diff options
Diffstat (limited to 'arch/s390')
| -rw-r--r-- | arch/s390/Kconfig | 2 | ||||
| -rw-r--r-- | arch/s390/crypto/crypt_s390.h | 122 | ||||
| -rw-r--r-- | arch/s390/crypto/prng.c | 850 | ||||
| -rw-r--r-- | arch/s390/hypfs/inode.c | 18 | ||||
| -rw-r--r-- | arch/s390/include/asm/dma-mapping.h | 2 | ||||
| -rw-r--r-- | arch/s390/include/asm/kexec.h | 3 | ||||
| -rw-r--r-- | arch/s390/include/asm/mmu.h | 4 | ||||
| -rw-r--r-- | arch/s390/include/asm/mmu_context.h | 3 | ||||
| -rw-r--r-- | arch/s390/include/asm/pci.h | 10 | ||||
| -rw-r--r-- | arch/s390/include/asm/pgalloc.h | 1 | ||||
| -rw-r--r-- | arch/s390/include/asm/pgtable.h | 169 | ||||
| -rw-r--r-- | arch/s390/kvm/kvm-s390.c | 2 | ||||
| -rw-r--r-- | arch/s390/mm/hugetlbpage.c | 66 | ||||
| -rw-r--r-- | arch/s390/mm/pgtable.c | 142 | ||||
| -rw-r--r-- | arch/s390/net/bpf_jit.S | 197 | ||||
| -rw-r--r-- | arch/s390/net/bpf_jit.h | 58 | ||||
| -rw-r--r-- | arch/s390/net/bpf_jit_comp.c | 1781 | ||||
| -rw-r--r-- | arch/s390/pci/pci.c | 6 | ||||
| -rw-r--r-- | arch/s390/pci/pci_debug.c | 33 | ||||
| -rw-r--r-- | arch/s390/pci/pci_dma.c | 8 |
20 files changed, 2316 insertions, 1161 deletions
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index de2726a487b0..b06dc3839268 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -115,7 +115,7 @@ config S390 select HAVE_ARCH_SECCOMP_FILTER select HAVE_ARCH_TRACEHOOK select HAVE_ARCH_TRANSPARENT_HUGEPAGE - select HAVE_BPF_JIT if PACK_STACK + select HAVE_BPF_JIT if PACK_STACK && HAVE_MARCH_Z196_FEATURES select HAVE_CMPXCHG_DOUBLE select HAVE_CMPXCHG_LOCAL select HAVE_DEBUG_KMEMLEAK diff --git a/arch/s390/crypto/crypt_s390.h b/arch/s390/crypto/crypt_s390.h index ba3b2aefddf5..d9c4c313fbc6 100644 --- a/arch/s390/crypto/crypt_s390.h +++ b/arch/s390/crypto/crypt_s390.h @@ -3,9 +3,10 @@ * * Support for s390 cryptographic instructions. * - * Copyright IBM Corp. 2003, 2007 + * Copyright IBM Corp. 2003, 2015 * Author(s): Thomas Spatzier * Jan Glauber (jan.glauber@de.ibm.com) + * Harald Freudenberger (freude@de.ibm.com) * * This program is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License as published by the Free @@ -28,15 +29,17 @@ #define CRYPT_S390_MSA 0x1 #define CRYPT_S390_MSA3 0x2 #define CRYPT_S390_MSA4 0x4 +#define CRYPT_S390_MSA5 0x8 /* s390 cryptographic operations */ enum crypt_s390_operations { - CRYPT_S390_KM = 0x0100, - CRYPT_S390_KMC = 0x0200, - CRYPT_S390_KIMD = 0x0300, - CRYPT_S390_KLMD = 0x0400, - CRYPT_S390_KMAC = 0x0500, - CRYPT_S390_KMCTR = 0x0600 + CRYPT_S390_KM = 0x0100, + CRYPT_S390_KMC = 0x0200, + CRYPT_S390_KIMD = 0x0300, + CRYPT_S390_KLMD = 0x0400, + CRYPT_S390_KMAC = 0x0500, + CRYPT_S390_KMCTR = 0x0600, + CRYPT_S390_PPNO = 0x0700 }; /* @@ -138,6 +141,16 @@ enum crypt_s390_kmac_func { KMAC_TDEA_192 = CRYPT_S390_KMAC | 3 }; +/* + * function codes for PPNO (PERFORM PSEUDORANDOM NUMBER + * OPERATION) instruction + */ +enum crypt_s390_ppno_func { + PPNO_QUERY = CRYPT_S390_PPNO | 0, + PPNO_SHA512_DRNG_GEN = CRYPT_S390_PPNO | 3, + PPNO_SHA512_DRNG_SEED = CRYPT_S390_PPNO | 0x83 +}; + /** * crypt_s390_km: * @func: the function code passed to KM; see crypt_s390_km_func @@ -162,11 +175,11 @@ static inline int crypt_s390_km(long func, void *param, int ret; asm volatile( - "0: .insn rre,0xb92e0000,%3,%1 \n" /* KM opcode */ - "1: brc 1,0b \n" /* handle partial completion */ + "0: .insn rre,0xb92e0000,%3,%1\n" /* KM opcode */ + "1: brc 1,0b\n" /* handle partial completion */ " la %0,0\n" "2:\n" - EX_TABLE(0b,2b) EX_TABLE(1b,2b) + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) : "=d" (ret), "+a" (__src), "+d" (__src_len), "+a" (__dest) : "d" (__func), "a" (__param), "0" (-1) : "cc", "memory"); if (ret < 0) @@ -198,11 +211,11 @@ static inline int crypt_s390_kmc(long func, void *param, int ret; asm volatile( - "0: .insn rre,0xb92f0000,%3,%1 \n" /* KMC opcode */ - "1: brc 1,0b \n" /* handle partial completion */ + "0: .insn rre,0xb92f0000,%3,%1\n" /* KMC opcode */ + "1: brc 1,0b\n" /* handle partial completion */ " la %0,0\n" "2:\n" - EX_TABLE(0b,2b) EX_TABLE(1b,2b) + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) : "=d" (ret), "+a" (__src), "+d" (__src_len), "+a" (__dest) : "d" (__func), "a" (__param), "0" (-1) : "cc", "memory"); if (ret < 0) @@ -233,11 +246,11 @@ static inline int crypt_s390_kimd(long func, void *param, int ret; asm volatile( - "0: .insn rre,0xb93e0000,%1,%1 \n" /* KIMD opcode */ - "1: brc 1,0b \n" /* handle partial completion */ + "0: .insn rre,0xb93e0000,%1,%1\n" /* KIMD opcode */ + "1: brc 1,0b\n" /* handle partial completion */ " la %0,0\n" "2:\n" - EX_TABLE(0b,2b) EX_TABLE(1b,2b) + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) : "=d" (ret), "+a" (__src), "+d" (__src_len) : "d" (__func), "a" (__param), "0" (-1) : "cc", "memory"); if (ret < 0) @@ -267,11 +280,11 @@ static inline int crypt_s390_klmd(long func, void *param, int ret; asm volatile( - "0: .insn rre,0xb93f0000,%1,%1 \n" /* KLMD opcode */ - "1: brc 1,0b \n" /* handle partial completion */ + "0: .insn rre,0xb93f0000,%1,%1\n" /* KLMD opcode */ + "1: brc 1,0b\n" /* handle partial completion */ " la %0,0\n" "2:\n" - EX_TABLE(0b,2b) EX_TABLE(1b,2b) + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) : "=d" (ret), "+a" (__src), "+d" (__src_len) : "d" (__func), "a" (__param), "0" (-1) : "cc", "memory"); if (ret < 0) @@ -302,11 +315,11 @@ static inline int crypt_s390_kmac(long func, void *param, int ret; asm volatile( - "0: .insn rre,0xb91e0000,%1,%1 \n" /* KLAC opcode */ - "1: brc 1,0b \n" /* handle partial completion */ + "0: .insn rre,0xb91e0000,%1,%1\n" /* KLAC opcode */ + "1: brc 1,0b\n" /* handle partial completion */ " la %0,0\n" "2:\n" - EX_TABLE(0b,2b) EX_TABLE(1b,2b) + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) : "=d" (ret), "+a" (__src), "+d" (__src_len) : "d" (__func), "a" (__param), "0" (-1) : "cc", "memory"); if (ret < 0) @@ -340,11 +353,11 @@ static inline int crypt_s390_kmctr(long func, void *param, u8 *dest, int ret = -1; asm volatile( - "0: .insn rrf,0xb92d0000,%3,%1,%4,0 \n" /* KMCTR opcode */ - "1: brc 1,0b \n" /* handle partial completion */ + "0: .insn rrf,0xb92d0000,%3,%1,%4,0\n" /* KMCTR opcode */ + "1: brc 1,0b\n" /* handle partial completion */ " la %0,0\n" "2:\n" - EX_TABLE(0b,2b) EX_TABLE(1b,2b) + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) : "+d" (ret), "+a" (__src), "+d" (__src_len), "+a" (__dest), "+a" (__ctr) : "d" (__func), "a" (__param) : "cc", "memory"); @@ -354,6 +367,47 @@ static inline int crypt_s390_kmctr(long func, void *param, u8 *dest, } /** + * crypt_s390_ppno: + * @func: the function code passed to PPNO; see crypt_s390_ppno_func + * @param: address of parameter block; see POP for details on each func + * @dest: address of destination memory area + * @dest_len: size of destination memory area in bytes + * @seed: address of seed data + * @seed_len: size of seed data in bytes + * + * Executes the PPNO (PERFORM PSEUDORANDOM NUMBER OPERATION) + * operation of the CPU. + * + * Returns -1 for failure, 0 for the query func, number of random + * bytes stored in dest buffer for generate function + */ +static inline int crypt_s390_ppno(long func, void *param, + u8 *dest, long dest_len, + const u8 *seed, long seed_len) +{ + register long __func asm("0") = func & CRYPT_S390_FUNC_MASK; + register void *__param asm("1") = param; /* param block (240 bytes) */ + register u8 *__dest asm("2") = dest; /* buf for recv random bytes */ + register long __dest_len asm("3") = dest_len; /* requested random bytes */ + register const u8 *__seed asm("4") = seed; /* buf with seed data */ + register long __seed_len asm("5") = seed_len; /* bytes in seed buf */ + int ret = -1; + + asm volatile ( + "0: .insn rre,0xb93c0000,%1,%5\n" /* PPNO opcode */ + "1: brc 1,0b\n" /* handle partial completion */ + " la %0,0\n" + "2:\n" + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) + : "+d" (ret), "+a"(__dest), "+d"(__dest_len) + : "d"(__func), "a"(__param), "a"(__seed), "d"(__seed_len) + : "cc", "memory"); + if (ret < 0) + return ret; + return (func & CRYPT_S390_FUNC_MASK) ? dest_len - __dest_len : 0; +} + +/** * crypt_s390_func_available: * @func: the function code of the specific function; 0 if op in general * @@ -373,6 +427,9 @@ static inline int crypt_s390_func_available(int func, return 0; if (facility_mask & CRYPT_S390_MSA4 && !test_facility(77)) return 0; + if (facility_mask & CRYPT_S390_MSA5 && !test_facility(57)) + return 0; + switch (func & CRYPT_S390_OP_MASK) { case CRYPT_S390_KM: ret = crypt_s390_km(KM_QUERY, &status, NULL, NULL, 0); @@ -390,8 +447,12 @@ static inline int crypt_s390_func_available(int func, ret = crypt_s390_kmac(KMAC_QUERY, &status, NULL, 0); break; case CRYPT_S390_KMCTR: - ret = crypt_s390_kmctr(KMCTR_QUERY, &status, NULL, NULL, 0, - NULL); + ret = crypt_s390_kmctr(KMCTR_QUERY, &status, + NULL, NULL, 0, NULL); + break; + case CRYPT_S390_PPNO: + ret = crypt_s390_ppno(PPNO_QUERY, &status, + NULL, 0, NULL, 0); break; default: return 0; @@ -419,15 +480,14 @@ static inline int crypt_s390_pcc(long func, void *param) int ret = -1; asm volatile( - "0: .insn rre,0xb92c0000,0,0 \n" /* PCC opcode */ - "1: brc 1,0b \n" /* handle partial