2.2.x on Alpha? Who wants to test this?

2.2.x on Alpha? Who wants to test this?

Post by Jan-Benedict Gla » Wed, 09 Oct 2002 20:30:08



Hi!

Are *you* currently using a 2.2.x kernel on a Alpha with SRM firmware?
If so, I'd kindly like to ask you to test a backport of my srm_env.o
(used to access SRM environment variables). If you give it a try, please
report success, no success and ksymoopsed Oopses to me:-)

Here's the patch. It still has some warnings about not-used functions
(init/exit function, #define'd here to init_module and cleanup_module),
but seems I don't see the point... Could someone please enlighten me on
this?

MfG, JBG

diff -Nur linux-2.2.22-pure/Documentation/Configure.help linux-2.2.22-srm/Documentation/Configure.help
--- linux-2.2.22-pure/Documentation/Configure.help      2002-09-16 18:26:11.000000000 +0200
+++ linux-2.2.22-srm/Documentation/Configure.help       2002-10-08 19:59:19.000000000 +0200
@@ -15138,6 +15138,27 @@
    boards supported by this driver, and for further information
    on the use of this driver.

+SRM environment access through procfs
+CONFIG_SRM_ENV
+  If you enable this option, a subdirectory inside /proc called
+  /proc/srm_environment will give you access to the all important
+  SRM environment variables (those which have a name) and also
+  to all others (by their internal number).
+
+  SRM is something like a BIOS for Alpha machines. There are some
+  other such BIOSes, like AlphaBIOS, which this driver cannot
+  support (hey, that's not SRM!).
+
+  Despite the fact that this driver doesn't work on all Alphas (but
+  only on those which have SRM as their firmware), it's save to
+  build it even if your particular machine doesn't know about SRM
+  (or if you intend to compile a generic kernel). It will simply
+  not create those subdirectory in /proc (and give you some warning,
+  of course).
+
+  This driver is also available as a module and will be called
+  srm_env.o then.
+
 Compaq Smart Array support
 CONFIG_BLK_CPQ_CISS_DA
    This is the driver for Compaq Smart Array 5xxx controllers.
diff -Nur linux-2.2.22-pure/arch/alpha/config.in linux-2.2.22-srm/arch/alpha/config.in
--- linux-2.2.22-pure/arch/alpha/config.in      2001-03-25 18:37:29.000000000 +0200
+++ linux-2.2.22-srm/arch/alpha/config.in       2002-10-08 20:05:54.000000000 +0200
@@ -165,7 +165,7 @@
         -o "$CONFIG_ALPHA_TAKARA" = "y" -o "$CONFIG_ALPHA_EB164" = "y" \
         -o "$CONFIG_ALPHA_ALCOR" = "y"  -o "$CONFIG_ALPHA_MIATA" = "y" \
         -o "$CONFIG_ALPHA_LX164" = "y"  -o "$CONFIG_ALPHA_SX164" = "y" \
-        -o "$CONFIG_ALPHA_NAUTILUS" = "y" ]
+        -o "$CONFIG_ALPHA_NAUTILUS" = "y" -o "$CONFIG_ALPHA_NONAME" = "y" ]
 then
   bool 'Use SRM as bootloader' CONFIG_ALPHA_SRM
 #  if [ "$CONFIG_EXPERIMENTAL" = "y" ]; then
@@ -209,6 +209,9 @@
   tristate 'Kernel support for JAVA binaries (obsolete)' CONFIG_BINFMT_JAVA
 fi
 tristate 'Kernel support for Linux/Intel ELF binaries' CONFIG_BINFMT_EM86
+if [ "$CONFIG_ALPHA_SRM" = "y" ]; then
+  tristate 'SRM environment access through procfs' CONFIG_ALPHA_SRM_ENV
+fi
 tristate 'Parallel port support' CONFIG_PARPORT
 if [ "$CONFIG_PARPORT" != "n" ]; then
   dep_tristate '  PC-style hardware' CONFIG_PARPORT_PC $CONFIG_PARPORT
diff -Nur linux-2.2.22-pure/arch/alpha/kernel/Makefile linux-2.2.22-srm/arch/alpha/kernel/Makefile
--- linux-2.2.22-pure/arch/alpha/kernel/Makefile        2001-03-25 18:31:46.000000000 +0200
+++ linux-2.2.22-srm/arch/alpha/kernel/Makefile 2002-10-08 20:05:00.000000000 +0200
@@ -17,6 +17,9 @@
            bios32.o ptrace.o time.o fpreg.o
 OX_OBJS  := alpha_ksyms.o

