameters < destination ?
+ jl 0f
+ lhi %r1,1 # copy direction is upwards
+ j 1f
+0: lhi %r1,-1 # copy direction is downwards
+ ar %r2,%r0
+ ar %r3,%r0
+ ar %r2,%r1
+ ar %r3,%r1
+1: mvc 0(1,%r3),0(%r2) # finally copy ipl parameters
+ ar %r3,%r1
+ ar %r2,%r1
+ sr %r0,%r4
+ jne 1b
+ b 0(%r14)
+
+ .align 4
+.Lipl_schib:
+ .rept 13
+ .long 0
+ .endr
+
+ .globl ipl_parameter_flags
+ipl_parameter_flags:
+ .long 0
+ .globl ipl_devno
+ipl_devno:
+ .word 0
+
#ifdef CONFIG_SHARED_KERNEL
.org 0x100000
#endif
@@ -764,11 +826,11 @@ _pend:
#
.globl _stext
_stext: basr %r13,0 # get base
-.LPG2:
+.LPG3:
#
# Setup stack
#
- l %r15,.Linittu-.LPG2(%r13)
+ l %r15,.Linittu-.LPG3(%r13)
mvc __LC_CURRENT(4),__TI_task(%r15)
ahi %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union + THREAD_SIZE
st %r15,__LC_KERNEL_STACK # set end of kernel stack
@@ -782,8 +844,8 @@ _stext: basr %r13,0
lctl %c0,%c15,0(%r15)
#
- lam 0,15,.Laregs-.LPG2(%r13) # load access regs needed by uaccess
- l %r14,.Lstart-.LPG2(%r13)
+ lam 0,15,.Laregs-.LPG3(%r13) # load access regs needed by uaccess
+ l %r14,.Lstart-.LPG3(%r13)
basr %r14,%r14 # call start_kernel
#
# We returned from start_kernel ?!? PANIK
diff -urpN linux-2.6/arch/s390/kernel/setup.c linux-2.6-patched/arch/s390/kernel/setup.c
--- linux-2.6/arch/s390/kernel/setup.c 2005-09-14 16:48:18.000000000 +0200
+++ linux-2.6-patched/arch/s390/kernel/setup.c 2005-09-14 16:48:18.000000000 +0200
@@ -682,3 +682,70 @@ struct seq_operations cpuinfo_op = {
.show = show_cpuinfo,
};
+#ifdef CONFIG_SYSFS
+
+static ssize_t
+ipl_parameter_read(struct kobject *kobj, char *buf, loff_t off, size_t count)
+{
+ unsigned int size;
+
+ if (!IPL_PARMBLOCK_VALID)
+ return 0;
+
+ size = IPL_PARMBLOCK_SIZE;
+
+ if (off > size)
+ return 0;
+ if (off + count > size)
+ count = size - off;
+
+ memcpy(buf, (void *) IPL_PARMBLOCK_START + off, count);
+ return count;
+}
+
+static struct bin_attribute ipl_parameter_attr = {
+ .attr = {
+ .name = "ipl_parameter",
+ .mode = S_IRUGO,
+ .owner = THIS_MODULE,
+ },
+ .size = PAGE_SIZE,
+ .read = &ipl_parameter_read,
+};
+
+static ssize_t
+ipl_device_show(struct subsystem *subsys, char *page)
+{
+ struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
+
+ if (!IPL_DEVNO_VALID)
+ goto type_unknown;
+ if (!IPL_PARMBLOCK_VALID)
+ goto type_ccw;
+ if (ipl->hdr.header.version > IPL_MAX_SUPPORTED_VERSION)
+ goto type_unknown;
+ if (ipl->fcp.pbt != IPL_TYPE_FCP)
+ goto type_unknown;
+
+ return sprintf(page, "fcp 0.0.%04x,0x%016llx,0x%016llx\n",
+ ipl->fcp.devno,
+ (unsigned long long) ipl->fcp.wwpn,
+ (unsigned long long) ipl->fcp.lun);
+ type_unknown:
+ return sprintf(page, "unknown\n");
+ type_ccw:
+ return sprintf(page, "ccw 0.0.%04x\n",ipl_devno);
+}
+
+static struct subsys_attribute ipl_device_attr = __ATTR_RO(ipl_device);
+
+static int __init
+ipl_device_sysfs_register(void) {
+ sysfs_create_file(&kernel_subsys.kset.kobj, &ipl_device_attr.attr);
+ sysfs_create_bin_file(&kernel_subsys.kset.kobj, &ipl_parameter_attr);
+ return 0;
+}
+
+__initcall(ipl_device_sysfs_register);
+
+#endif /* CONFIG_SYSFS */
diff -urpN linux-2.6/include/asm-s390/setup.h linux-2.6-patched/include/asm-s390/setup.h
--- linux-2.6/include/asm-s390/setup.h 2005-08-29 01:41:01.000000000 +0200
+++ linux-2.6-patched/include/asm-s390/setup.h 2005-09-14 16:48:18.000000000 +0200
@@ -8,11 +8,14 @@
#ifndef _ASM_S390_SETUP_H
#define _ASM_S390_SETUP_H
+#include <asm/types.h>
+
#define PARMAREA 0x10400
#define COMMAND_LINE_SIZE 896
#define RAMDISK_ORIGIN 0x800000
