Add the rt linux 4.1.3-rt3 as base
[kvmfornfv.git] / kernel / drivers / s390 / char / zcore.c
diff --git a/kernel/drivers/s390/char/zcore.c b/kernel/drivers/s390/char/zcore.c
new file mode 100644 (file)
index 0000000..a68fcfd
--- /dev/null
@@ -0,0 +1,750 @@
+/*
+ * zcore module to export memory content and register sets for creating system
+ * dumps on SCSI disks (zfcpdump). The "zcore/mem" debugfs file shows the same
+ * dump format as s390 standalone dumps.
+ *
+ * For more information please refer to Documentation/s390/zfcpdump.txt
+ *
+ * Copyright IBM Corp. 2003, 2008
+ * Author(s): Michael Holzheu
+ */
+
+#define KMSG_COMPONENT "zdump"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/miscdevice.h>
+#include <linux/debugfs.h>
+#include <linux/module.h>
+#include <linux/memblock.h>
+
+#include <asm/asm-offsets.h>
+#include <asm/ipl.h>
+#include <asm/sclp.h>
+#include <asm/setup.h>
+#include <asm/uaccess.h>
+#include <asm/debug.h>
+#include <asm/processor.h>
+#include <asm/irqflags.h>
+#include <asm/checksum.h>
+#include <asm/switch_to.h>
+#include "sclp.h"
+
+#define TRACE(x...) debug_sprintf_event(zcore_dbf, 1, x)
+
+#define TO_USER                1
+#define TO_KERNEL      0
+#define CHUNK_INFO_SIZE        34 /* 2 16-byte char, each followed by blank */
+
+enum arch_id {
+       ARCH_S390       = 0,
+       ARCH_S390X      = 1,
+};
+
+/* dump system info */
+
+struct sys_info {
+       enum arch_id     arch;
+       unsigned long    sa_base;
+       u32              sa_size;
+       int              cpu_map[NR_CPUS];
+       unsigned long    mem_size;
+       struct save_area lc_mask;
+};
+
+struct ipib_info {
+       unsigned long   ipib;
+       u32             checksum;
+}  __attribute__((packed));
+
+static struct sys_info sys_info;
+static struct debug_info *zcore_dbf;
+static int hsa_available;
+static struct dentry *zcore_dir;
+static struct dentry *zcore_file;
+static struct dentry *zcore_memmap_file;
+static struct dentry *zcore_reipl_file;
+static struct dentry *zcore_hsa_file;
+static struct ipl_parameter_block *ipl_block;
+
+/*
+ * Copy memory from HSA to kernel or user memory (not reentrant):
+ *
+ * @dest:  Kernel or user buffer where memory should be copied to
+ * @src:   Start address within HSA where data should be copied
+ * @count: Size of buffer, which should be copied
+ * @mode:  Either TO_KERNEL or TO_USER
+ */
+int memcpy_hsa(void *dest, unsigned long src, size_t count, int mode)
+{
+       int offs, blk_num;
+       static char buf[PAGE_SIZE] __attribute__((__aligned__(PAGE_SIZE)));
+
+       if (!hsa_available)
+               return -ENODATA;
+       if (count == 0)
+               return 0;
+
+       /* copy first block */
+       offs = 0;
+       if ((src % PAGE_SIZE) != 0) {
+               blk_num = src / PAGE_SIZE + 2;
+               if (sclp_sdias_copy(buf, blk_num, 1)) {
+                       TRACE("sclp_sdias_copy() failed\n");
+                       return -EIO;
+               }
+               offs = min((PAGE_SIZE - (src % PAGE_SIZE)), count);
+               if (mode == TO_USER) {
+                       if (copy_to_user((__force __user void*) dest,
+                                        buf + (src % PAGE_SIZE), offs))
+                               return -EFAULT;
+               } else
+                       memcpy(dest, buf + (src % PAGE_SIZE), offs);
+       }
+       if (offs == count)
