X-Git-Url: https://git.proxmox.com/?a=blobdiff_plain;f=module%2Fspl%2Fspl-generic.c;h=f6782dae73af08eec0b54c2da385ece22c405812;hb=ec06701b5473894514b25e4054dd5597e25a22e8;hp=1dedf76cf22cf78d6e84f540dd6f129e51b35984;hpb=6b3b569df30d13ed7bbbff877cffc71290a52f12;p=mirror_spl-debian.git diff --git a/module/spl/spl-generic.c b/module/spl/spl-generic.c index 1dedf76..f6782da 100644 --- a/module/spl/spl-generic.c +++ b/module/spl/spl-generic.c @@ -6,7 +6,7 @@ * UCRL-CODE-235197 * * This file is part of the SPL, Solaris Porting Layer. - * For details, see . + * For details, see . * * The SPL is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License as published by the @@ -29,6 +29,8 @@ #include #include #include +#include +#include #include #include #include @@ -37,67 +39,130 @@ #include #include #include -#include #include +#include +#include +#include #include +#include #include -#include -#ifdef SS_DEBUG_SUBSYS -#undef SS_DEBUG_SUBSYS -#endif - -#define SS_DEBUG_SUBSYS SS_GENERIC - -char spl_version[16] = "SPL v" SPL_META_VERSION; +char spl_version[32] = "SPL v" SPL_META_VERSION "-" SPL_META_RELEASE; EXPORT_SYMBOL(spl_version); -unsigned long spl_hostid = HW_INVALID_HOSTID; +unsigned long spl_hostid = 0; EXPORT_SYMBOL(spl_hostid); module_param(spl_hostid, ulong, 0644); MODULE_PARM_DESC(spl_hostid, "The system hostid."); -char hw_serial[HW_HOSTID_LEN] = ""; -EXPORT_SYMBOL(hw_serial); - -proc_t p0 = { 0 }; +proc_t p0; EXPORT_SYMBOL(p0); -#ifndef HAVE_KALLSYMS_LOOKUP_NAME -kallsyms_lookup_name_t spl_kallsyms_lookup_name_fn = SYMBOL_POISON; -#endif +/* + * Xorshift Pseudo Random Number Generator based on work by Sebastiano Vigna + * + * "Further scramblings of Marsaglia's xorshift generators" + * http://vigna.di.unimi.it/ftp/papers/xorshiftplus.pdf + * + * random_get_pseudo_bytes() is an API function on Illumos whose sole purpose + * is to provide bytes containing random numbers. It is mapped to /dev/urandom + * on Illumos, which uses a "FIPS 186-2 algorithm". No user of the SPL's + * random_get_pseudo_bytes() needs bytes that are of cryptographic quality, so + * we can implement it using a fast PRNG that we seed using Linux' actual + * equivalent to random_get_pseudo_bytes(). We do this by providing each CPU + * with an independent seed so that all calls to random_get_pseudo_bytes() are + * free of atomic instructions. + * + * A consequence of using a fast PRNG is that using random_get_pseudo_bytes() + * to generate words larger than 128 bits will paradoxically be limited to + * `2^128 - 1` possibilities. This is because we have a sequence of `2^128 - 1` + * 128-bit words and selecting the first will implicitly select the second. If + * a caller finds this behavior undesireable, random_get_bytes() should be used + * instead. + * + * XXX: Linux interrupt handlers that trigger within the critical section + * formed by `s[1] = xp[1];` and `xp[0] = s[0];` and call this function will + * see the same numbers. Nothing in the code