completion */ + "0: .insn rre,0xb92c0000,0,0\n" /* PCC opcode */ + "1: brc 1,0b\n" /* handle partial completion */ " la %0,0\n" "2:\n" - EX_TABLE(0b,2b) EX_TABLE(1b,2b) + EX_TABLE(0b, 2b) EX_TABLE(1b, 2b) : "+d" (ret) : "d" (__func), "a" (__param) : "cc", "memory"); return ret; } - #endif /* _CRYPTO_ARCH_S390_CRYPT_S390_H */ diff --git a/arch/s390/crypto/prng.c b/arch/s390/crypto/prng.c index 94a35a4c1b48..9d5192c94963 100644 --- a/arch/s390/crypto/prng.c +++ b/arch/s390/crypto/prng.c @@ -1,106 +1,529 @@ /* - * Copyright IBM Corp. 2006, 2007 + * Copyright IBM Corp. 2006, 2015 * Author(s): Jan Glauber <jan.glauber@de.ibm.com> + * Harald Freudenberger <freude@de.ibm.com> * Driver for the s390 pseudo random number generator */ + +#define KMSG_COMPONENT "prng" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include <linux/fs.h> +#include <linux/fips.h> #include <linux/init.h> #include <linux/kernel.h> +#include <linux/device.h> #include <linux/miscdevice.h> #include <linux/module.h> #include <linux/moduleparam.h> +#include <linux/mutex.h> #include <linux/random.h> #include <linux/slab.h> #include <asm/debug.h> #include <asm/uaccess.h> +#include <asm/timex.h> #include "crypt_s390.h" MODULE_LICENSE("GPL"); -MODULE_AUTHOR("Jan Glauber <jan.glauber@de.ibm.com>"); +MODULE_AUTHOR("IBM Corporation"); MODULE_DESCRIPTION("s390 PRNG interface"); -static int prng_chunk_size = 256; -module_param(prng_chunk_size, int, S_IRUSR | S_IRGRP | S_IROTH); + +#define PRNG_MODE_AUTO 0 +#define PRNG_MODE_TDES 1 +#define PRNG_MODE_SHA512 2 + +static unsigned int prng_mode = PRNG_MODE_AUTO; +module_param_named(mode, prng_mode, int, 0); +MODULE_PARM_DESC(prng_mode, "PRNG mode: 0 - auto, 1 - TDES, 2 - SHA512"); + + +#define PRNG_CHUNKSIZE_TDES_MIN 8 +#define PRNG_CHUNKSIZE_TDES_MAX (64*1024) +#define PRNG_CHUNKSIZE_SHA512_MIN 64 +#define PRNG_CHUNKSIZE_SHA512_MAX (64*1024) + +static unsigned int prng_chunk_size = 256; +module_param_named(chunksize, prng_chunk_size, int, 0); MODULE_PARM_DESC(prng_chunk_size, "PRNG read chunk size in bytes"); -static int prng_entropy_limit = 4096; -module_param(prng_entropy_limit, int, S_IRUSR | S_IRGRP | S_IROTH | S_IWUSR); -MODULE_PARM_DESC(prng_entropy_limit, - "PRNG add entropy after that much bytes were produced"); + +#define PRNG_RESEED_LIMIT_TDES 4096 +#define PRNG_RESEED_LIMIT_TDES_LOWER 4096 +#define PRNG_RESEED_LIMIT_SHA512 100000 +#define PRNG_RESEED_LIMIT_SHA512_LOWER 10000 + +static unsigned int prng_reseed_limit; +module_param_named(reseed_limit, prng_reseed_limit, int, 0); +MODULE_PARM_DESC(prng_reseed_limit, "PRNG reseed limit"); + /* * Any one who considers arithmetical methods of producing random digits is, * of course, in a state of sin. -- John von Neumann */ -struct s390_prng_data { - unsigned long count; /* how many bytes were produced */ - char *buf; +static int prng_errorflag; + +#define PRNG_GEN_ENTROPY_FAILED 1 +#define PRNG_SELFTEST_FAILED 2 +#define PRNG_INSTANTIATE_FAILED 3 +#define PRNG_SEED_FAILED 4 +#define PRNG_RESEED_FAILED 5 +#define PRNG_GEN_FAILED 6 + +struct prng_ws_s { + u8 parm_block[32]; + u32 reseed_counter; + u64 byte_counter; }; -static struct s390_prng_data *p; +struct ppno_ws_s { + u32 res; + u32 reseed_counter; + u64 stream_bytes; + u8 V[112]; + u8 C[112]; +}; -/* copied from libica, use a non-zero initial parameter block */ -static unsigned char parm_block[32] = { -0x0F,0x2B,0x8E,0x63,0x8C,0x8E,0xD2,0x52,0x64,0xB7,0xA0,0x7B,0x75,0x28,0xB8,0xF4, -0x75,0x5F,0xD2,0xA6,0x8D,0x97,0x11,0xFF,0x49,0xD8,0x23,0xF3,0x7E,0x21,0xEC,0xA0, +struct prng_data_s { + struct mutex mutex; + union { + struct prng_ws_s prngws; + struct ppno_ws_s ppnows; + }; + u8 *buf; + u32 rest; + u8 *prev; }; -static int prng_open(struct inode *inode, struct file *file) +static struct prng_data_s *prng_data; + +/* initial parameter block for tdes mode, copied from libica */ +static const u8 initial_parm_block[32] __initconst = { + 0x0F, 0x2B, 0x8E, 0x63, 0x8C, 0x8E, 0xD2, 0x52, + 0x64, 0xB7, 0xA0, 0x7B, 0x75, 0x28, 0xB8, 0xF4, + 0x75, 0x5F, 0xD2, 0xA6, 0x8D, 0x97, 0x11, 0xFF, + 0x49, 0xD8, 0x23, 0xF3, 0x7E, 0x21, 0xEC, 0xA0 }; + + +/*** helper functions ***/ + +static int generate_entropy(u8 *ebuf, size_t nbytes) { - return nonseekable_open(inode, file); + int n, ret = 0; + u8 *pg, *h, hash[32]; + + pg = (u8 *) __get_free_page(GFP_KERNEL); + if (!pg) { + prng_errorflag = PRNG_GEN_ENTROPY_FAILED; + return -ENOMEM; + } + + while (nbytes) { + /* fill page with urandom bytes */ + get_random_bytes(pg, PAGE_SIZE); + /* exor page with stckf values */ + for (n = 0; n < PAGE_SIZE / sizeof(u64); n++) { + u64 *p = ((u64 *)pg) + n; + *p ^= get_tod_clock_fast(); + } + n = (nbytes < sizeof(hash)) ? nbytes : sizeof(hash); + if (n < sizeof(hash)) + h = hash; + else + h = ebuf; + /* generate sha256 from this page */ + if (crypt_s390_kimd(KIMD_SHA_256, h, + pg, PAGE_SIZE) != PAGE_SIZE) { + prng_errorflag = PRNG_GEN_ENTROPY_FAILED; + ret = -EIO; + goto out; + } + if (n < sizeof(hash)) + memcpy(ebuf, hash, n); + ret += n; + ebuf += n; + nbytes -= n; + } + +out: + free_page((unsigned long)pg); + return ret; } -static void prng_add_entropy(void) + +/*** tdes functions ***/ + +static void prng_tdes_add_entropy(void) { __u64 entropy[4]; unsigned int i; int ret; for (i = 0; i < 16; i++) { - ret = crypt_s390_kmc(KMC_PRNG, parm_block, (char *)entropy, - (char *)entropy, sizeof(entropy)); + ret = crypt_s390_kmc(KMC_PRNG, prng_data->prngws.parm_block, + (char *)entropy, (char *)entropy, + sizeof(entropy)); BUG_ON(ret < 0 || ret != sizeof(entropy)); - memcpy(parm_block, entropy, sizeof(entropy)); + memcpy(prng_data->prngws.parm_block, entropy, sizeof(entropy)); } } -static void prng_seed(int nbytes) + +static void prng_tdes_seed(int nbytes) { char buf[16]; int i = 0; - BUG_ON(nbytes > 16); + BUG_ON(nbytes > sizeof(buf)); + get_random_bytes(buf, nbytes); /* Add the entropy */ while (nbytes >= 8) { - *((__u64 *)parm_block) ^= *((__u64 *)(buf+i)); - prng_add_entropy(); + *((__u64 *)prng_data->prngws.parm_block) ^= *((__u64 *)(buf+i)); + prng_tdes_add_entropy(); i += 8; nbytes -= 8; } - prng_add_entropy(); + prng_tdes_add_entropy(); + prng_data->prngws.reseed_counter = 0; +} + + +static int __init prng_tdes_instantiate(void) +{ + int datalen; + + pr_debug("prng runs in TDES mode with " + "chunksize=%d and reseed_limit=%u\n", + prng_chunk_size, prng_reseed_limit); + + /* memory allocation, prng_data struct init, mutex init */ + datalen = sizeof(struct prng_data_s) + prng_chunk_size; + prng_data = kzalloc(datalen, GFP_KERNEL); + if (!prng_data) { + prng_errorflag = PRNG_INSTANTIATE_FAILED; + return -ENOMEM; + } + mutex_init(&prng_data->mutex); + prng_data->buf = ((u8 *)prng_data) + sizeof(struct prng_data_s); + memcpy(prng_data->prngws.parm_block, initial_parm_block, 32); + + /* initialize the PRNG, add 128 bits of entropy */ + prng_tdes_seed(16); + + return 0; } -static ssize_t prng_read(struct file *file, char __user *ubuf, size_t nbytes, - loff_t *ppos) + +static void prng_tdes_deinstantiate(void) +{ + pr_debug("The prng module stopped " + "after running in triple DES mode\n"); + kzfree(prng_data); +} + + +/*** sha512 functions ***/ + +static int __init prng_sha512_selftest(void) { - int chunk, n; + /* NIST DRBG testvector for Hash Drbg, Sha-512, Count #0 */ + static const u8 seed[] __initconst = { + 0x6b, 0x50, 0xa7, 0xd8, 0xf8, 0xa5, 0x5d, 0x7a, + 0x3d, 0xf8, 0xbb, 0x40, 0xbc, 0xc3, 0xb7, 0x22, + 0xd8, 0x70, 0x8d, 0xe6, 0x7f, 0xda, 0x01, 0x0b, + 0x03, 0xc4, 0xc8, 0x4d, 0x72, 0x09, 0x6f, 0x8c, + 0x3e, 0xc6, 0x49, 0xcc, 0x62, 0x56, 0xd9, 0xfa, + 0x31, 0xdb, 0x7a, 0x29, 0x04, 0xaa, 0xf0, 0x25 }; + static const u8 V0[] __initconst = { + 0x00, 0xad, 0xe3, 0x6f, 0x9a, 0x01, 0xc7, 0x76, + 0x61, 0x34, 0x35, 0xf5, 0x4e, 0x24, 0x74, 0x22, + 0x21, 0x9a, 0x29, 0x89, 0xc7, 0x93, 0x2e, 0x60, + 0x1e, 0xe8, 0x14, 0x24, 0x8d, 0xd5, 0x03, 0xf1, + 0x65, 0x5d, 0x08, 0x22, 0x72, 0xd5, 0xad, 0x95, + 0xe1, 0x23, 0x1e, 0x8a, 0xa7, 0x13, 0xd9, 0x2b, + 0x5e, 0xbc, 0xbb, 0x80, 0xab, 0x8d, 0xe5, 0x79, + 0xab, 0x5b, 0x47, 0x4e, 0xdd, 0xee, 0x6b, 0x03, + 0x8f, 0x0f, 0x5c, 0x5e, 0xa9, 0x1a, 0x83, 0xdd, + 0xd3, 0x88, 0xb2, 0x75, 0x4b, 0xce, 0x83, 0x36, + 0x57, 0x4b, 0xf1, 0x5c, 0xca, 0x7e, 0x09, 0xc0, + 0xd3, 0x89, 0xc6, 0xe0, 0xda, 0xc4, 0x81, 0x7e, + 0x5b, 0xf9, 0xe1, 0x01, 0xc1, 0x92, 0x05, 0xea, + 0xf5, 0x2f, 0xc6, 0xc6, 0xc7, 0x8f, 0xbc, 0xf4 }; + static const u8 C0[] __initconst = { + 0x00, 0xf4, 0xa3, 0xe5, 0xa0, 0x72, 0x63, 0x95, + 0xc6, 0x4f, 0x48, 0xd0, 0x8b, 0x5b, 0x5f, 0x8e, + 0x6b, 0x96, 0x1f, 0x16, 0xed, 0xbc, 0x66, 0x94, + 0x45, 0x31, 0xd7, 0x47, 0x73, 0x22, 0xa5, 0x86, + 0xce, 0xc0, 0x4c, 0xac, 0x63, 0xb8, 0x39, 0x50, + 0xbf, 0xe6, 0x59, 0x6c, 0x38, 0x58, 0x99, 0x1f, + 0x27, 0xa7, 0x9d, 0x71, 0x2a, 0xb3, 0x7b, 0xf9, + 0xfb, 0x17, 0x86, 0xaa, 0x99, 0x81, 0xaa, 0x43, + 0xe4, 0x37, 0xd3, 0x1e, 0x6e, 0xe5, 0xe6, 0xee, + 0xc2, 0xed, 0x95, 0x4f, 0x53, 0x0e, 0x46, 0x8a, + 0xcc, 0x45, 0xa5, 0xdb, 0x69, 0x0d, 0x81, 0xc9, + 0x32, 0x92, 0xbc, 0x8f, 0x33, 0xe6, 0xf6, 0x09, + 0x7c, 0x8e, 0x05, 0x19, 0x0d, 0xf1, 0xb6, 0xcc, + 0xf3, 0x02, 0x21, 0x90, 0x25, 0xec, 0xed, 0x0e }; + static const u8 random[] __initconst = { + 0x95, 0xb7, 0xf1, 0x7e, 