+               goto out;
+
+       /* copy middle */
+       for (; (offs + PAGE_SIZE) <= count; offs += PAGE_SIZE) {
+               blk_num = (src + offs) / PAGE_SIZE + 2;
+               if (sclp_sdias_copy(buf, blk_num, 1)) {
+                       TRACE("sclp_sdias_copy() failed\n");
+                       return -EIO;
+               }
+               if (mode == TO_USER) {
+                       if (copy_to_user((__force __user void*) dest + offs,
+                                        buf, PAGE_SIZE))
+                               return -EFAULT;
+               } else
+                       memcpy(dest + offs, buf, PAGE_SIZE);
+       }
+       if (offs == count)
+               goto out;
+
+       /* copy last block */
+       blk_num = (src + offs) / PAGE_SIZE + 2;
+       if (sclp_sdias_copy(buf, blk_num, 1)) {
+               TRACE("sclp_sdias_copy() failed\n");
+               return -EIO;
+       }
+       if (mode == TO_USER) {
+               if (copy_to_user((__force __user void*) dest + offs, buf,
+                                count - offs))
+                       return -EFAULT;
+       } else
+               memcpy(dest + offs, buf, count - offs);
+out:
+       return 0;
+}
+
+static int memcpy_hsa_user(void __user *dest, unsigned long src, size_t count)
+{
+       return memcpy_hsa((void __force *) dest, src, count, TO_USER);
+}
+
+static int memcpy_hsa_kernel(void *dest, unsigned long src, size_t count)
+{
+       return memcpy_hsa(dest, src, count, TO_KERNEL);
+}
+
+static int __init init_cpu_info(enum arch_id arch)
+{
+       struct save_area_ext *sa_ext;
+
+       /* get info for boot cpu from lowcore, stored in the HSA */
+
+       sa_ext = dump_save_area_create(0);
+       if (!sa_ext)
+               return -ENOMEM;
+       if (memcpy_hsa_kernel(&sa_ext->sa, sys_info.sa_base,
+                             sys_info.sa_size) < 0) {
+               TRACE("could not copy from HSA\n");
+               kfree(sa_ext);
+               return -EIO;
+       }
+       if (MACHINE_HAS_VX)
+               save_vx_regs_safe(sa_ext->vx_regs);
+       return 0;
+}
+
+static DEFINE_MUTEX(zcore_mutex);
+
+#define DUMP_VERSION   0x5
+#define DUMP_MAGIC     0xa8190173618f23fdULL
+#define DUMP_ARCH_S390X        2
+#define DUMP_ARCH_S390 1
+#define HEADER_SIZE    4096
+
+/* dump header dumped according to s390 crash dump format */
+
+struct zcore_header {
+       u64 magic;
+       u32 version;
+       u32 header_size;
+       u32 dump_level;
+       u32 page_size;
+       u64 mem_size;
+       u64 mem_start;
+       u64 mem_end;
+       u32 num_pages;
+       u32 pad1;
+       u64 tod;
+       struct cpuid cpu_id;
+       u32 arch_id;
+       u32 volnr;
+       u32 build_arch;
+       u64 rmem_size;
+       u8 mvdump;
+       u16 cpu_cnt;
+       u16 real_cpu_cnt;
+       u8 end_pad1[0x200-0x061];
+       u64 mvdump_sign;
+       u64 mvdump_zipl_time;
+       u8 end_pad2[0x800-0x210];
+       u32 lc_vec[512];
+} __attribute__((packed,__aligned__(16)));
+
+static struct zcore_header zcore_header = {
+       .magic          = DUMP_MAGIC,
+       .version        = DUMP_VERSION,
+       .header_size    = 4096,
+       .dump_level     = 0,
+       .page_size      = PAGE_SIZE,
+       .mem_start      = 0,
+       .build_arch     = DUMP_ARCH_S390X,
+};
+
+/*
+ * Copy lowcore info to buffer. Use map in order to copy only register parts.