currently calls this in an + * interrupt handler, so this is considered to be okay. If that becomes a + * problem, we could create a set of per-cpu variables for interrupt handlers + * and use them when in_interrupt() from linux/preempt_mask.h evaluates to + * true. + */ +static DEFINE_PER_CPU(uint64_t[2], spl_pseudo_entropy); + +/* + * spl_rand_next()/spl_rand_jump() are copied from the following CC-0 licensed + * file: + * + * http://xorshift.di.unimi.it/xorshift128plus.c + */ + +static inline uint64_t +spl_rand_next(uint64_t *s) { + uint64_t s1 = s[0]; + const uint64_t s0 = s[1]; + s[0] = s0; + s1 ^= s1 << 23; // a + s[1] = s1 ^ s0 ^ (s1 >> 18) ^ (s0 >> 5); // b, c + return (s[1] + s0); +} + +static inline void +spl_rand_jump(uint64_t *s) { + static const uint64_t JUMP[] = { 0x8a5cd789635d2dff, 0x121fd2155c472f96 }; + + uint64_t s0 = 0; + uint64_t s1 = 0; + int i, b; + for(i = 0; i < sizeof JUMP / sizeof *JUMP; i++) + for(b = 0; b < 64; b++) { + if (JUMP[i] & 1ULL << b) { + s0 ^= s[0]; + s1 ^= s[1]; + } + (void) spl_rand_next(s); + } + + s[0] = s0; + s[1] = s1; +} int -highbit(unsigned long i) +random_get_pseudo_bytes(uint8_t *ptr, size_t len) { - register int h = 1; - SENTRY; - - if (i == 0) - SRETURN(0); -#if BITS_PER_LONG == 64 - if (i & 0xffffffff00000000ul) { - h += 32; i >>= 32; - } -#endif - if (i & 0xffff0000) { - h += 16; i >>= 16; - } - if (i & 0xff00) { - h += 8; i >>= 8; - } - if (i & 0xf0) { - h += 4; i >>= 4; - } - if (i & 0xc) { - h += 2; i >>= 2; - } - if (i & 0x2) { - h += 1; - } - SRETURN(h); + uint64_t *xp, s[2]; + + ASSERT(ptr); + + xp = get_cpu_var(spl_pseudo_entropy); + + s[0] = xp[0]; + s[1] = xp[1]; + + while (len) { + union { + uint64_t ui64; + uint8_t byte[sizeof (uint64_t)]; + }entropy; + int i = MIN(len, sizeof (uint64_t)); + + len -= i; + entropy.ui64 = spl_rand_next(s); + + while (i--) + *ptr++ = entropy.byte[i]; + } + + xp[0] = s[0]; + xp[1] = s[1]; + + put_cpu_var(spl_pseudo_entropy); + + return (0); } -EXPORT_SYMBOL(highbit); + + +EXPORT_SYMBOL(random_get_pseudo_bytes); #if BITS_PER_LONG == 32 /* @@ -183,7 +248,7 @@ __udivdi3(uint64_t u, uint64_t v) q0 = q0 - 1; // too small by 1. if ((u - q0 * v) >= v) q0 = q0 + 1; // Now q0 is correct. - + return q0; } } @@ -212,6 +277,105 @@ __umoddi3(uint64_t dividend, uint64_t divisor) } EXPORT_SYMBOL(__umoddi3); +/* + * Implementation of 64-bit unsigned division/modulo for 32-bit machines. + */ +uint64_t +__udivmoddi4(uint64_t n, uint64_t d, uint64_t *r) +{ + uint64_t q = __udivdi3(n, d); + if (r) + *r = n - d * q; + return (q); +} +EXPORT_SYMBOL(__udivmoddi4); + +/* + * Implementation of 64-bit signed division/modulo for 32-bit machines. + */ +int64_t +__divmoddi4(int64_t n, int64_t d, int64_t *r) +{ + int64_t q, rr; + boolean_t nn = B_FALSE; + boolean_t nd = B_FALSE; + if (n < 0) { + nn = B_TRUE; + n = -n; + } + if (d < 0) { + nd = B_TRUE; + d = -d; + } + + q = __udivmoddi4(n, d, (uint64_t *)&rr); + + if (nn != nd) + q = -q; + if (nn) + rr = -rr; + if (r) + *r = rr; + return (q); +} +EXPORT_SYMBOL(__divmoddi4); + +#if defined(__arm) || defined(__arm__) +/* + * Implementation of 64-bit (un)signed division for 32-bit arm machines. + * + * Run-time ABI for the ARM Architecture (page 20). A pair of (unsigned) + * long longs is returned in {{r0, r1}, {r2,r3}}, the quotient in {r0, r1}, + * and the remainder in {r2, r3}. The return type is specifically left + * set to 'void' to ensure the compiler does not overwrite these registers + * during the return. All results are in registers as per ABI + */ +void +__aeabi_uldivmod(uint64_t u, uint64_t v) +{ + uint64_t res; + uint64_t mod; + + res = __udivdi3(u, v); + mod = __umoddi3(u, v); + { + register uint32_t r0 asm("r0") = (res & 0xFFFFFFFF); + register uint32_t r1 asm("r1") = (res >> 32); + register uint32_t r2 asm("r2") = (mod & 0xFFFFFFFF); + register uint32_t r3 asm("r3") = (mod >> 32); + + asm volatile("" + : "+r"(r0), "+r"(r1), "+r"(r2),"+r"(r3) /* output */ + : "r"(r0), "r"(r1), "r"(r2), "r"(r3)); /* input */ + + return; /* r0; */ + } +} +EXPORT_SYMBOL(__aeabi_uldivmod); + +void +__aeabi_ldivmod(int64_t u, int64_t v) +{ + int64_t res; + uint64_t mod; + + res = __divdi3(u, v); + mod = __umoddi3(u, v); + { + register uint32_t r0 asm("r0") = (res & 0xFFFFFFFF); + register uint32_t r1 asm("r1") = (res >> 32); + register uint32_t r2 asm("r2") = (mod & 0xFFFFFFFF); + register uint32_t r3 asm("r3") = (mod >> 32); + + asm volatile("" + : "+r"(r0), "+r"(r1), "+r"(r2),"+r"(r3) /* output */ + : "r"(r0), "r"(r1), "r"(r2), "r"(r3)); /* input */ + + return; /* r0; */ + } +} +EXPORT_SYMBOL(__aeabi_ldivmod); +#endif /* __arm || __arm__ */ #endif /* BITS_PER_LONG */ /* NOTE: The strtoxx behavior is solely based on my reading of the Solaris @@ -338,33 +502,6 @@ ddi_copyout(const void *from, void *to, size_t len, int flags) } EXPORT_SYMBOL(ddi_copyout); -#ifndef HAVE_PUT_TASK_STRUCT -/* - * This is only a stub function which should never be used. The SPL should - * never be putting away the last reference on a task structure so this will - * not be called. However, we still need to define it so the module does not - * have undefined symbol at load time. That all said if this impossible - * thing does somehow happen PANIC immediately so we know about it. - */ -void -__put_task_struct(struct task_struct *t) -{ - PANIC("Unexpectly put last reference on task %d\n", (int)t->pid); -} -EXPORT_SYMBOL(__put_task_struct); -#endif /* HAVE_PUT_TASK_STRUCT */ - -struct new_utsname *__utsname(void) -{ -#ifdef HAVE_INIT_UTSNAME - return init_utsname(); -#else - return &system_utsname; -#endif -} -EXPORT_SYMBOL(__utsname); - - /* * Read the unique system identifier from the /etc/hostid file. * @@ -405,270 +542,218 @@ module_param(spl_hostid_path, charp, 0444); MODULE_PARM_DESC(spl_hostid_path, "The system hostid file (/etc/hostid)"); static int -hostid_read(void) +hostid_read(uint32_t *hostid) { - int result; uint64_t size; struct _buf *file; - unsigned long