0x98, 0x02, 0xd3, 0x57, + 0x73, 0x92, 0xc6, 0xa9, 0xc0, 0x80, 0x83, 0xb6, + 0x7d, 0xd1, 0x29, 0x22, 0x65, 0xb5, 0xf4, 0x2d, + 0x23, 0x7f, 0x1c, 0x55, 0xbb, 0x9b, 0x10, 0xbf, + 0xcf, 0xd8, 0x2c, 0x77, 0xa3, 0x78, 0xb8, 0x26, + 0x6a, 0x00, 0x99, 0x14, 0x3b, 0x3c, 0x2d, 0x64, + 0x61, 0x1e, 0xee, 0xb6, 0x9a, 0xcd, 0xc0, 0x55, + 0x95, 0x7c, 0x13, 0x9e, 0x8b, 0x19, 0x0c, 0x7a, + 0x06, 0x95, 0x5f, 0x2c, 0x79, 0x7c, 0x27, 0x78, + 0xde, 0x94, 0x03, 0x96, 0xa5, 0x01, 0xf4, 0x0e, + 0x91, 0x39, 0x6a, 0xcf, 0x8d, 0x7e, 0x45, 0xeb, + 0xdb, 0xb5, 0x3b, 0xbf, 0x8c, 0x97, 0x52, 0x30, + 0xd2, 0xf0, 0xff, 0x91, 0x06, 0xc7, 0x61, 0x19, + 0xae, 0x49, 0x8e, 0x7f, 0xbc, 0x03, 0xd9, 0x0f, + 0x8e, 0x4c, 0x51, 0x62, 0x7a, 0xed, 0x5c, 0x8d, + 0x42, 0x63, 0xd5, 0xd2, 0xb9, 0x78, 0x87, 0x3a, + 0x0d, 0xe5, 0x96, 0xee, 0x6d, 0xc7, 0xf7, 0xc2, + 0x9e, 0x37, 0xee, 0xe8, 0xb3, 0x4c, 0x90, 0xdd, + 0x1c, 0xf6, 0xa9, 0xdd, 0xb2, 0x2b, 0x4c, 0xbd, + 0x08, 0x6b, 0x14, 0xb3, 0x5d, 0xe9, 0x3d, 0xa2, + 0xd5, 0xcb, 0x18, 0x06, 0x69, 0x8c, 0xbd, 0x7b, + 0xbb, 0x67, 0xbf, 0xe3, 0xd3, 0x1f, 0xd2, 0xd1, + 0xdb, 0xd2, 0xa1, 0xe0, 0x58, 0xa3, 0xeb, 0x99, + 0xd7, 0xe5, 0x1f, 0x1a, 0x93, 0x8e, 0xed, 0x5e, + 0x1c, 0x1d, 0xe2, 0x3a, 0x6b, 0x43, 0x45, 0xd3, + 0x19, 0x14, 0x09, 0xf9, 0x2f, 0x39, 0xb3, 0x67, + 0x0d, 0x8d, 0xbf, 0xb6, 0x35, 0xd8, 0xe6, 0xa3, + 0x69, 0x32, 0xd8, 0x10, 0x33, 0xd1, 0x44, 0x8d, + 0x63, 0xb4, 0x03, 0xdd, 0xf8, 0x8e, 0x12, 0x1b, + 0x6e, 0x81, 0x9a, 0xc3, 0x81, 0x22, 0x6c, 0x13, + 0x21, 0xe4, 0xb0, 0x86, 0x44, 0xf6, 0x72, 0x7c, + 0x36, 0x8c, 0x5a, 0x9f, 0x7a, 0x4b, 0x3e, 0xe2 }; + int ret = 0; - int tmp; + u8 buf[sizeof(random)]; + struct ppno_ws_s ws; + + memset(&ws, 0, sizeof(ws)); + + /* initial seed */ + ret = crypt_s390_ppno(PPNO_SHA512_DRNG_SEED, + &ws, NULL, 0, + seed, sizeof(seed)); + if (ret < 0) { + pr_err("The prng self test seed operation for the " + "SHA-512 mode failed with rc=%d\n", ret); + prng_errorflag = PRNG_SELFTEST_FAILED; + return -EIO; + } + + /* check working states V and C */ + if (memcmp(ws.V, V0, sizeof(V0)) != 0 + || memcmp(ws.C, C0, sizeof(C0)) != 0) { + pr_err("The prng self test state test " + "for the SHA-512 mode failed\n"); + prng_errorflag = PRNG_SELFTEST_FAILED; + return -EIO; + } + + /* generate random bytes */ + ret = crypt_s390_ppno(PPNO_SHA512_DRNG_GEN, + &ws, buf, sizeof(buf), + NULL, 0); + if (ret < 0) { + pr_err("The prng self test generate operation for " + "the SHA-512 mode failed with rc=%d\n", ret); + prng_errorflag = PRNG_SELFTEST_FAILED; + return -EIO; + } + ret = crypt_s390_ppno(PPNO_SHA512_DRNG_GEN, + &ws, buf, sizeof(buf), + NULL, 0); + if (ret < 0) { + pr_err("The prng self test generate operation for " + "the SHA-512 mode failed with rc=%d\n", ret); + prng_errorflag = PRNG_SELFTEST_FAILED; + return -EIO; + } + + /* check against expected data */ + if (memcmp(buf, random, sizeof(random)) != 0) { + pr_err("The prng self test data test " + "for the SHA-512 mode failed\n"); + prng_errorflag = PRNG_SELFTEST_FAILED; + return -EIO; + } + + return 0; +} + + +static int __init prng_sha512_instantiate(void) +{ + int ret, datalen; + u8 seed[64]; + + pr_debug("prng runs in SHA-512 mode " + "with chunksize=%d and reseed_limit=%u\n", + prng_chunk_size, prng_reseed_limit); + + /* memory allocation, prng_data struct init, mutex init */ + datalen = sizeof(struct prng_data_s) + prng_chunk_size; + if (fips_enabled) + datalen += prng_chunk_size; + prng_data = kzalloc(datalen, GFP_KERNEL); + if (!prng_data) { + prng_errorflag = PRNG_INSTANTIATE_FAILED; + return -ENOMEM; + } + mutex_init(&prng_data->mutex); + prng_data->buf = ((u8 *)prng_data) + sizeof(struct prng_data_s); + + /* selftest */ + ret = prng_sha512_selftest(); + if (ret) + goto outfree; + + /* generate initial seed bytestring, first 48 bytes of entropy */ + ret = generate_entropy(seed, 48); + if (ret != 48) + goto outfree; + /* followed by 16 bytes of unique nonce */ + get_tod_clock_ext(seed + 48); + + /* initial seed of the ppno drng */ + ret = crypt_s390_ppno(PPNO_SHA512_DRNG_SEED, + &prng_data->ppnows, NULL, 0, + seed, sizeof(seed)); + if (ret < 0) { + prng_errorflag = PRNG_SEED_FAILED; + ret = -EIO; + goto outfree; + } + + /* if fips mode is enabled, generate a first block of random + bytes for the FIPS 140-2 Conditional Self Test */ + if (fips_enabled) { + prng_data->prev = prng_data->buf + prng_chunk_size; + ret = crypt_s390_ppno(PPNO_SHA512_DRNG_GEN, + &prng_data->ppnows, + prng_data->prev, + prng_chunk_size, + NULL, 0); + if (ret < 0 || ret != prng_chunk_size) { + prng_errorflag = PRNG_GEN_FAILED; + ret = -EIO; + goto outfree; + } + } + + return 0; + +outfree: + kfree(prng_data); + return ret; +} + + +static void prng_sha512_deinstantiate(void) +{ + pr_debug("The prng module stopped after running in SHA-512 mode\n"); + kzfree(prng_data); +} + + +static int prng_sha512_reseed(void) +{ + int ret; + u8 seed[32]; + + /* generate 32 bytes of fresh entropy */ + ret = generate_entropy(seed, sizeof(seed)); + if (ret != sizeof(seed)) + return ret; + + /* do a reseed of the ppno drng with this bytestring */ + ret = crypt_s390_ppno(PPNO_SHA512_DRNG_SEED, + &prng_data->ppnows, NULL, 0, + seed, sizeof(seed)); + if (ret) { + prng_errorflag = PRNG_RESEED_FAILED; + return -EIO; + } + + return 0; +} + + +static int prng_sha512_generate(u8 *buf, size_t nbytes) +{ + int ret; + + /* reseed needed ? */ + if (prng_data->ppnows.reseed_counter > prng_reseed_limit) { + ret = prng_sha512_reseed(); + if (ret) + return ret; + } + + /* PPNO generate */ + ret = crypt_s390_ppno(PPNO_SHA512_DRNG_GEN, + &prng_data->ppnows, buf, nbytes, + NULL, 0); + if (ret < 0 || ret != nbytes) { + prng_errorflag = PRNG_GEN_FAILED; + return -EIO; + } + + /* FIPS 140-2 Conditional Self Test */ + if (fips_enabled) { + if (!memcmp(prng_data->prev, buf, nbytes)) { + prng_errorflag = PRNG_GEN_FAILED; + return -EILSEQ; + } + memcpy(prng_data->prev, buf, nbytes); + } + + return ret; +} + + +/*** file io functions ***/ + +static int prng_open(struct inode *inode, struct file *file) +{ + return nonseekable_open(inode, file); +} + + +static ssize_t prng_tdes_read(struct file *file, char __user *ubuf, + size_t nbytes, loff_t *ppos) +{ + int chunk, n, tmp, ret = 0; + + /* lock prng_data struct */ + if (mutex_lock_interruptible(&prng_data->mutex)) + return -ERESTARTSYS; - /* nbytes can be arbitrary length, we split it into chunks */ while (nbytes) { - /* same as in extract_entropy_user in random.c */ if (need_resched()) { if (signal_pending(current)) { if (ret == 0) ret = -ERESTARTSYS; break; } + /* give mutex free before calling schedule() */ + mutex_unlock(&prng_data->mutex); schedule(); + /* occopy mutex again */ + if (mutex_lock_interruptible(&prng_data->mutex)) { + if (ret == 0) + ret = -ERESTARTSYS; + return ret; + } } /* @@ -112,12 +535,11 @@ static ssize_t prng_read(struct file *file, char __user *ubuf, size_t nbytes, /* PRNG only likes multiples of 8 bytes */ n = (chunk + 7) & -8; - if (p->count > prng_entropy_limit) - prng_seed(8); + if (prng_data->prngws.reseed_counter > prng_reseed_limit) + prng_tdes_seed(8); /* if the CPU supports PRNG stckf is present too */ - asm volatile(".insn s,0xb27c0000,%0" - : "=m" (*((unsigned long long *)p->buf)) : : "cc"); + *((unsigned long long *)prng_data->buf) = get_tod_clock_fast(); /* * Beside the STCKF the input for the TDES-EDE is the output @@ -132,35 +554,259 @@ static ssize_t prng_read(struct file *file, char __user *ubuf, size_t nbytes, * Note: you can still get strict X9.17 conformity by setting * prng_chunk_size to 8 bytes. */ - tmp = crypt_s390_kmc(KMC_PRNG, parm_block, p->buf, p->buf, n); - BUG_ON((tmp < 0) || (tmp != n)); + tmp = crypt_s390_kmc(KMC_PRNG, prng_data->prngws.parm_block, + prng_data->buf, prng_data->buf, n); + if (tmp < 0 || tmp != n) { + ret = -EIO; + break; + } - p->count += n; + prng_data->prngws.byte_counter += n; + prng_data->prngws.reseed_counter += n; - if (copy_to_user(ubuf, p->buf, chunk)) + if (copy_to_user(ubuf, prng_data->buf, chunk)) return -EFAULT; nbytes -= chunk; ret += chunk; ubuf += chunk; } + + /* unlock prng_data struct */ + mutex_unlock(&prng_data->mutex); + return ret; } -static const struct file_operations prng_fops = { + +static ssize_t prng_sha512_read(struct file *file, char __user *ubuf, + size_t nbytes, loff_t *ppos) +{ + int n, ret = 0; + u8 *p; + + /* if errorflag is set do nothing and return 'broken pipe' */ + if (prng_errorflag) + return -EPIPE; + + /* lock prng_data struct */ + if (mutex_lock_interruptible(&prng_data->mutex)) + return -ERESTARTSYS; + + while (nbytes) { + if (need_resched()) { + if (signal_pending(current)) { + if (ret == 0) + ret = -ERESTARTSYS; + break; + } + /* give mutex free before calling schedule() */ + mutex_unlock(&prng_data->mutex); + schedule(); + /* occopy mutex again */ + if (mutex_lock_interruptible(&prng_data->mutex)) { + if (ret == 0) + ret = -ERESTARTSYS; + return ret; + } + } + if (prng_data->rest) { + /* push left over random bytes from the previous read */ + p = prng_data->buf + prng_chunk_size - prng_data->rest; + n = (nbytes < prng_data->rest) ? + nbytes : prng_data->rest; + prng_data->rest -= n; + } else { + /* generate one chunk of random