+ *
+ * @buf:    User buffer
+ * @sa:     Pointer to save area
+ * @sa_off: Offset in save area to copy
+ * @len:    Number of bytes to copy
+ */
+static int copy_lc(void __user *buf, void *sa, int sa_off, int len)
+{
+       int i;
+       char *lc_mask = (char*)&sys_info.lc_mask;
+
+       for (i = 0; i < len; i++) {
+               if (!lc_mask[i + sa_off])
+                       continue;
+               if (copy_to_user(buf + i, sa + sa_off + i, 1))
+                       return -EFAULT;
+       }
+       return 0;
+}
+
+/*
+ * Copy lowcores info to memory, if necessary
+ *
+ * @buf:   User buffer
+ * @addr:  Start address of buffer in dump memory
+ * @count: Size of buffer
+ */
+static int zcore_add_lc(char __user *buf, unsigned long start, size_t count)
+{
+       unsigned long end;
+       int i;
+
+       if (count == 0)
+               return 0;
+
+       end = start + count;
+       for (i = 0; i < dump_save_areas.count; i++) {
+               unsigned long cp_start, cp_end; /* copy range */
+               unsigned long sa_start, sa_end; /* save area range */
+               unsigned long prefix;
+               unsigned long sa_off, len, buf_off;
+               struct save_area *save_area = &dump_save_areas.areas[i]->sa;
+
+               prefix = save_area->pref_reg;
+               sa_start = prefix + sys_info.sa_base;
+               sa_end = prefix + sys_info.sa_base + sys_info.sa_size;
+
+               if ((end < sa_start) || (start > sa_end))
+                       continue;
+               cp_start = max(start, sa_start);
+               cp_end = min(end, sa_end);
+
+               buf_off = cp_start - start;
+               sa_off = cp_start - sa_start;
+               len = cp_end - cp_start;
+
+               TRACE("copy_lc for: %lx\n", start);
+               if (copy_lc(buf + buf_off, save_area, sa_off, len))
+                       return -EFAULT;
+       }
+       return 0;
+}
+
+/*
+ * Release the HSA
+ */
+static void release_hsa(void)
+{
+       diag308(DIAG308_REL_HSA, NULL);
+       hsa_available = 0;
+}
+
+/*
+ * Read routine for zcore character device
+ * First 4K are dump header
+ * Next 32MB are HSA Memory
+ * Rest is read from absolute Memory
+ */
+static ssize_t zcore_read(struct file *file, char __user *buf, size_t count,
+                         loff_t *ppos)
+{
+       unsigned long mem_start; /* Start address in memory */
+       size_t mem_offs;         /* Offset in dump memory */
+       size_t hdr_count;        /* Size of header part of output buffer */
+       size_t size;
+       int rc;
+
+       mutex_lock(&zcore_mutex);
+
+       if (*ppos > (sys_info.mem_size + HEADER_SIZE)) {
+               rc = -EINVAL;
+               goto fail;
+       }
+
+       count = min(count, (size_t) (sys_info.mem_size + HEADER_SIZE - *ppos));
+
+       /* Copy dump header */
+       if (*ppos < HEADER_SIZE) {
+               size = min(count, (size_t) (HEADER_SIZE - *ppos));
+               if (copy_to_user(buf, &zcore_header + *ppos, size)) {
+                       rc = -EFAULT;
+                       goto fail;
+               }
+               hdr_count = size;
+               mem_start = 0;
+       } else {
+               hdr_count = 0;
+               mem_start = *ppos - HEADER_SIZE;
+       }
+
+       mem_offs = 0;
+
+       /* Copy from HSA data */
+       if (*ppos < sclp_get_hsa_size() + HEADER_SIZE) {
+               size = min((count - hdr_count),
+                          (size_t) (sclp_get_hsa_size() - mem_start));
+               rc = memcpy_hsa_user(buf + hdr_count, mem_start, size);
+               if (rc)
+                       goto fail;
+
+               mem_offs += size;
+       }
+
+       /* Copy from real mem */
+       size = count - mem_offs - hdr_count;
+       rc = copy_to_user_real(buf + hdr_count + mem_offs,
+                              (void *) mem_start + mem_offs, size);
+       if (rc)
+               goto fail;
+
+       /*
+        * Since s390 dump analysis tools like lcrash or crash
+        * expect register sets in the prefix pages of the cpus,
+        * we copy them into the read buffer, if necessary.