+ifdef CONFIG_ALPHA_SRM_ENV
+M_OBJS   := srm_env.o
+endif

 ifdef CONFIG_ALPHA_GENERIC

diff -Nur linux-2.2.22-pure/arch/alpha/kernel/srm_env.c linux-2.2.22-srm/arch/alpha/kernel/srm_env.c
--- linux-2.2.22-pure/arch/alpha/kernel/srm_env.c       1970-01-01 01:00:00.000000000 +0100
+++ linux-2.2.22-srm/arch/alpha/kernel/srm_env.c        2002-10-08 20:11:56.000000000 +0200
@@ -0,0 +1,384 @@
+/*
+ * srm_env.c - Access to SRM environment
+ *             variables through linux' procfs
+ *
+ * Copyright (C) 2001-2002 Jan-Benedict Glaw <jbg...@lug-owl.de>
+ *
+ * This driver is at all a modified version of Erik Mouw's
+ * ./linux/Documentation/DocBook/procfs_example.c, so: thank
+ * you, Erik! He can be reached via email at
+ * <J.A.K.M...@its.tudelft.nl>. It is based on an idea
+ * provided by DEC^WCompaq^WIntel's "Jumpstart" CD. They
+ * included a patch like this as well. Thanks for idea!
+ *
+ * This program is free software; you can redistribute
+ * it and/or modify it under the terms of the GNU General
+ * Public License version 2 as published by the Free Software
+ * Foundation.
+ *
+ * This program is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied
+ * warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * PURPOSE.  See the GNU General Public License for more
+ * details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place,
+ * Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+
+/*
+ * Changelog
+ * ~~~~~~~~~
+ *
+ * Tue, 08 Oct 2002 20:10:21 +0200
+ *     - Port srm_env to 2.2.x.
+ *     - This is v0.0.5-2.2.x-1 (1st try...)
+ *
+ * Thu, 22 Aug 2002 15:10:43 +0200
+ *     - Update Config.help entry. I got a number of emails asking
+ *       me to tell their senders if they could make use of this
+ *       piece of code... So: "SRM is something like BIOS for your
+ *       Alpha"
+ *     - Update code formatting a bit to better conform CodingStyle
+ *       rules.
+ *     - So this is v0.0.5, with no changes (except formatting)
+ *    
+ * Wed, 22 May 2002 00:11:21 +0200
+ *     - Fix typo on comment (SRC -> SRM)
+ *     - Call this "Version 0.0.4"
+ *
+ * Tue,  9 Apr 2002 18:44:40 +0200
+ *     - Implement access by variable name and additionally
+ *       by number. This is done by creating two subdirectories
+ *       where one holds all names (like the old directory
+ *       did) and the other holding 256 files named like "0",
+ *       "1" and so on.
+ *     - Call this "Version 0.0.3"
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/proc_fs.h>
+#include <linux/version.h>
+#include <asm/console.h>
+#include <asm/uaccess.h>
+#include <asm/machvec.h>
+
+/*
+ * 2.2.x compatibility cruft
+ */
+#if !defined(__init)
+#      define __init
+#endif
+#if !defined(__edit)
+#      define __exit
+#endif
+
+
+
+#define BASE_DIR       "srm_environment"     /* Subdir in /proc/             */
+#define NAMED_DIR      "named_variables"     /* Subdir for known variables   */
+#define NUMBERED_DIR   "numbered_variables"  /* Subdir for all variables     */
+#define VERSION                "0.0.5-2.2.x-1"               /* Module version               */
+#define NAME           "srm_env"             /* Module name                  */
+
+MODULE_AUTHOR("Jan-Benedict Glaw <jbg...@lug-owl.de>");
+MODULE_DESCRIPTION("Accessing Alpha SRM environment through procfs interface");
+MODULE_LICENSE("GPL");
+
+typedef struct _srm_env {
+       char                    *name;
+       unsigned long           id;
+       struct proc_dir_entry   *proc_entry;
+} srm_env_t;
+
+static struct proc_dir_entry   *base_dir;
+static struct proc_dir_entry   *named_dir;
+static struct proc_dir_entry   *numbered_dir;
+static char                    number[256][4];
+
+static srm_env_t       srm_named_entries[] = {
+       { "auto_action",      ENV_AUTO_ACTION         },
+       { "boot_dev",         ENV_BOOT_DEV            },
+       { "bootdef_dev",      ENV_BOOTDEF_DEV         },
+       { "booted_dev",               ENV_BOOTED_DEV          },
+       { "boot_file",                ENV_BOOT_FILE           },
+       { "booted_file",      ENV_BOOTED_FILE         },
+       { "boot_osflags",     ENV_BOOT_OSFLAGS        },
+       { "booted_osflags",   ENV_BOOTED_OSFLAGS      },
+       { "boot_reset",               ENV_BOOT_RESET          },