bytes into read buf */ + p = prng_data->buf; + n = prng_sha512_generate(p, prng_chunk_size); + if (n < 0) { + ret = n; + break; + } + if (nbytes < prng_chunk_size) { + n = nbytes; + prng_data->rest = prng_chunk_size - n; + } else { + n = prng_chunk_size; + prng_data->rest = 0; + } + } + if (copy_to_user(ubuf, p, n)) { + ret = -EFAULT; + break; + } + ubuf += n; + nbytes -= n; + ret += n; + } + + /* unlock prng_data struct */ + mutex_unlock(&prng_data->mutex); + + return ret; +} + + +/*** sysfs stuff ***/ + +static const struct file_operations prng_sha512_fops = { + .owner = THIS_MODULE, + .open = &prng_open, + .release = NULL, + .read = &prng_sha512_read, + .llseek = noop_llseek, +}; +static const struct file_operations prng_tdes_fops = { .owner = THIS_MODULE, .open = &prng_open, .release = NULL, - .read = &prng_read, + .read = &prng_tdes_read, .llseek = noop_llseek, }; -static struct miscdevice prng_dev = { +static struct miscdevice prng_sha512_dev = { + .name = "prandom", + .minor = MISC_DYNAMIC_MINOR, + .fops = &prng_sha512_fops, +}; +static struct miscdevice prng_tdes_dev = { .name = "prandom", .minor = MISC_DYNAMIC_MINOR, - .fops = &prng_fops, + .fops = &prng_tdes_fops, }; + +/* chunksize attribute (ro) */ +static ssize_t prng_chunksize_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, PAGE_SIZE, "%u\n", prng_chunk_size); +} +static DEVICE_ATTR(chunksize, 0444, prng_chunksize_show, NULL); + +/* counter attribute (ro) */ +static ssize_t prng_counter_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + u64 counter; + + if (mutex_lock_interruptible(&prng_data->mutex)) + return -ERESTARTSYS; + if (prng_mode == PRNG_MODE_SHA512) + counter = prng_data->ppnows.stream_bytes; + else + counter = prng_data->prngws.byte_counter; + mutex_unlock(&prng_data->mutex); + + return snprintf(buf, PAGE_SIZE, "%llu\n", counter); +} +static DEVICE_ATTR(byte_counter, 0444, prng_counter_show, NULL); + +/* errorflag attribute (ro) */ +static ssize_t prng_errorflag_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, PAGE_SIZE, "%d\n", prng_errorflag); +} +static DEVICE_ATTR(errorflag, 0444, prng_errorflag_show, NULL); + +/* mode attribute (ro) */ +static ssize_t prng_mode_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + if (prng_mode == PRNG_MODE_TDES) + return snprintf(buf, PAGE_SIZE, "TDES\n"); + else + return snprintf(buf, PAGE_SIZE, "SHA512\n"); +} +static DEVICE_ATTR(mode, 0444, prng_mode_show, NULL); + +/* reseed attribute (w) */ +static ssize_t prng_reseed_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + if (mutex_lock_interruptible(&prng_data->mutex)) + return -ERESTARTSYS; + prng_sha512_reseed(); + mutex_unlock(&prng_data->mutex); + + return count; +} +static DEVICE_ATTR(reseed, 0200, NULL, prng_reseed_store); + +/* reseed limit attribute (rw) */ +static ssize_t prng_reseed_limit_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, PAGE_SIZE, "%u\n", prng_reseed_limit); +} +static ssize_t prng_reseed_limit_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + unsigned limit; + + if (sscanf(buf, "%u\n", &limit) != 1) + return -EINVAL; + + if (prng_mode == PRNG_MODE_SHA512) { + if (limit < PRNG_RESEED_LIMIT_SHA512_LOWER) + return -EINVAL; + } else { + if (limit < PRNG_RESEED_LIMIT_TDES_LOWER) + return -EINVAL; + } + + prng_reseed_limit = limit; + + return count; +} +static DEVICE_ATTR(reseed_limit, 0644, + prng_reseed_limit_show, prng_reseed_limit_store); + +/* strength attribute (ro) */ +static ssize_t prng_strength_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, PAGE_SIZE, "256\n"); +} +static DEVICE_ATTR(strength, 0444, prng_strength_show, NULL); + +static struct attribute *prng_sha512_dev_attrs[] = { + &dev_attr_errorflag.attr, + &dev_attr_chunksize.attr, + &dev_attr_byte_counter.attr, + &dev_attr_mode.attr, + &dev_attr_reseed.attr, + &dev_attr_reseed_limit.attr, + &dev_attr_strength.attr, + NULL +}; +static struct attribute *prng_tdes_dev_attrs[] = { + &dev_attr_chunksize.attr, + &dev_attr_byte_counter.attr, + &dev_attr_mode.attr, + NULL +}; + +static struct attribute_group prng_sha512_dev_attr_group = { + .attrs = prng_sha512_dev_attrs +}; +static struct attribute_group prng_tdes_dev_attr_group = { + .attrs = prng_tdes_dev_attrs +}; + + +/*** module init and exit ***/ + static int __init prng_init(void) { int ret; @@ -169,43 +815,105 @@ static int __init prng_init(void) if (!crypt_s390_func_available(KMC_PRNG, CRYPT_S390_MSA)) return -EOPNOTSUPP; - if (prng_chunk_size < 8) - return -EINVAL; + /* choose prng mode */ + if (prng_mode != PRNG_MODE_TDES) { + /* check for MSA5 support for PPNO operations */ + if (!crypt_s390_func_available(PPNO_SHA512_DRNG_GEN, + CRYPT_S390_MSA5)) { + if (prng_mode == PRNG_MODE_SHA512) { + pr_err("The prng module cannot " + "start in SHA-512 mode\n"); + return -EOPNOTSUPP; + } + prng_mode = PRNG_MODE_TDES; + } else + prng_mode = PRNG_MODE_SHA512; + } - p = kmalloc(sizeof(struct s390_prng_data), GFP_KERNEL); - if (!p) - return -ENOMEM; - p->count = 0; + if (prng_mode == PRNG_MODE_SHA512) { - p->buf = kmalloc(prng_chunk_size, GFP_KERNEL); - if (!p->buf) { - ret = -ENOMEM; - goto out_free; - } + /* SHA512 mode */ - /* initialize the PRNG, add 128 bits of entropy */ - prng_seed(16); + if (prng_chunk_size < PRNG_CHUNKSIZE_SHA512_MIN + || prng_chunk_size > PRNG_CHUNKSIZE_SHA512_MAX) + return -EINVAL; + prng_chunk_size = (prng_chunk_size + 0x3f) & ~0x3f; - ret = misc_register(&prng_dev); - if (ret) - goto out_buf; - return 0; + if (prng_reseed_limit == 0) + prng_reseed_limit = PRNG_RESEED_LIMIT_SHA512; + else if (prng_reseed_limit < PRNG_RESEED_LIMIT_SHA512_LOWER) + return -EINVAL; + + ret = prng_sha512_instantiate(); + if (ret) + goto out; + + ret = misc_register(&prng_sha512_dev); + if (ret) { + prng_sha512_deinstantiate(); + goto out; + } + ret = sysfs_create_group(&prng_sha512_dev.this_device->kobj, + &prng_sha512_dev_attr_group); + if (ret) { + misc_deregister(&prng_sha512_dev); + prng_sha512_deinstantiate(); + goto out; + } -out_buf: - kfree(p->buf); -out_free: - kfree(p); + } else { + + /* TDES mode */ + + if (prng_chunk_size < PRNG_CHUNKSIZE_TDES_MIN + || prng_chunk_size > PRNG_CHUNKSIZE_TDES_MAX) + return -EINVAL; + prng_chunk_size = (prng_chunk_size + 0x07) & ~0x07; + + if (prng_reseed_limit == 0) + prng_reseed_limit = PRNG_RESEED_LIMIT_TDES; + else if (prng_reseed_limit < PRNG_RESEED_LIMIT_TDES_LOWER) + return -EINVAL; + + ret = prng_tdes_instantiate(); + if (ret) + goto out; + + ret = misc_register(&prng_tdes_dev); + if (ret) { + prng_tdes_deinstantiate(); + goto out; + } + ret = sysfs_create_group(&prng_tdes_dev.this_device->kobj, + &prng_tdes_dev_attr_group); + if (ret) { + misc_deregister(&prng_tdes_dev); + prng_tdes_deinstantiate(); + goto out; + } + + } + +out: return ret; } + static void __exit prng_exit(void) { - /* wipe me */ - kzfree(p->buf); - kfree(p); - - misc_deregister(&prng_dev); + if (prng_mode == PRNG_MODE_SHA512) { + sysfs_remove_group(&prng_sha512_dev.this_device->kobj, + &prng_sha512_dev_attr_group); + misc_deregister(&prng_sha512_dev); + prng_sha512_deinstantiate(); + } else { + sysfs_remove_group(&prng_tdes_dev.this_device->kobj, + &prng_tdes_dev_attr_group); + misc_deregister(&prng_tdes_dev); + prng_tdes_deinstantiate(); + } } + module_init(prng_init); module_exit(prng_exit); diff --git a/arch/s390/hypfs/inode.c b/arch/s390/hypfs/inode.c index 3f5c799b7fb5..d3f896a35b98 100644 --- a/arch/s390/hypfs/inode.c +++ b/arch/s390/hypfs/inode.c @@ -48,7 +48,7 @@ static struct dentry *hypfs_last_dentry; static void hypfs_update_update(struct super_block *sb) { struct hypfs_sb_info *sb_info = sb->s_fs_info; - struct inode *inode = sb_info->update_file->d_inode; + struct inode *inode = d_inode(sb_info->update_file); sb_info->last_update = get_seconds(); inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -64,7 +64,7 @@ static void hypfs_add_dentry(struct dentry *dentry) static inline int hypfs_positive(struct dentry *dentry) { - return dentry->d_inode && !d_unhashed(dentry); + return d_really_is_positive(dentry) && !d_unhashed(dentry); } static void hypfs_remove(struct dentry *dentry) @@ -72,16 +72,16 @@ static void hypfs_remove(struct dentry *dentry) struct dentry *parent; parent = dentry->d_parent; - mutex_lock(&parent->d_inode->i_mutex); + mutex_lock(&d_inode(parent)->i_mutex); if (hypfs_positive(dentry)) { if (d_is_dir(dentry)) - simple_rmdir(parent->d_inode, dentry); + simple_rmdir(d_inode(parent), dentry); else - simple_unlink(parent->d_inode, dentry); + simple_unlink(d_inode(parent), dentry); } d_delete(dentry); dput(dentry); - mutex_unlock(&parent->d_inode->i_mutex); + mutex_unlock(&d_inode(parent)->i_mutex); } static void hypfs_delete_tree(struct dentry *root) @@ -336,7 +336,7 @@ static struct dentry *hypfs_create_file(struct dentry *parent, const char *name, struct dentry *dentry; struct inode *inode; - mutex_lock(&parent->d_inode->i_mutex); + mutex_lock(&d_inode(parent)->i_mutex); dentry = lookup_one_len(name, parent, strlen(name)); if (IS_ERR(dentry)) { dentry = ERR_PTR(-ENOMEM); @@ -357,14 +357,14 @@ static struct dentry *hypfs_create_file(struct dentry *parent, const char *name, } else if (S_ISDIR(mode)) { inode->i_op = &simple_dir_inode_operations; inode->i_fop = &simple_dir_operations; - inc_nlink(parent->d_inode); + inc_nlink(d_inode(parent)); } else BUG(); inode->i_private = data; d_instantiate(dentry, inode); dget(dentry); fail: - mutex_unlock(&parent->d_inode->i_mutex); + mutex_unlock(&d_inode(parent)->i_mutex); return dentry; } diff --git a/arch/s390/include/asm/dma-mapping.h b/arch/s390/include/asm/dma-mapping.h index 709955ddaa4d..9d395961e713 100644 --- a/arch/s390/include/asm/dma-mapping.h +++ b/arch/s390/include/asm/dma-mapping.h @@ -42,7 +42,7 @@ static inline int dma_supported(struct device *dev, u64 mask) static inline bool dma_capable(struct device *dev, dma_addr_t addr, size_t size) { if (!dev->dma_mask) - return 0; + return false; return addr + size - 1 <= *dev->dma_mask; } diff --git a/arch/s390/include/asm/kexec.h b/arch/s390/include/asm/kexec.h index 694bcd6bd927..2f924bc30e35 100644 --- a/arch/s390/include/asm/kexec.h +++ b/arch/s390/include/asm/kexec.h @@ -26,6 +26,9 @@ /* Not more than 2GB */ #define KEXEC_CONTROL_MEMORY_LIMIT (1UL<<31) +/* Allocate control page with GFP_DMA */ +#define KEXEC_CONTROL_MEMORY_GFP GFP_DMA + /* Maximum address we can use for the crash control pages */ #define KEXEC_CRASH_CONTROL_MEMORY_LIMIT (-1UL) diff --git a/arch/s390/include/asm/mmu.h b/arch/s390/include/asm/mmu.h index a5e656260a70..d29ad9545b41 100644 --- a/arch/s390/include/asm/mmu.h +++ b/arch/s390/include/asm/mmu.h @@ -14,7 +14,9 @@ typedef struct { unsigned long asce_bits; unsigned long asce_limit; unsigned long vdso_base; - /* The mmu context has extended page tables. */ + /* The mmu context allocates 4K page tables. */ + unsigned int alloc_pgste:1; + /* The mmu context uses extended page tables. */ unsigned int has_pgste:1; /* The mmu context uses storage keys. */ unsigned int use_skey:1; diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h index d25d9ff10ba8..fb1b93ea3e3f 100644 --- a/arch/s390/include/asm/mmu_context.h +++ b/arch/s390/include/asm/mmu_context.h @@ -20,8 +20,11 @@ static inline int init_new_context(struct task_struct *tsk, mm->context.flush_mm = 0; mm->context.asce_bits = _ASCE_TABLE_LENGTH | _ASCE_USER_BITS; mm->context.asce_bits |= _ASCE_TYPE_REGION3; +#ifdef CONFIG_PGSTE + mm->context.alloc_pgste = page_table_allocate_pgste; mm->context.has_pgste = 0; mm->context.use_skey = 0; +#endif mm->context.asce_limit = STACK_TOP_MAX; crst_table_init((unsigned long *) mm->pgd, pgd_entry_type(mm)); return 0; diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h index ef803c202d42..a648338c434a 100644 --- a/arch/s390/include/asm/pci.h +++ b/arch/s390/include/asm/pci.h @@ -7,6 +7,7 @@ #define PCI_BAR_COUNT 6 #include <linux/pci.h> +#include <linux/mutex.h> #include <asm-generic/pci.h> #include <asm-generic/pci-dma-compat.h> #include <asm/pci_clp.h> @@ -44,10 +45,6 @@ struct zpci_fmb { u64 rpcit_ops; u64 dma_rbytes; u64 dma_wbytes; - /* software counters */ - atomic64_t allocated_pages; - atomic64_t mapped_pages; - atomic64_t unmapped_pages; } __packed __aligned(16); enum zpci_state { @@ -80,6 +77,7 @@ struct zpci_dev { u8 pft; /* pci function type */ u16 domain; + struct mutex lock; u8 pfip[CLP_PFIP_NR_SEGMENTS]; /* pci function internal path */ u32 uid; /* user defined id */ u8 util_str[CLP_UTIL_STR_LEN]; /* utility string */ @@ -111,6 +109,10 @@ struct zpci_dev { /* Function measurement block */ struct zpci_fmb *fmb; u16 fmb_update; /* update interval */ + /* software counters */ + atomic64_t allocated_pages; + atomic64_t mapped_pages; + atomic64_t unmapped_pages; enum pci_bus_speed max_bus_speed; diff --git a/arch/s390/include/asm/pgalloc.h b/arch/s390/include/asm/pgalloc.h index 51e7fb634ebc..7b7858f158b4 100644 --- a/arch/s390/include/asm/pgalloc.h +++ b/arch/s390/include/asm/pgalloc.h @@ -21,6 +21,7 @@ void crst_table_free(struct mm_struct *, unsigned long *); unsigned long *page_table_alloc(struct mm_struct *); void page_table_free(struct mm_struct *, unsigned long *); void page_table_free_rcu(struct mmu_gather *, unsigned long *, unsigned long); +extern int page_table_allocate_pgste; int set_guest_storage_key(struct mm_struct *mm, unsigned long addr, unsigned long key, bool nq); diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 989cfae9e202..ef24a212eeb7 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -12,12 +12,9 @@ #define _ASM_S390_PGTABLE_H /* - * The Linux memory management assumes a three-level page table setup. For - * s390 31 bit we "fold" the mid level into the top-level page table, so - * that we physically have the same two-level page table as the s390 mmu - * expects in 31 bit mode. For s390 64 bit we use three of the five levels - * the hardware provides (region first and region second tables are not - * used). + * The Linux memory management assumes a three-level page table setup. + * For s390 64 bit we use up to four of the five levels the hardware + * provides (region first tables are not used). * * The "pgd_xxx()" functions are trivial for a folded two-level * setup: the pgd is never bad, and a pmd always exists (as it's folded @@ -101,8 +98,8 @@ extern unsigned long zero_page_mask; #ifndef __ASSEMBLY__ /* - * The vmalloc and module area will always be on the topmost area of the kernel - * mapping. We reserve 96MB (31bit) / 128GB (64bit) for vmalloc and modules. + * The vmalloc and module area will always be on the topmost area of the + * kernel mapping. We reserve 128GB (64bit) for vmalloc and modules. * On 64 bit kernels we have a 2GB area at the top of the vmalloc area where * modules will reside. That makes sure that inter module branches always * happen without trampolines and in addition the placement within a 2GB frame @@ -131,38 +128,6 @@ static inline int is_module_addr(void *addr) } /* - * A 31 bit pagetable entry of S390 has following format: - * | PFRA | | OS | - * 0 0IP0 - * 00000000001111111111222222222233 - * 01234567890123456789012345678901 - * - * I Page-Invalid Bit: Page is not available for address-translation - * P Page-Protection Bit: Store access not possible for page - * - * A 31 bit segmenttable entry of S390 has following format: - * | P-table origin | |PTL - * 0 IC - * 00000000001111111111222222222233 - * 01234567890123456789012345678901 - * - * I Segment-Invalid Bit: Segment is not available for address-translation - * C Common-Segment Bit: Segment is not private (PoP 3-30) - * PTL Page-Table-Length: Page-table length (PTL+1*16 entries -> up to 256) - * - * The 31 bit segmenttable origin of S390 has following format: - * - * |S-table origin | | STL | - * X **GPS - * 00000000001111111111222222222233 - * 01234567890123456789012345678901 - * - * X Space-Switch event: - * G Segment-Invalid Bit: * - * P Private-Space Bit: Segment is not private (PoP 3-30) - * S Storage-Alteration: - * STL Segment-Table-Length: Segment-table length (STL+1*16 entries -> up to 2048) - * * A 64 bit pagetable entry of S390 has following format: * | PFRA |0IPC| OS | * 0000000000111111111122222222223333333333444444444455555555556666 @@ -220,7 +185,6 @@ static inline int is_module_addr(void *addr) /* Software bits in the page table entry */ #define _PAGE_PRESENT 0x001 /* SW pte present bit */ -#define _PAGE_TYPE 0x002 /* SW pte type bit */ #define _PAGE_YOUNG 0x004 /* SW pte young bit */ #define _PAGE_DIRTY 0x008 /* SW pte dirty bit */ #define _PAGE_READ 0x010 /* SW pte read bit */ @@ -240,31 +204,34 @@ static inline int is_module_addr(void *addr) * table lock held. * * The following table gives the different possible bit combinations for - * the pte hardware and software bits in the last 12 bits of a pte: + * the pte hardware and software bits in the last 12 bits of a pte + * (. unassigned bit, x don't care, t swap type): * * 842100000000 * 000084210000 * 000000008421 - * .IR...wrdytp - * empty .10...000000 - * swap .10...xxxx10 - * file .11...xxxxx0 - * prot-none, clean, old .11...000001 - * prot-none, clean, young .11...000101 - * prot-none, dirty, old .10...001001 - * prot-none, dirty, young .10...001101 - * read-only, clean, old .11...010001 - * read-only, clean, young .01...010101 - * read-only, dirty, old .11...011001 - * read-only, dirty, young .01...011101 - * read-write, clean, old .11...110001 - * read-write, clean, young .01...110101 - * read-write, dirty, old .10...111001 - * read-write, dirty, young .00...111101 + * .IR.uswrdy.p + * empty .10.00000000 + * swap .11..ttttt.0 + * prot-none, clean, old .11.xx0000.1 + * prot-none, clean, young .11.xx0001.1 + * prot-none, dirty, old .10.xx0010.1 + * prot-none, dirty, young .10.xx0011.1 + * read-only, clean, old .11.xx0100.1 + * read-only, clean, young .01.xx0101.1 + * read-only, dirty, old .11.xx0110.1 + * read-only, dirty, young .01.xx0111.1 + * read-write, clean, old .11.xx1100.1 + * read-write, clean, young .01.xx1101.1 + * read-write, dirty, old .10.xx1110.1 + * read-write, dirty, young .00.xx1111.1 + * HW-bits: R read-only, I invalid + * SW-bits: p present, y young, d dirty, r read, w write, s special, + * u unused, l large * - * pte_present is true for the bit pattern .xx...xxxxx1, (pte & 0x001) == 0x001 - * pte_none is true for the bit pattern .10...xxxx00, (pte & 0x603) == 0x400 - * pte_swap is true for the bit pattern .10...xxxx10, (pte & 0x603) == 0x402 + * pte_none is true for the bit pattern .10.00000000, pte == 0x400 + * pte_swap is true for the bit pattern .11..ooooo.0, (pte & 0x201) == 0x200 + * pte_present is true for the bit pattern .xx.xxxxxx.1, (pte & 0x001) == 0x001 */ /* Bits in the segment/region table address-space-control-element */ @@ -335,6 +302,8 @@ static inline int is_module_addr(void *addr) * read-write, dirty, young 11..0...0...11 * The segment table origin is used to distinguish empty (origin==0) from * read-write, old segment table entries (origin!