+        * buf + hdr_count: Start of memory part of output buffer
+        * mem_start: Start memory address to copy from
+        * count - hdr_count: Size of memory area to copy
+        */
+       if (zcore_add_lc(buf + hdr_count, mem_start, count - hdr_count)) {
+               rc = -EFAULT;
+               goto fail;
+       }
+       *ppos += count;
+fail:
+       mutex_unlock(&zcore_mutex);
+       return (rc < 0) ? rc : count;
+}
+
+static int zcore_open(struct inode *inode, struct file *filp)
+{
+       if (!hsa_available)
+               return -ENODATA;
+       else
+               return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
+}
+
+static int zcore_release(struct inode *inode, struct file *filep)
+{
+       if (hsa_available)
+               release_hsa();
+       return 0;
+}
+
+static loff_t zcore_lseek(struct file *file, loff_t offset, int orig)
+{
+       loff_t rc;
+
+       mutex_lock(&zcore_mutex);
+       switch (orig) {
+       case 0:
+               file->f_pos = offset;
+               rc = file->f_pos;
+               break;
+       case 1:
+               file->f_pos += offset;
+               rc = file->f_pos;
+               break;
+       default:
+               rc = -EINVAL;
+       }
+       mutex_unlock(&zcore_mutex);
+       return rc;
+}
+
+static const struct file_operations zcore_fops = {
+       .owner          = THIS_MODULE,
+       .llseek         = zcore_lseek,
+       .read           = zcore_read,
+       .open           = zcore_open,
+       .release        = zcore_release,
+};
+
+static ssize_t zcore_memmap_read(struct file *filp, char __user *buf,
+                                size_t count, loff_t *ppos)
+{
+       return simple_read_from_buffer(buf, count, ppos, filp->private_data,
+                                      memblock.memory.cnt * CHUNK_INFO_SIZE);
+}
+
+static int zcore_memmap_open(struct inode *inode, struct file *filp)
+{
+       struct memblock_region *reg;
+       char *buf;
+       int i = 0;
+
+       buf = kzalloc(memblock.memory.cnt * CHUNK_INFO_SIZE, GFP_KERNEL);
+       if (!buf) {
+               return -ENOMEM;
+       }
+       for_each_memblock(memory, reg) {
+               sprintf(buf + (i++ * CHUNK_INFO_SIZE), "%016llx %016llx ",
+                       (unsigned long long) reg->base,
+                       (unsigned long long) reg->size);
+       }
+       filp->private_data = buf;
+       return nonseekable_open(inode, filp);
+}
+
+static int zcore_memmap_release(struct inode *inode, struct file *filp)
+{
+       kfree(filp->private_data);
+       return 0;
+}
+
+static const struct file_operations zcore_memmap_fops = {
+       .owner          = THIS_MODULE,
+       .read           = zcore_memmap_read,
+       .open           = zcore_memmap_open,
+       .release        = zcore_memmap_release,
+       .llseek         = no_llseek,
+};
+
+static ssize_t zcore_reipl_write(struct file *filp, const char __user *buf,
+                                size_t count, loff_t *ppos)
+{
+       if (ipl_block) {
+               diag308(DIAG308_SET, ipl_block);
+               diag308(DIAG308_IPL, NULL);
+       }
+       return count;
+}
+
+static int zcore_reipl_open(struct inode *inode, struct file *filp)
+{
+       return nonseekable_open(inode, filp);
+}
+
+static int zcore_reipl_release(struct inode *inode, struct file *filp)
+{
+       return 0;
+}
+
+static const struct file_operations zcore_reipl_fops = {
+       .owner          = THIS_MODULE,
+       .write          = zcore_reipl_write,
+       .open           = zcore_reipl_open,
+       .release        = zcore_reipl_release,
+       .llseek         = no_llseek,
+};
+
+static ssize_t zcore_hsa_read(struct file *filp, char __user *buf,
+                             size_t count, loff_t *ppos)
+{
+       static char str[18];
+
+       if (hsa_available)
+               snprintf(str, sizeof(str), "%lx\n", sclp_get_hsa_size());
+       else
+               snprintf(str, sizeof(str), "0\n");
+       return simple_read_from_buffer(buf, count, ppos, str, strlen(str));