#define RAMDISK_SIZE 0x800000
#define MEMORY_CHUNKS 16 /* max 0x7fff */
+#define IPL_PARMBLOCK_ORIGIN 0x2000
#ifndef __ASSEMBLY__
@@ -64,6 +67,47 @@ extern unsigned int console_irq;
#define SET_CONSOLE_3215 do { console_mode = 2; } while (0)
#define SET_CONSOLE_3270 do { console_mode = 3; } while (0)
+struct ipl_list_header {
+ u32 length;
+ u8 reserved[3];
+ u8 version;
+} __attribute__((packed));
+
+struct ipl_block_fcp {
+ u32 length;
+ u8 pbt;
+ u8 reserved1[322-1];
+ u16 devno;
+ u8 reserved2[4];
+ u64 wwpn;
+ u64 lun;
+} __attribute__((packed));
+
+struct ipl_parameter_block {
+ union {
+ u32 length;
+ struct ipl_list_header header;
+ } hdr;
+ struct ipl_block_fcp fcp;
+} __attribute__((packed));
+
+#define IPL_MAX_SUPPORTED_VERSION (0)
+
+#define IPL_TYPE_FCP (0)
+
+/*
+ * IPL validity flags and parameters as detected in head.S
+ */
+extern u32 ipl_parameter_flags;
+extern u16 ipl_devno;
+
+#define IPL_DEVNO_VALID (ipl_parameter_flags & 1)
+#define IPL_PARMBLOCK_VALID (ipl_parameter_flags & 2)
+
+#define IPL_PARMBLOCK_START ((struct ipl_parameter_block *) \
+ IPL_PARMBLOCK_ORIGIN)
+#define IPL_PARMBLOCK_SIZE (IPL_PARMBLOCK_START->hdr.length)
+
#else
#ifndef __s390x__
On Wed, Sep 14, 2005 at 05:55:09PM +0200, Martin Schwidefsky wrote:
> [patch 6/7] s390: ipl device.
>
> From: Heiko Carstens <[email protected]>
>
> Export the ipl device settings to userspace via the sysfs:
> * /sys/kernel/ipl_device
What? Why that location? Why not in the proper location for your
device, on your bus?
> Contains a string in on of the following formats:
> 1) "ccw <bus_id>", or 2) "fcp <bus_id>,<wwpn>,<lun>".
> * /sys/kernel/ipl_parameter
> is a binary interface that exports the ipl parameter block for
> scsi ipl. For non-scsi ipl the ipl_paramter is irrelevant.
Again, put this in your device directory, not in /sys/kernel/
> +static ssize_t
> +ipl_device_show(struct subsystem *subsys, char *page)
> +{
> + struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
> +
> + if (!IPL_DEVNO_VALID)
> + goto type_unknown;
> + if (!IPL_PARMBLOCK_VALID)
> + goto type_ccw;
> + if (ipl->hdr.header.version > IPL_MAX_SUPPORTED_VERSION)
> + goto type_unknown;
> + if (ipl->fcp.pbt != IPL_TYPE_FCP)
> + goto type_unknown;
> +
> + return sprintf(page, "fcp 0.0.%04x,0x%016llx,0x%016llx\n",
> + ipl->fcp.devno,
> + (unsigned long long) ipl->fcp.wwpn,
> + (unsigned long long) ipl->fcp.lun);
> + type_unknown:
> + return sprintf(page, "unknown\n");
> + type_ccw:
> + return sprintf(page, "ccw 0.0.%04x\n",ipl_devno);
That doesn't look like a "single value" from a single file there. Can't
you break that up into individual files, based on what exactly is
present at the time?
thanks,
greg k-h
> > Export the ipl device settings to userspace via the sysfs:
> > * /sys/kernel/ipl_device
> What? Why that location? Why not in the proper location for your
> device, on your bus?
This interface tells from where the kernel was booted from. I don't
think a device should have an attribute where the meaning would be
"the current running kernel came via this device into memory".
IMHO this should be an attribute of the kernel and therefore I
thought /sys/kernel would be a good idea.
> > Contains a string in on of the following formats:
> > 1) "ccw <bus_id>", or 2) "fcp <bus_id>,<wwpn>,<lun>".
> > * /sys/kernel/ipl_parameter
> > is a binary interface that exports the ipl parameter block for
> > scsi ipl. For non-scsi ipl the ipl_paramter is irrelevant.
> Again, put this in your device directory, not in /sys/kernel/
Same here.