=0) + * HW-bits: R read-only, I invalid + * SW-bits: y young, d dirty, r read, w write */ #define _SEGMENT_ENTRY_SPLIT_BIT 11 /* THP splitting bit number */ @@ -423,6 +392,15 @@ static inline int mm_has_pgste(struct mm_struct *mm) return 0; } +static inline int mm_alloc_pgste(struct mm_struct *mm) +{ +#ifdef CONFIG_PGSTE + if (unlikely(mm->context.alloc_pgste)) + return 1; +#endif + return 0; +} + /* * In the case that a guest uses storage keys * faults should no longer be backed by zero pages @@ -516,7 +494,7 @@ static inline int pmd_large(pmd_t pmd) return (pmd_val(pmd) & _SEGMENT_ENTRY_LARGE) != 0; } -static inline int pmd_pfn(pmd_t pmd) +static inline unsigned long pmd_pfn(pmd_t pmd) { unsigned long origin_mask; @@ -582,10 +560,9 @@ static inline int pte_none(pte_t pte) static inline int pte_swap(pte_t pte) { - /* Bit pattern: (pte & 0x603) == 0x402 */ - return (pte_val(pte) & (_PAGE_INVALID | _PAGE_PROTECT | - _PAGE_TYPE | _PAGE_PRESENT)) - == (_PAGE_INVALID | _PAGE_TYPE); + /* Bit pattern: (pte & 0x201) == 0x200 */ + return (pte_val(pte) & (_PAGE_PROTECT | _PAGE_PRESENT)) + == _PAGE_PROTECT; } static inline int pte_special(pte_t pte) @@ -1586,51 +1563,51 @@ static inline int has_transparent_hugepage(void) #endif /* CONFIG_TRANSPARENT_HUGEPAGE */ /* - * 31 bit swap entry format: - * A page-table entry has some bits we have to treat in a special way. - * Bits 0, 20 and bit 23 have to be zero, otherwise an specification - * exception will occur instead of a page translation exception. The - * specifiation exception has the bad habit not to store necessary - * information in the lowcore. - * Bits 21, 22, 30 and 31 are used to indicate the page type. - * A swap pte is indicated by bit pattern (pte & 0x603) == 0x402 - * This leaves the bits 1-19 and bits 24-29 to store type and offset. - * We use the 5 bits from 25-29 for the type and the 20 bits from 1-19 - * plus 24 for the offset. - * 0| offset |0110|o|type |00| - * 0 0000000001111111111 2222 2 22222 33 - * 0 1234567890123456789 0123 4 56789 01 - * * 64 bit swap entry format: * A page-table entry has some bits we have to treat in a special way. * Bits 52 and bit 55 have to be zero, otherwise an specification * exception will occur instead of a page translation exception. The * specifiation exception has the bad habit not to store necessary * information in the lowcore. - * Bits 53, 54, 62 and 63 are used to indicate the page type. - * A swap pte is indicated by bit pattern (pte & 0x603) == 0x402 - * This leaves the bits 0-51 and bits 56-61 to store type and offset. - * We use the 5 bits from 57-61 for the type and the 53 bits from 0-51 - * plus 56 for the offset. - * | offset |0110|o|type |00| - * 0000000000111111111122222222223333333333444444444455 5555 5 55566 66 - * 0123456789012345678901234567890123456789012345678901 2345 6 78901 23 + * Bits 54 and 63 are used to indicate the page type. + * A swap pte is indicated by bit pattern (pte & 0x201) == 0x200 + * This leaves the bits 0-51 and bits 56-62 to store type and offset. + * We use the 5 bits from 57-61 for the type and the 52 bits from 0-51 + * for the offset. + * | offset |01100|type |00| + * |0000000000111111111122222222223333333333444444444455|55555|55566|66| + * |0123456789012345678901234567890123456789012345678901|23456|78901|23| */ -#define __SWP_OFFSET_MASK (~0UL >> 11) +#define __SWP_OFFSET_MASK ((1UL << 52) - 1) +#define __SWP_OFFSET_SHIFT 12 +#define __SWP_TYPE_MASK ((1UL << 5) - 1) +#define __SWP_TYPE_SHIFT 2 static inline pte_t mk_swap_pte(unsigned long type, unsigned long offset) { pte_t pte; - offset &= __SWP_OFFSET_MASK; - pte_val(pte) = _PAGE_INVALID | _PAGE_TYPE | ((type & 0x1f) << 2) | - ((offset & 1UL) << 7) | ((offset & ~1UL) << 11); + + pte_val(pte) = _PAGE_INVALID | _PAGE_PROTECT; + pte_val(pte) |= (offset & __SWP_OFFSET_MASK) << __SWP_OFFSET_SHIFT; + pte_val(pte) |= (type & __SWP_TYPE_MASK) << __SWP_TYPE_SHIFT; return pte; } -#define __swp_type(entry) (((entry).val >> 2) & 0x1f) -#define __swp_offset(entry) (((entry).val >> 11) | (((entry).val >> 7) & 1)) -#define __swp_entry(type,offset) ((swp_entry_t) { pte_val(mk_swap_pte((type),(offset))) }) +static inline unsigned long __swp_type(swp_entry_t entry) +{ + return (entry.val >> __SWP_TYPE_SHIFT) & __SWP_TYPE_MASK; +} + +static inline unsigned long __swp_offset(swp_entry_t entry) +{ + return (entry.val >> __SWP_OFFSET_SHIFT) & __SWP_OFFSET_MASK; +} + +static inline swp_entry_t __swp_entry(unsigned long type, unsigned long offset) +{ + return (swp_entry_t) { pte_val(mk_swap_pte(type, offset)) }; +} #define __pte_to_swp_entry(pte) ((swp_entry_t) { pte_val(pte) }) #define __swp_entry_to_pte(x) ((pte_t) { (x).val }) diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index afa2bd750ffc..8cd8e7b288c5 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -110,7 +110,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = { /* upper facilities limit for kvm */ unsigned long kvm_s390_fac_list_mask[] = { 0xffe6fffbfcfdfc40UL, - 0x205c800000000000UL, + 0x005c800000000000UL, }; unsigned long kvm_s390_fac_list_mask_size(void) diff --git a/arch/s390/mm/hugetlbpage.c b/arch/s390/mm/hugetlbpage.c index 210ffede0153..e617e74b7be2 100644 --- a/arch/s390/mm/hugetlbpage.c +++ b/arch/s390/mm/hugetlbpage.c @@ -14,20 +14,23 @@ static inline pmd_t __pte_to_pmd(pte_t pte) /* * Convert encoding pte bits pmd bits - * .IR...wrdytp dy..R...I...wr - * empty .10...000000 -> 00..0...1...00 - * prot-none, clean, old .11...000001 -> 00..1...1...00 - * prot-none, clean, young .11...000101 -> 01..1...1...00 - * prot-none, dirty, old .10...001001 -> 10..1...1...00 - * prot-none, dirty, young .10...001101 -> 11..1...1...00 - * read-only, clean, old .11...010001 -> 00..1...1...01 - * read-only, clean, young .01...010101 -> 01..1...0...01 - * read-only, dirty, old .11...011001 -> 10..1...1...01 - * read-only, dirty, young .01...011101 -> 11..1...0...01 - * read-write, clean, old .11...110001 -> 00..0...1...11 - * read-write, clean, young .01...110101 -> 01..0...0...11 - * read-write, dirty, old .10...111001 -> 10..0...1...11 - * read-write, dirty, young .00...111101 -> 11..0...0...11 + * lIR.uswrdy.p dy..R...I...wr + * empty 010.000000.0 -> 00..0...1...00 + * prot-none, clean, old 111.000000.1 -> 00..1...1...00 + * prot-none, clean, young 111.000001.1 -> 01..1...1...00 + * prot-none, dirty, old 111.000010.1 -> 10..1...1...00 + * prot-none, dirty, young 111.000011.1 -> 11..1...1...00 + * read-only, clean, old 111.000100.1 -> 00..1...1...01 + * read-only, clean, young 101.000101.1 -> 01..1...0...01 + * read-only, dirty, old 111.000110.1 -> 10..1...1...01 + * read-only, dirty, young 101.000111.1 -> 11..1...0...01 + * read-write, clean, old 111.001100.1 -> 00..1...1...11 + * read-write, clean, young 101.001101.1 -> 01..1...0...11 + * read-write, dirty, old 110.001110.1 -> 10..0...1...11 + * read-write, dirty, young 100.001111.1 -> 11..0...0...11 + * HW-bits: R read-only, I invalid + * SW-bits: p present, y young, d dirty, r read, w write, s special, + * u unused, l large */ if (pte_present(pte)) { pmd_val(pmd) = pte_val(pte) & PAGE_MASK; @@ -48,20 +51,23 @@ static inline pte_t __pmd_to_pte(pmd_t pmd) /* * Convert encoding pmd bits pte bits - * dy..R...I...wr .IR...wrdytp - * empty 00..0...1...00 -> .10...001100 - * prot-none, clean, old 00..0...1...00 -> .10...000001 - * prot-none, clean, young 01..0...1...00 -> .10...000101 - * prot-none, dirty, old 10..0...1...00 -> .10...001001 - * prot-none, dirty, young 11..0...1...00 -> .10...001101 - * read-only, clean, old 00..1...1...01 -> .11...010001 - * read-only, clean, young 01..1...1...01 -> .11...010101 - * read-only, dirty, old 10..1...1...01 -> .11...011001 - * read-only, dirty, young 11..1...1...01 -> .11...011101 - * read-write, clean, old 00..0...1...11 -> .10...110001 - * read-write, clean, young 01..0...1...11 -> .10...110101 - * read-write, dirty, old 10..0...1...11 -> .10...111001 - * read-write, dirty, young 11..0...1...11 -> .10...111101 + * dy..R...I...wr lIR.uswrdy.p + * empty 00..0...1...00 -> 010.000000.0 + * prot-none, clean, old 00..1...1...00 -> 111.000000.1 + * prot-none, clean, young 01..1...1...00 -> 111.000001.1 + * prot-none, dirty, old 10..1...1...00 -> 111.000010.1 + * prot-none, dirty, young 11..1...1...00 -> 111.000011.1 + * read-only, clean, old 00..1...1...01 -> 111.000100.1 + * read-only, clean, young 01..1...0...01 -> 101.000101.1 + * read-only, dirty, old 10..1...1...01 -> 111.000110.1 + * read-only, dirty, young 11..1...0...01 -> 101.000111.1 + * read-write, clean, old 00..1...1...11 -> 111.001100.1 + * read-write, clean, young 01..1...0...11 -> 101.001101.1 + * read-write, dirty, old 10..0...1...11 -> 110.001110.1 + * read-write, dirty, young 11..0...0...11 -> 100.001111.1 + * HW-bits: R read-only, I invalid + * SW-bits: p present, y young, d dirty, r read, w write, s special, + * u unused, l large */ if (pmd_present(pmd)) { pte_val(pte) = pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN_LARGE; @@ -70,8 +76,8 @@ static inline pte_t __pmd_to_pte(pmd_t pmd) pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_WRITE) << 4; pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_INVALID) << 5; pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT); - pmd_val(pmd) |= (pte_val(pte) & _PAGE_DIRTY) << 10; - pmd_val(pmd) |= (pte_val(pte) & _PAGE_YOUNG) << 10; + pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_DIRTY) >> 10; + pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_YOUNG) >> 10; } else pte_val(pte) = _PAGE_INVALID; return pte; diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 33f589459113..b33f66110ca9 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -18,6 +18,7 @@ #include <linux/rcupdate.h> #include <linux/slab.h> #include <linux/swapops.h> +#include <linux/sysctl.h> #include <linux/ksm.h> #include <linux/mman.h> @@ -920,6 +921,40 @@ unsigned long get_guest_storage_key(struct mm_struct *mm, unsigned long addr) } EXPORT_SYMBOL(get_guest_storage_key); +static int page_table_allocate_pgste_min = 0; +static int page_table_allocate_pgste_max = 1; +int page_table_allocate_pgste = 0; +EXPORT_SYMBOL(page_table_allocate_pgste); + +static struct ctl_table page_table_sysctl[] = { + { + .procname = "allocate_pgste", + .data = &page_table_allocate_pgste, + .maxlen = sizeof(int), + .mode = S_IRUGO | S_IWUSR, + .proc_handler = proc_dointvec, + .extra1 = &page_table_allocate_pgste_min, + .extra2 = &page_table_allocate_pgste_max, + }, + { } +}; + +static struct ctl_table page_table_sysctl_dir[] = { + { + .procname = "vm", + .maxlen = 0, + .mode = 0555, + .child = page_table_sysctl, + }, + { } +}; + +static int __init page_table_register_sysctl(void) +{ + return register_sysctl_table(page_table_sysctl_dir) ? 0 : -ENOMEM; +} +__initcall(page_table_register_sysctl); + #else /* CONFIG_PGSTE */ static inline int page_table_with_pgste(struct page *page) @@ -963,7 +998,7 @@ unsigned long *page_table_alloc(struct mm_struct *mm) struct page *uninitialized_var(page); unsigned int mask, bit; - if (mm_has_pgste(mm)) + if (mm_alloc_pgste(mm)) return page_table_alloc_pgste(mm); /* Allocate fragments of a 4K page as 1K/2K page table */ spin_lock_bh(&mm->context.list_lock); @@ -1165,116 +1200,25 @@ static inline void thp_split_mm(struct mm_struct *mm) } #endif /* CONFIG_TRANSPARENT_HUGEPAGE */ -static unsigned long page_table_realloc_pmd(struct mmu_gather *tlb, - struct mm_struct *mm, pud_t *pud, - unsigned long addr, unsigned long end) -{ - unsigned long next, *table, *new; - struct page *page; - spinlock_t *ptl; - pmd_t *pmd; - - pmd = pmd_offset(pud, addr); - do { - next = pmd_addr_end(addr, end); -again: - if (pmd_none_or_clear_bad(pmd)) - continue; - table = (unsigned long *) pmd_deref(*pmd); - page = pfn_to_page(__pa(table) >> PAGE_SHIFT); - if (page_table_with_pgste(page)) - continue; - /* Allocate new page table with pgstes */ - new = page_table_alloc_pgste(mm); - if (!new) - return -ENOMEM; - - ptl = pmd_lock(mm, pmd); - if (likely((unsigned long *) pmd_deref(*pmd) == table)) { - /* Nuke pmd entry pointing to the "short" page table */ - pmdp_flush_lazy(mm, addr, pmd); - pmd_clear(pmd); - /* Copy ptes from old table to new table */ - memcpy(new, table, PAGE_SIZE/2); - clear_table(table, _PAGE_INVALID, PAGE_SIZE/2); - /* Establish new table */ - pmd_populate(mm, pmd, (pte_t *) new); - /* Free old table with rcu, there might be a walker! */ - page_table_free_rcu(tlb, table, addr); - new = NULL; - } - spin_unlock(ptl); - if (new) { - page_table_free_pgste(new); - goto again; - } - } while (pmd++, addr = next, addr != end); - - return addr; -} - -static unsigned long page_table_realloc_pud(struct mmu_gather *tlb, - struct mm_struct *mm, pgd_t *pgd, - unsigned long addr, unsigned long end) -{ - unsigned long next; - pud_t *pud; - - pud = pud_offset(pgd, addr); - do { - next = pud_addr_end(addr, end); - if (pud_none_or_clear_bad(pud)) - continue; - next = page_table_realloc_pmd(tlb, mm, pud, addr, next); - if (unlikely(IS_ERR_VALUE(next))) - return next; - } while (pud++, addr = next, addr != end); - - return addr; -} - -static unsigned long page_table_realloc(struct mmu_gather *tlb, struct mm_struct *mm, - unsigned long addr, unsigned long end) -{ - unsigned long next; - pgd_t *pgd; - - pgd = pgd_offset(mm, addr); - do { - next = pgd_addr_end(addr, end); - if (pgd_none_or_clear_bad(pgd)) - continue; - next = page_table_realloc_pud(tlb, mm, pgd, addr, next); - if (unlikely(IS_ERR_VALUE(next))) - return next; - } while (pgd++, addr = next, addr != end); - - return 0; -} - /* * switch on pgstes for its userspace process (for kvm) */ int s390_enable_sie(void) { - struct task_struct *tsk = current; - struct mm_struct *mm = tsk->mm; - struct mmu_gather tlb; + struct mm_struct *mm = current->mm; /* Do we have pgstes? if yes, we are done */ - if (mm_has_pgste(tsk->mm)) + if (mm_has_pgste(mm)) return 0; - + /* Fail if the page tables are 2K */ + if (!mm_alloc_pgste(mm)) + return -EINVAL; down_write(&mm->mmap_sem); + mm->context.has_pgste = 1; /* split thp mappings and disable thp for future mappings */ thp_split_mm(mm); - /* Reallocate the page tables with pgstes */ - tlb_gather_mmu(&tlb, mm, 0, TASK_SIZE); - if (!page_table_realloc(&tlb, mm, 0, TASK_SIZE)) - mm->context.has_pgste = 1; - tlb_finish_mmu(&tlb, 0, TASK_SIZE); up_write(&mm->mmap_sem); - return mm->context.has_pgste ? 0 : -ENOMEM; + return 0; } EXPORT_SYMBOL_GPL(s390_enable_sie); diff --git a/arch/s390/net/bpf_jit.S b/arch/s390/net/bpf_jit.S index ba44c9f55346..a1c917d881ec 100644 --- a/arch/s390/net/bpf_jit.S +++ b/arch/s390/net/bpf_jit.S @@ -1,134 +1,115 @@ /* * BPF Jit compiler for s390, help functions. * - * Copyright IBM Corp. 2012 + * Copyright IBM Corp. 2012,2015 * * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> + * Michael Holzheu <holzheu@linux.vnet.ibm.com> */ + #include <linux/linkage.h> +#include "bpf_jit.h" /* * Calling convention: - * registers %r2, %r6-%r8, %r10-%r11, %r13, %r15 are call saved - * %r2: skb pointer - * %r3: offset parameter - * %r5: BPF A accumulator - * %r8: return address - * %r9: save register for skb pointer - * %r10: skb->data - * %r11: skb->len - skb->data_len (headlen) - * %r12: BPF X accumulator + * registers %r7-%r10, %r11,%r13, and %r15 are call saved + * + * Input (64 bit): + * %r3 (%b2) = offset into skb data + * %r6 (%b5) = return address + * %r7 (%b6) = skb pointer + * %r12 = skb data pointer + * + * Output: + * %r14= %b0 = return value (read skb value) + * + * Work registers: %r2,%r4,%r5,%r14 * * skb_copy_bits takes 4 parameters: * %r2 = skb pointer * %r3 = offset into skb data * %r4 = pointer to temp buffer * %r5 = length to copy + * Return value in %r2: 0 = ok + * + * bpf_internal_load_pointer_neg_helper takes 3 parameters: + * %r2 = skb pointer + * %r3 = offset into data + * %r4 = length to copy + * Return value in %r2: Pointer to data */ -#define SKBDATA %r8 - - /* A = *(u32 *) (skb->data+K+X) */ -ENTRY(sk_load_word_ind) - ar %r3,%r12 # offset += X - bmr %r8 # < 0 -> return with cc - - /* A = *(u32 *) (skb->data+K) */ -ENTRY(sk_load_word) - llgfr %r1,%r3 # extend offset - ahi %r3,4 # offset + 4 - clr %r11,%r3 # hlen <= offset + 4 ? - jl sk_load_word_slow - l %r5,0(%r1,%r10) # get word from skb - xr %r1,%r1 # set cc to zero - br %r8 -sk_load_word_slow: - lgr %r9,%r2 # save %r2 - lgr %r3,%r1 # offset - la %r4,160(%r15) # pointer to temp buffer - lghi %r5,4 # 4 bytes - brasl %r14,skb_copy_bits # get data from skb - l %r5,160(%r15) # load result from temp buffer - ltgr %r2,%r2 # set cc to (%r2 != 0) - lgr %r2,%r9 # restore %r2 - br %r8 +#define SKF_MAX_NEG_OFF -0x200000 /* SKF_LL_OFF from filter.h */ - /* A = *(u16 *) (skb->data+K+X) */ -ENTRY(sk_load_half_ind) - ar %r3,%r12 # offset += X - bmr %r8 # < 0 -> return with cc - - /* A = *(u16 *) (skb->data+K) */ -ENTRY(sk_load_half) - llgfr %r1,%r3 # extend offset - ahi %r3,2 # offset + 2 - clr %r11,%r3 # hlen <= offset + 2 ? - jl sk_load_half_slow - llgh %r5,0(%r1,%r10) # get half from skb - xr %r1,%r1 # set cc to zero - br %r8 - -sk_load_half_slow: - lgr %r9,%r2 # save %r2 - lgr %r3,%r1 # offset - la %r4,162(%r15) # pointer to temp buffer - lghi %r5,2 # 2 bytes - brasl %r14,skb_copy_bits # get data from skb - xc 160(2,%r15),160(%r15) - l %r5,160(%r15) # load result from temp buffer - ltgr %r2,%r2 # set cc to (%r2 != 0) - lgr %r2,%r9 # restore %r2 - br %r8 +/* + * Load SIZE bytes from SKB + */ +#define sk_load_common(NAME, SIZE, LOAD) \ +ENTRY(sk_load_##NAME); \ + ltgr %r3,%r3; /* Is offset negative? */ \ + jl sk_load_##NAME##_slow_neg; \ +ENTRY(sk_load_##NAME##_pos); \ + aghi %r3,SIZE; /* Offset + SIZE */ \ + clg %r3,STK_OFF_HLEN(%r15); /* Offset + SIZE > hlen? */ \ + jh sk_load_##NAME##_slow; \ + LOAD %r14,-SIZE(%r3,%r12); /* Get