+}
+
+static ssize_t zcore_hsa_write(struct file *filp, const char __user *buf,
+                              size_t count, loff_t *ppos)
+{
+       char value;
+
+       if (*ppos != 0)
+               return -EPIPE;
+       if (copy_from_user(&value, buf, 1))
+               return -EFAULT;
+       if (value != '0')
+               return -EINVAL;
+       release_hsa();
+       return count;
+}
+
+static const struct file_operations zcore_hsa_fops = {
+       .owner          = THIS_MODULE,
+       .write          = zcore_hsa_write,
+       .read           = zcore_hsa_read,
+       .open           = nonseekable_open,
+       .llseek         = no_llseek,
+};
+
+static void __init set_lc_mask(struct save_area *map)
+{
+       memset(&map->fp_regs, 0xff, sizeof(map->fp_regs));
+       memset(&map->gp_regs, 0xff, sizeof(map->gp_regs));
+       memset(&map->psw, 0xff, sizeof(map->psw));
+       memset(&map->pref_reg, 0xff, sizeof(map->pref_reg));
+       memset(&map->fp_ctrl_reg, 0xff, sizeof(map->fp_ctrl_reg));
+       memset(&map->tod_reg, 0xff, sizeof(map->tod_reg));
+       memset(&map->timer, 0xff, sizeof(map->timer));
+       memset(&map->clk_cmp, 0xff, sizeof(map->clk_cmp));
+       memset(&map->acc_regs, 0xff, sizeof(map->acc_regs));
+       memset(&map->ctrl_regs, 0xff, sizeof(map->ctrl_regs));
+}
+
+/*
+ * Initialize dump globals for a given architecture
+ */
+static int __init sys_info_init(enum arch_id arch, unsigned long mem_end)
+{
+       int rc;
+
+       switch (arch) {
+       case ARCH_S390X:
+               pr_alert("DETECTED 'S390X (64 bit) OS'\n");
+               break;
+       case ARCH_S390:
+               pr_alert("DETECTED 'S390 (32 bit) OS'\n");
+               break;
+       default:
+               pr_alert("0x%x is an unknown architecture.\n",arch);
+               return -EINVAL;
+       }
+       sys_info.sa_base = SAVE_AREA_BASE;
+       sys_info.sa_size = sizeof(struct save_area);
+       sys_info.arch = arch;
+       set_lc_mask(&sys_info.lc_mask);
+       rc = init_cpu_info(arch);
+       if (rc)
+               return rc;
+       sys_info.mem_size = mem_end;
+
+       return 0;
+}
+
+static int __init check_sdias(void)
+{
+       if (!sclp_get_hsa_size()) {
+               TRACE("Could not determine HSA size\n");
+               return -ENODEV;
+       }
+       return 0;
+}
+
+static int __init get_mem_info(unsigned long *mem, unsigned long *end)
+{
+       struct memblock_region *reg;
+
+       for_each_memblock(memory, reg) {
+               *mem += reg->size;
+               *end = max_t(unsigned long, *end, reg->base + reg->size);
+       }
+       return 0;
+}
+
+static void __init zcore_header_init(int arch, struct zcore_header *hdr,
+                                    unsigned long mem_size)
+{
+       u32 prefix;
+       int i;
+
+       if (arch == ARCH_S390X)
+               hdr->arch_id = DUMP_ARCH_S390X;
+       else
+               hdr->arch_id = DUMP_ARCH_S390;
+       hdr->mem_size = mem_size;
+       hdr->rmem_size = mem_size;
+       hdr->mem_end = sys_info.mem_size;
+       hdr->num_pages = mem_size / PAGE_SIZE;
+       hdr->tod = get_tod_clock();
+       get_cpu_id(&hdr->cpu_id);
+       for (i = 0; i < dump_save_areas.count; i++) {
+               prefix = dump_save_areas.areas[i]->sa.pref_reg;
+               hdr->real_cpu_cnt++;
+               if (!prefix)
+                       continue;
+               hdr->lc_vec[hdr->cpu_cnt] = prefix;
+               hdr->cpu_cnt++;
+       }
+}
+
+/*
+ * Provide IPL parameter information block from either HSA or memory
+ * for future reipl
+ */
+static int __init zcore_reipl_init(void)
+{
+       struct ipib_info ipib_info;
+       int rc;
+
+       rc = memcpy_hsa_kernel(&ipib_info, __LC_DUMP_REIPL, sizeof(ipib_info));
+       if (rc)
+               return rc;
+       if (ipib_info.ipib == 0)
+               return 0;
+       ipl_block = (void *) __get_free_page(GFP_KERNEL);
+       if (!ipl_block)
+               return -ENOMEM;