hostid = 0; + uint32_t value = 0; + int error; file = kobj_open_file(spl_hostid_path); - if (file == (struct _buf *)-1) - return -1; - - result = kobj_get_filesize(file, &size); + return (ENOENT); - if (result != 0) { - printk(KERN_WARNING - "SPL: kobj_get_filesize returned %i on %s\n", - result, spl_hostid_path); + error = kobj_get_filesize(file, &size); + if (error) { kobj_close_file(file); - return -2; + return (error); } if (size < sizeof(HW_HOSTID_MASK)) { - printk(KERN_WARNING - "SPL: Ignoring the %s file because it is %llu bytes; " - "expecting %lu bytes instead.\n", - spl_hostid_path, size, sizeof(HW_HOSTID_MASK)); kobj_close_file(file); - return -3; + return (EINVAL); } - /* Read directly into the variable like eglibc does. */ - /* Short reads are okay; native behavior is preserved. */ - result = kobj_read_file(file, (char *)&hostid, sizeof(hostid), 0); - - if (result < 0) { - printk(KERN_WARNING - "SPL: kobj_read_file returned %i on %s\n", - result, spl_hostid_path); + /* + * Read directly into the variable like eglibc does. + * Short reads are okay; native behavior is preserved. + */ + error = kobj_read_file(file, (char *)&value, sizeof(value), 0); + if (error < 0) { kobj_close_file(file); - return -4; + return (EIO); } /* Mask down to 32 bits like coreutils does. */ - spl_hostid = hostid & HW_HOSTID_MASK; + *hostid = (value & HW_HOSTID_MASK); kobj_close_file(file); + return 0; } -#define GET_HOSTID_CMD \ - "exec 0/proc/sys/kernel/spl/hostid " \ - " 2>/dev/null; " \ - "hostid" - -static int -hostid_exec(void) +/* + * Return the system hostid. Preferentially use the spl_hostid module option + * when set, otherwise use the value in the /etc/hostid file. + */ +uint32_t +zone_get_hostid(void *zone) { - char *argv[] = { "/bin/sh", - "-c", - GET_HOSTID_CMD, - NULL }; - char *envp[] = { "HOME=/", - "TERM=linux", - "PATH=/sbin:/usr/sbin:/bin:/usr/bin", - NULL }; - int rc; - - /* Doing address resolution in the kernel is tricky and just - * not a good idea in general. So to set the proper 'hw_serial' - * use the usermodehelper support to ask '/bin/sh' to run - * '/usr/bin/hostid' and redirect the result to /proc/sys/spl/hostid - * for us to use. It's a horrific solution but it will do for now. - */ - rc = call_usermodehelper(argv[0], argv, envp, 1); - if (rc) - printk("SPL: Failed user helper '%s %s %s', rc = %d\n", - argv[0], argv[1], argv[2], rc); + uint32_t hostid; + + ASSERT3P(zone, ==, NULL); + + if (spl_hostid != 0) + return ((uint32_t)(spl_hostid & HW_HOSTID_MASK)); - return rc; + if (hostid_read(&hostid) == 0) + return (hostid); + + return (0); } +EXPORT_SYMBOL(zone_get_hostid); -uint32_t -zone_get_hostid(void *zone) +static int +spl_kvmem_init(void) { - unsigned long hostid; + int rc = 0; - /* Only the global zone is supported */ - ASSERT(zone == NULL); + rc = spl_kmem_init(); + if (rc) + return (rc); - if (ddi_strtoul(hw_serial, NULL, HW_HOSTID_LEN-1, &hostid) != 0) - return HW_INVALID_HOSTID; + rc = spl_vmem_init(); + if (rc) { + spl_kmem_fini(); + return (rc); + } - return (uint32_t)hostid; + return (rc); } -EXPORT_SYMBOL(zone_get_hostid); -#ifndef HAVE_KALLSYMS_LOOKUP_NAME /* - * The kallsyms_lookup_name() kernel function is not an exported symbol in - * Linux 2.6.19 through 2.6.32 inclusive. - * - * This function replaces the functionality by performing an upcall to user - * space where /proc/kallsyms is consulted for the requested address. - * + * We initialize the random number generator with 128 bits of entropy from the + * system random number generator. In the improbable case that we have a zero + * seed, we fallback to the system jiffies, unless it is also zero, in which + * situation we use a preprogrammed seed. We step forward by 2^64 iterations to + * initialize each of the per-cpu seeds so that the sequences generated on each + * CPU are guaranteed to never overlap in practice. */ +static void __init +spl_random_init(void) +{ + uint64_t s[2]; + int i; -#define GET_KALLSYMS_ADDR_CMD \ - "exec 0/proc/sys/kernel/spl/kallsyms_lookup_name " \ - " 2>/dev/null; " \ - "awk '{ if ( $3 == \"kallsyms_lookup_name\" ) { print $1 } }' " \ - " /proc/kallsyms " + get_random_bytes(s, sizeof (s)); -static int -set_kallsyms_lookup_name(void) -{ - char *argv[] = { "/bin/sh", - "-c", - GET_KALLSYMS_ADDR_CMD, - NULL }; - char *envp[] = { "HOME=/", - "TERM=linux", - "PATH=/sbin:/usr/sbin:/bin:/usr/bin", - NULL }; - int rc; - - rc = call_usermodehelper(argv[0], argv, envp, 1); - if (rc) - printk("SPL: Failed user helper '%s %s %s', rc = %d\n", - argv[0], argv[1], argv[2], rc); + if (s[0] == 0 && s[1] == 0) { + if (jiffies != 0) { + s[0] = jiffies; + s[1] = ~0 - jiffies; + } else { + (void) memcpy(s, "improbable seed", sizeof (s)); + } + printk("SPL: get_random_bytes() returned 0 " + "when generating random seed. Setting initial seed to " + "0x%016llx%016llx.", cpu_to_be64(s[0]), cpu_to_be64(s[1])); + } - return rc; + for_each_possible_cpu(i) { + uint64_t *wordp = per_cpu(spl_pseudo_entropy, i); + + spl_rand_jump(s); + + wordp[0] = s[0]; + wordp[1] = s[1]; + } } -#endif -static int -__init spl_init(void) +static void +spl_kvmem_fini(void) +{ + spl_vmem_fini(); + spl_kmem_fini(); +} + +static int __init +spl_init(void) { int rc = 0; - if ((rc = debug_init())) - return rc; + bzero(&p0, sizeof (proc_t)); + spl_random_init(); - if ((rc = spl_kmem_init())) - SGOTO(out1, rc); + if ((rc = spl_kvmem_init())) + goto out1; if ((rc = spl_mutex_init())) - SGOTO(out2, rc); + goto out2; if ((rc = spl_rw_init())) - SGOTO(out3, rc); - - if ((rc = spl_taskq_init())) - SGOTO(out4, rc); - - if ((rc = vn_init())) - SGOTO(out5, rc); + goto out3; - if ((rc = proc_init())) - SGOTO(out6, rc); + if ((rc = spl_tsd_init())) + goto out4; - if ((rc = kstat_init())) - SGOTO(out7, rc); + if ((rc = spl_taskq_init())) + goto out5; - if ((rc = tsd_init())) - SGOTO(out8, rc); + if ((rc = spl_kmem_cache_init())) + goto out6; - if ((rc = zlib_init())) - SGOTO(out9, rc); + if ((rc = spl_vn_init())) + goto out7; - /* - * Get the hostid if it was not passed as a module parameter. Try - * reading the /etc/hostid file directly, and then fall back to calling - * the /usr/bin/hostid utility. - */ + if ((rc = spl_proc_init())) + goto out8; - if (spl_hostid == HW_INVALID_HOSTID - && (rc = hostid_read()) && (rc = hostid_exec())) - SGOTO(out10, rc = -EADDRNOTAVAIL); + if ((rc = spl_kstat_init())) + goto out9; -#ifndef HAVE_KALLSYMS_LOOKUP_NAME - if ((rc = set_kallsyms_lookup_name())) - SGOTO(out10, rc = -EADDRNOTAVAIL); -#endif /* HAVE_KALLSYMS_LOOKUP_NAME */ + if ((rc = spl_zlib_init())) + goto out10; - if ((rc = spl_kmem_init_kallsyms_lookup())) - SGOTO(out10, rc); + printk(KERN_NOTICE "SPL: Loaded module v%s-%s%s\n", SPL_META_VERSION, + SPL_META_RELEASE, SPL_DEBUG_STR); + return (rc); - printk(KERN_NOTICE "SPL: Loaded module v%s%s, using hostid 0x%08x\n", - SPL_META_VERSION, SPL_DEBUG_STR, (unsigned int) spl_hostid); - SRETURN(rc); out10: - zlib_fini(); + spl_kstat_fini(); out9: - tsd_fini(); + spl_proc_fini(); out8: - kstat_fini(); + spl_vn_fini(); out7: - proc_fini(); + spl_kmem_cache_fini(); out6: - vn_fini(); -out5: spl_taskq_fini(); +out5: + spl_tsd_fini(); out4: spl_rw_fini(); out3: spl_mutex_fini(); out2: - spl_kmem_fini(); + spl_kvmem_fini(); out1: - debug_fini(); + printk(KERN_NOTICE "SPL: Failed to Load Solaris Porting Layer " + "v%s-%s%s, rc = %d\n", SPL_META_VERSION, SPL_META_RELEASE, + SPL_DEBUG_STR, rc); - printk(KERN_NOTICE "SPL: Failed to Load Solaris Porting Layer v%s%s" - ", rc = %d\n", SPL_META_VERSION, SPL_DEBUG_STR, rc); - return rc; + return (rc); } -static void +static void __exit spl_fini(void) { - SENTRY; - - printk(KERN_NOTICE "SPL: Unloaded module v%s%s\n", - SPL_META_VERSION, SPL_DEBUG_STR); - zlib_fini(); - tsd_fini(); - kstat_fini(); - proc_fini(); - vn_fini(); + printk(KERN_NOTICE "SPL: Unloaded module v%s-%s%s\n", + SPL_META_VERSION, SPL_META_RELEASE, SPL_DEBUG_STR); + spl_zlib_fini(); + spl_kstat_fini(); + spl_proc_fini(); + spl_vn_fini(); + spl_kmem_cache_fini(); spl_taskq_fini(); + spl_tsd_fini(); spl_rw_fini(); spl_mutex_fini(); - spl_kmem_fini(); - debug_fini(); -} - -/* Called when a dependent module is loaded */ -void -spl_setup(void) -{ - int rc; - - /* - * At module load time the pwd is set to '/' on a Solaris system. - * On a Linux system will be set to whatever directory the caller - * was in when executing insmod/modprobe. - */ - rc = vn_set_pwd("/"); - if (rc) - printk("SPL: Warning unable to set pwd to '/': %d\n", rc); -} -EXPORT_SYMBOL(spl_setup); - -/* Called when a dependent module is unloaded */ -void -spl_cleanup(void) -{ + spl_kvmem_fini(); } -EXPORT_SYMBOL(spl_cleanup); module_init(spl_init); module_exit(spl_fini); -MODULE_AUTHOR("Lawrence Livermore National Labs"); MODULE_DESCRIPTION("Solaris Porting Layer"); -MODULE_LICENSE("GPL"); +MODULE_AUTHOR(SPL_META_AUTHOR); +MODULE_LICENSE(SPL_META_LICENSE); +MODULE_VERSION(SPL_META_VERSION "-" SPL_META_RELEASE);