+       { "dump_dev",         ENV_DUMP_DEV            },
+       { "enable_audit",     ENV_ENABLE_AUDIT        },
+       { "license",          ENV_LICENSE             },
+       { "char_set",         ENV_CHAR_SET            },
+       { "language",         ENV_LANGUAGE            },
+       { "tty_dev",          ENV_TTY_DEV             },
+       { NULL,                 0                       },
+};
+static srm_env_t       srm_numbered_entries[256];
+
+
+
+void
+srm_env_set_owner(struct proc_dir_entry *entry)
+{
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 4, 0)
+       /*
+        * /proc/ entries are "owned" since 2.4.0 or so...
+        */
+       entry->owner = THIS_MODULE;
+#endif /* LINUX_VERSION_CODE >= 2.4.0 */
+
+         return;
+}
+
+
+static int
+srm_env_read(char *page, char **start, off_t off, int count, int *eof,
+               void *data)
+{
+       int             nbytes;
+       unsigned long   ret;
+       srm_env_t       *entry;
+
+       MOD_INC_USE_COUNT;
+
+       if(off != 0) {
+               MOD_DEC_USE_COUNT;
+               return -EFAULT;
+       }
+
+       entry   = (srm_env_t *) data;
+       ret     = callback_getenv(entry->id, page, count);
+
+       if((ret >> 61) == 0)
+               nbytes = (int) ret;
+       else
+               nbytes = -EFAULT;
+
+       MOD_DEC_USE_COUNT;
+
+       return nbytes;
+}
+
+
+static int
+srm_env_write(struct file *file, const char *buffer, unsigned long count,
+               void *data)
+{
+#define BUFLEN 512
+       int             nbytes;
+       srm_env_t       *entry;
+       char            buf[BUFLEN];
+       unsigned long   ret1, ret2;
+
+       MOD_INC_USE_COUNT;
+
+       entry = (srm_env_t *) data;
+
+       nbytes = strlen(buffer) + 1;
+       if(nbytes > BUFLEN) {
+               MOD_DEC_USE_COUNT;
+               return -ENOMEM;
+       }
+
+       /* memcpy(aligned_buffer, buffer, nbytes) */
+
+       if(copy_from_user(buf, buffer, count)) {
+               MOD_DEC_USE_COUNT;
+               return -EFAULT;
+       }
+       buf[count] = 0x00;
+
+       ret1 = callback_setenv(entry->id, buf, count);
+       if((ret1 >> 61) == 0) {
+               do
+                       ret2 = callback_save_env();
+               while((ret2 >> 61) == 1);
+               nbytes = (int) ret1;
+       } else
+               nbytes = -EFAULT;
+
+       MOD_DEC_USE_COUNT;
+
+       return nbytes;
+}
+
+
+static void
+srm_env_cleanup(void)
+{
+       srm_env_t       *entry;
+       unsigned long   var_num;
+
+       if(base_dir) {
+               /*
+                * Remove named entries
+                */
+               if(named_dir) {
+                       entry = srm_named_entries;
+                       while(entry->name != NULL && entry->id != 0) {
+                               if(entry->proc_entry) {
+                                       remove_proc_entry(entry->name,
+                                                       named_dir);
+                                       entry->proc_entry = NULL;
+                               }
+                               entry++;
+                       }
+                       remove_proc_entry(NAMED_DIR, base_dir);
+               }
+
+               /*
+                * Remove numbered entries
+                */
+               if(numbered_dir) {
+                       for(var_num = 0; var_num <= 255; var_num++) {
+                               entry = &srm_numbered_entries[var_num];
+
+                               if(entry->proc_entry) {
+                                       remove_proc_entry(entry->name,
+                                                       numbered_dir);
+              
...

read more »

  application_pgp-signature_part
< 1K Download
 
 
 

2.2.x on Alpha? Who wants to test this?

Post by Krzysiek Taraszk » Sun, 15 Dec 2002 19:10:07



> Hi!

> Are *you* currently using a 2.2.x kernel on a Alpha with SRM firmware?
> If so, I'd kindly like to ask you to test a backport of my srm_env.o
> (used to access SRM environment variables). If you give it a try, please
> report success, no success and ksymoopsed Oopses to me:-)

did you test it ? or this is theoreticaly patch ?


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/