> > +static ssize_t
> > +ipl_device_show(struct subsystem *subsys, char *page)
> > +{
> > + struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
> > +
> > + if (!IPL_DEVNO_VALID)
> > + goto type_unknown;
> > + if (!IPL_PARMBLOCK_VALID)
> > + goto type_ccw;
> > + if (ipl->hdr.header.version > IPL_MAX_SUPPORTED_VERSION)
> > + goto type_unknown;
> > + if (ipl->fcp.pbt != IPL_TYPE_FCP)
> > + goto type_unknown;
> > +
> > + return sprintf(page, "fcp 0.0.%04x,0x%016llx,0x%016llx\n",
> > + ipl->fcp.devno,
> > + (unsigned long long) ipl->fcp.wwpn,
> > + (unsigned long long) ipl->fcp.lun);
> > + type_unknown:
> > + return sprintf(page, "unknown\n");
> > + type_ccw:
> > + return sprintf(page, "ccw 0.0.%04x\n",ipl_devno);
>
> That doesn't look like a "single value" from a single file there. Can't
> you break that up into individual files, based on what exactly is
> present at the time?
Sure, so I would end up with quite a few files:
always a file which tells the type of ipl e.g. ipl_type and dependent
on that additionally:
- in case of ccw ipl:
just the bus_id of the ipl device (e.g. ipl_bus_id).
- in case of fcp ipl:
the bus_id, the wwpn and the fcp_lun all exported via different files.
Does that sound reasonable?
Heiko
On Fri, 2005-09-16 at 09:14 +0200, Heiko Carstens wrote:
> > > Export the ipl device settings to userspace via the sysfs:
> > > * /sys/kernel/ipl_device
> > What? Why that location? Why not in the proper location for your
> > device, on your bus?
>
> This interface tells from where the kernel was booted from. I don't
> think a device should have an attribute where the meaning would be
> "the current running kernel came via this device into memory".
> IMHO this should be an attribute of the kernel and therefore I
> thought /sys/kernel would be a good idea.
If the additional ipl information is bound to the ipl device then we'd
have to search for the device if we'd want to get the ipl information.
And if we ever want to make /sys/kernel/ipl_device writable to be able
to change the ipl_device for a reboot then what? Have the ipl related
sysfs files for ALL devices you can ipl from to be able to move to
another device? I think we need a central place for this information.
--
blue skies,
Martin
Martin Schwidefsky
Linux for zSeries Development & Services
IBM Deutschland Entwicklung GmbH
On Fri, Sep 16, 2005 at 09:14:44AM +0200, Heiko Carstens wrote:
> > > Export the ipl device settings to userspace via the sysfs:
> > > * /sys/kernel/ipl_device
> > What? Why that location? Why not in the proper location for your
> > device, on your bus?
>
> This interface tells from where the kernel was booted from.
Then why not use /sys/firmware/ipl/ ? That matches the
/sys/firmware/edd usage we currently have on x86 boxes.
> > > +static ssize_t
> > > +ipl_device_show(struct subsystem *subsys, char *page)
> > > +{
> > > + struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
> > > +
> > > + if (!IPL_DEVNO_VALID)
> > > + goto type_unknown;
> > > + if (!IPL_PARMBLOCK_VALID)
> > > + goto type_ccw;
> > > + if (ipl->hdr.header.version > IPL_MAX_SUPPORTED_VERSION)
> > > + goto type_unknown;
> > > + if (ipl->fcp.pbt != IPL_TYPE_FCP)
> > > + goto type_unknown;
> > > +
> > > + return sprintf(page, "fcp 0.0.%04x,0x%016llx,0x%016llx\n",
> > > + ipl->fcp.devno,
> > > + (unsigned long long) ipl->fcp.wwpn,
> > > + (unsigned long long) ipl->fcp.lun);
> > > + type_unknown:
> > > + return sprintf(page, "unknown\n");
> > > + type_ccw:
> > > + return sprintf(page, "ccw 0.0.%04x\n",ipl_devno);
> >
> > That doesn't look like a "single value" from a single file there. Can't
> > you break that up into individual files, based on what exactly is
> > present at the time?
>
> Sure, so I would end up with quite a few files:
That's fine to have.
> always a file which tells the type of ipl e.g. ipl_type and dependent
> on that additionally:
> - in case of ccw ipl:
> just the bus_id of the ipl device (e.g. ipl_bus_id).
> - in case of fcp ipl:
> the bus_id, the wwpn and the fcp_lun all exported via different files.
>
> Does that sound reasonable?
Yes.
thanks,
greg k-h
> > > > Export the ipl device settings to userspace via the sysfs:
> > > > * /sys/kernel/ipl_device
> > > What? Why that location? Why not in the proper location for your
> > > device, on your bus?
> >
> > This interface tells from where the kernel was booted from.
>
> Then why not use /sys/firmware/ipl/ ? That matches the
> /sys/firmware/edd usage we currently have on x86 boxes.
Ok, I will do the changes. Thanks for reviewing!
Heiko