+       if (ipib_info.ipib < sclp_get_hsa_size())
+               rc = memcpy_hsa_kernel(ipl_block, ipib_info.ipib, PAGE_SIZE);
+       else
+               rc = memcpy_real(ipl_block, (void *) ipib_info.ipib, PAGE_SIZE);
+       if (rc || csum_partial(ipl_block, ipl_block->hdr.len, 0) !=
+           ipib_info.checksum) {
+               TRACE("Checksum does not match\n");
+               free_page((unsigned long) ipl_block);
+               ipl_block = NULL;
+       }
+       return 0;
+}
+
+static int __init zcore_init(void)
+{
+       unsigned long mem_size, mem_end;
+       unsigned char arch;
+       int rc;
+
+       mem_size = mem_end = 0;
+       if (ipl_info.type != IPL_TYPE_FCP_DUMP)
+               return -ENODATA;
+       if (OLDMEM_BASE)
+               return -ENODATA;
+
+       zcore_dbf = debug_register("zcore", 4, 1, 4 * sizeof(long));
+       debug_register_view(zcore_dbf, &debug_sprintf_view);
+       debug_set_level(zcore_dbf, 6);
+
+       TRACE("devno:  %x\n", ipl_info.data.fcp.dev_id.devno);
+       TRACE("wwpn:   %llx\n", (unsigned long long) ipl_info.data.fcp.wwpn);
+       TRACE("lun:    %llx\n", (unsigned long long) ipl_info.data.fcp.lun);
+
+       rc = sclp_sdias_init();
+       if (rc)
+               goto fail;
+
+       rc = check_sdias();
+       if (rc)
+               goto fail;
+       hsa_available = 1;
+
+       rc = memcpy_hsa_kernel(&arch, __LC_AR_MODE_ID, 1);
+       if (rc)
+               goto fail;
+
+       if (arch == ARCH_S390) {
+               pr_alert("The 64-bit dump tool cannot be used for a "
+                        "32-bit system\n");
+               rc = -EINVAL;
+               goto fail;
+       }
+
+       rc = get_mem_info(&mem_size, &mem_end);
+       if (rc)
+               goto fail;
+
+       rc = sys_info_init(arch, mem_end);
+       if (rc)
+               goto fail;
+       zcore_header_init(arch, &zcore_header, mem_size);
+
+       rc = zcore_reipl_init();
+       if (rc)
+               goto fail;
+
+       zcore_dir = debugfs_create_dir("zcore" , NULL);
+       if (!zcore_dir) {
+               rc = -ENOMEM;
+               goto fail;
+       }
+       zcore_file = debugfs_create_file("mem", S_IRUSR, zcore_dir, NULL,
+                                        &zcore_fops);
+       if (!zcore_file) {
+               rc = -ENOMEM;
+               goto fail_dir;
+       }
+       zcore_memmap_file = debugfs_create_file("memmap", S_IRUSR, zcore_dir,
+                                               NULL, &zcore_memmap_fops);
+       if (!zcore_memmap_file) {
+               rc = -ENOMEM;
+               goto fail_file;
+       }
+       zcore_reipl_file = debugfs_create_file("reipl", S_IRUSR, zcore_dir,
+                                               NULL, &zcore_reipl_fops);
+       if (!zcore_reipl_file) {
+               rc = -ENOMEM;
+               goto fail_memmap_file;
+       }
+       zcore_hsa_file = debugfs_create_file("hsa", S_IRUSR|S_IWUSR, zcore_dir,
+                                            NULL, &zcore_hsa_fops);
+       if (!zcore_hsa_file) {
+               rc = -ENOMEM;
+               goto fail_reipl_file;
+       }
+       return 0;
+
+fail_reipl_file:
+       debugfs_remove(zcore_reipl_file);
+fail_memmap_file:
+       debugfs_remove(zcore_memmap_file);
+fail_file:
+       debugfs_remove(zcore_file);
+fail_dir:
+       debugfs_remove(zcore_dir);
+fail:
+       diag308(DIAG308_REL_HSA, NULL);
+       return rc;
+}
+
+static void __exit zcore_exit(void)
+{
+       debug_unregister(zcore_dbf);
+       sclp_sdias_exit();
+       free_page((unsigned long) ipl_block);
+       debugfs_remove(zcore_hsa_file);
+       debugfs_remove(zcore_reipl_file);
+       debugfs_remove(zcore_memmap_file);
+       debugfs_remove(zcore_file);
+       debugfs_remove(zcore_dir);
+       diag308(DIAG308_REL_HSA, NULL);
+}
+
+MODULE_AUTHOR("Copyright IBM Corp. 2003,2008");
+MODULE_DESCRIPTION("zcore module for zfcpdump support");
+MODULE_LICENSE("GPL");
+
+subsys_initcall(zcore_init);
+module_exit(zcore_exit);