data from skb */ \ + b OFF_OK(%r6); /* Return */ \ + \ +sk_load_##NAME##_slow:; \ + lgr %r2,%r7; /* Arg1 = skb pointer */ \ + aghi %r3,-SIZE; /* Arg2 = offset */ \ + la %r4,STK_OFF_TMP(%r15); /* Arg3 = temp bufffer */ \ + lghi %r5,SIZE; /* Arg4 = size */ \ + brasl %r14,skb_copy_bits; /* Get data from skb */ \ + LOAD %r14,STK_OFF_TMP(%r15); /* Load from temp bufffer */ \ + ltgr %r2,%r2; /* Set cc to (%r2 != 0) */ \ + br %r6; /* Return */ - /* A = *(u8 *) (skb->data+K+X) */ -ENTRY(sk_load_byte_ind) - ar %r3,%r12 # offset += X - bmr %r8 # < 0 -> return with cc +sk_load_common(word, 4, llgf) /* r14 = *(u32 *) (skb->data+offset) */ +sk_load_common(half, 2, llgh) /* r14 = *(u16 *) (skb->data+offset) */ - /* A = *(u8 *) (skb->data+K) */ +/* + * Load 1 byte from SKB (optimized version) + */ + /* r14 = *(u8 *) (skb->data+offset) */ ENTRY(sk_load_byte) - llgfr %r1,%r3 # extend offset - clr %r11,%r3 # hlen < offset ? - jle sk_load_byte_slow - lhi %r5,0 - ic %r5,0(%r1,%r10) # get byte from skb - xr %r1,%r1 # set cc to zero - br %r8 + ltgr %r3,%r3 # Is offset negative? + jl sk_load_byte_slow_neg +ENTRY(sk_load_byte_pos) + clg %r3,STK_OFF_HLEN(%r15) # Offset >= hlen? + jnl sk_load_byte_slow + llgc %r14,0(%r3,%r12) # Get byte from skb + b OFF_OK(%r6) # Return OK sk_load_byte_slow: - lgr %r9,%r2 # save %r2 - lgr %r3,%r1 # offset - la %r4,163(%r15) # pointer to temp buffer - lghi %r5,1 # 1 byte - brasl %r14,skb_copy_bits # get data from skb - xc 160(3,%r15),160(%r15) - l %r5,160(%r15) # load result from temp buffer - ltgr %r2,%r2 # set cc to (%r2 != 0) - lgr %r2,%r9 # restore %r2 - br %r8 + lgr %r2,%r7 # Arg1 = skb pointer + # Arg2 = offset + la %r4,STK_OFF_TMP(%r15) # Arg3 = pointer to temp buffer + lghi %r5,1 # Arg4 = size (1 byte) + brasl %r14,skb_copy_bits # Get data from skb + llgc %r14,STK_OFF_TMP(%r15) # Load result from temp buffer + ltgr %r2,%r2 # Set cc to (%r2 != 0) + br %r6 # Return cc + +#define sk_negative_common(NAME, SIZE, LOAD) \ +sk_load_##NAME##_slow_neg:; \ + cgfi %r3,SKF_MAX_NEG_OFF; \ + jl bpf_error; \ + lgr %r2,%r7; /* Arg1 = skb pointer */ \ + /* Arg2 = offset */ \ + lghi %r4,SIZE; /* Arg3 = size */ \ + brasl %r14,bpf_internal_load_pointer_neg_helper; \ + ltgr %r2,%r2; \ + jz bpf_error; \ + LOAD %r14,0(%r2); /* Get data from pointer */ \ + xr %r3,%r3; /* Set cc to zero */ \ + br %r6; /* Return cc */ - /* X = (*(u8 *)(skb->data+K) & 0xf) << 2 */ -ENTRY(sk_load_byte_msh) - llgfr %r1,%r3 # extend offset - clr %r11,%r3 # hlen < offset ? - jle sk_load_byte_msh_slow - lhi %r12,0 - ic %r12,0(%r1,%r10) # get byte from skb - nill %r12,0x0f - sll %r12,2 - xr %r1,%r1 # set cc to zero - br %r8 +sk_negative_common(word, 4, llgf) +sk_negative_common(half, 2, llgh) +sk_negative_common(byte, 1, llgc) -sk_load_byte_msh_slow: - lgr %r9,%r2 # save %r2 - lgr %r3,%r1 # offset - la %r4,163(%r15) # pointer to temp buffer - lghi %r5,1 # 1 byte - brasl %r14,skb_copy_bits # get data from skb - xc 160(3,%r15),160(%r15) - l %r12,160(%r15) # load result from temp buffer - nill %r12,0x0f - sll %r12,2 - ltgr %r2,%r2 # set cc to (%r2 != 0) - lgr %r2,%r9 # restore %r2 - br %r8 +bpf_error: +# force a return 0 from jit handler + ltgr %r15,%r15 # Set condition code + br %r6 diff --git a/arch/s390/net/bpf_jit.h b/arch/s390/net/bpf_jit.h new file mode 100644 index 000000000000..ba8593a515ba --- /dev/null +++ b/arch/s390/net/bpf_jit.h @@ -0,0 +1,58 @@ +/* + * BPF Jit compiler defines + * + * Copyright IBM Corp. 2012,2015 + * + * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> + * Michael Holzheu <holzheu@linux.vnet.ibm.com> + */ + +#ifndef __ARCH_S390_NET_BPF_JIT_H +#define __ARCH_S390_NET_BPF_JIT_H + +#ifndef __ASSEMBLY__ + +#include <linux/filter.h> +#include <linux/types.h> + +extern u8 sk_load_word_pos[], sk_load_half_pos[], sk_load_byte_pos[]; +extern u8 sk_load_word[], sk_load_half[], sk_load_byte[]; + +#endif /* __ASSEMBLY__ */ + +/* + * Stackframe layout (packed stack): + * + * ^ high + * +---------------+ | + * | old backchain | | + * +---------------+ | + * | r15 - r6 | | + * BFP -> +===============+ | + * | | | + * | BPF stack | | + * | | | + * +---------------+ | + * | 8 byte hlen | | + * R15+168 -> +---------------+ | + * | 4 byte align | | + * +---------------+ | + * | 4 byte temp | | + * | for bpf_jit.S | | + * R15+160 -> +---------------+ | + * | new backchain | | + * R15+152 -> +---------------+ | + * | + 152 byte SA | | + * R15 -> +---------------+ + low + * + * We get 160 bytes stack space from calling function, but only use + * 11 * 8 byte (old backchain + r15 - r6) for storing registers. + */ +#define STK_OFF (MAX_BPF_STACK + 8 + 4 + 4 + (160 - 11 * 8)) +#define STK_OFF_TMP 160 /* Offset of tmp buffer on stack */ +#define STK_OFF_HLEN 168 /* Offset of SKB header length on stack */ + +/* Offset to skip condition code check */ +#define OFF_OK 4 + +#endif /* __ARCH_S390_NET_BPF_JIT_H */ diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c index bbd1981cc150..20c146d1251a 100644 --- a/arch/s390/net/bpf_jit_comp.c +++ b/arch/s390/net/bpf_jit_comp.c @@ -1,817 +1,1210 @@ /* * BPF Jit compiler for s390. * - * Copyright IBM Corp. 2012 + * Minimum build requirements: + * + * - HAVE_MARCH_Z196_FEATURES: laal, laalg + * - HAVE_MARCH_Z10_FEATURES: msfi, cgrj, clgrj + * - HAVE_MARCH_Z9_109_FEATURES: alfi, llilf, clfi, oilf, nilf + * - PACK_STACK + * - 64BIT + * + * Copyright IBM Corp. 2012,2015 * * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> + * Michael Holzheu <holzheu@linux.vnet.ibm.com> */ + +#define KMSG_COMPONENT "bpf_jit" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + #include <linux/netdevice.h> -#include <linux/if_vlan.h> #include <linux/filter.h> #include <linux/init.h> #include <asm/cacheflush.h> -#include <asm/facility.h> #include <asm/dis.h> +#include "bpf_jit.h" -/* - * Conventions: - * %r2 = skb pointer - * %r3 = offset parameter - * %r4 = scratch register / length parameter - * %r5 = BPF A accumulator - * %r8 = return address - * %r9 = save register for skb pointer - * %r10 = skb->data - * %r11 = skb->len - skb->data_len (headlen) - * %r12 = BPF X accumulator - * %r13 = literal pool pointer - * 0(%r15) - 63(%r15) scratch memory array with BPF_MEMWORDS - */ int bpf_jit_enable __read_mostly; +struct bpf_jit { + u32 seen; /* Flags to remember seen eBPF instructions */ + u32 seen_reg[16]; /* Array to remember which registers are used */ + u32 *addrs; /* Array with relative instruction addresses */ + u8 *prg_buf; /* Start of program */ + int size; /* Size of program and literal pool */ + int size_prg; /* Size of program */ + int prg; /* Current position in program */ + int lit_start; /* Start of literal pool */ + int lit; /* Current position in literal pool */ + int base_ip; /* Base address for literal pool */ + int ret0_ip; /* Address of return 0 */ + int exit_ip; /* Address of exit */ +}; + +#define BPF_SIZE_MAX 4096 /* Max size for program */ + +#define SEEN_SKB 1 /* skb access */ +#define SEEN_MEM 2 /* use mem[] for temporary storage */ +#define SEEN_RET0 4 /* ret0_ip points to a valid return 0 */ +#define SEEN_LITERAL 8 /* code uses literals */ +#define SEEN_FUNC 16 /* calls C functions */ +#define SEEN_STACK (SEEN_FUNC | SEEN_MEM | SEEN_SKB) + /* - * assembly code in arch/x86/net/bpf_jit.S + * s390 registers */ -extern u8 sk_load_word[], sk_load_half[], sk_load_byte[], sk_load_byte_msh[]; -extern u8 sk_load_word_ind[], sk_load_half_ind[], sk_load_byte_ind[]; +#define REG_W0 (__MAX_BPF_REG+0) /* Work register 1 (even) */ +#define REG_W1 (__MAX_BPF_REG+1) /* Work register 2 (odd) */ +#define REG_SKB_DATA (__MAX_BPF_REG+2) /* SKB data register */ +#define REG_L (__MAX_BPF_REG+3) /* Literal pool register */ +#define REG_15 (__MAX_BPF_REG+4) /* Register 15 */ +#define REG_0 REG_W0 /* Register 0 */ +#define REG_2 BPF_REG_1 /* Register 2 */ +#define REG_14 BPF_REG_0 /* Register 14 */ -struct bpf_jit { - unsigned int seen; - u8 *start; - u8 *prg; - u8 *mid; - u8 *lit; - u8 *end; - u8 *base_ip; - u8 *ret0_ip; - u8 *exit_ip; - unsigned int off_load_word; - unsigned int off_load_half; - unsigned int off_load_byte; - unsigned int off_load_bmsh; - unsigned int off_load_iword; - unsigned int off_load_ihalf; - unsigned int off_load_ibyte; +/* + * Mapping of BPF registers to s390 registers + */ +static const int reg2hex[] = { + /* Return code */ + [BPF_REG_0] = 14, + /* Function parameters */ + [BPF_REG_1] = 2, + [BPF_REG_2] = 3, + [BPF_REG_3] = 4, + [BPF_REG_4] = 5, + [BPF_REG_5] = 6, + /* Call saved registers */ + [BPF_REG_6] = 7, + [BPF_REG_7] = 8, + [BPF_REG_8] = 9, + [BPF_REG_9] = 10, + /* BPF stack pointer */ + [BPF_REG_FP] = 13, + /* SKB data pointer */ + [REG_SKB_DATA] = 12, + /* Work registers for s390x backend */ + [REG_W0] = 0, + [REG_W1] = 1, + [REG_L] = 11, + [REG_15] = 15, }; -#define BPF_SIZE_MAX 4096 /* Max size for program */ +static inline u32 reg(u32 dst_reg, u32 src_reg) +{ + return reg2hex[dst_reg] << 4 | reg2hex[src_reg]; +} + +static inline u32 reg_high(u32 reg) +{ + return reg2hex[reg] << 4; +} + +static inline void reg_set_seen(struct bpf_jit *jit, u32 b1) +{ + u32 r1 = reg2hex[b1]; + + if (!jit->seen_reg[r1] && r1 >= 6 && r1 <= 15) + jit->seen_reg[r1] = 1; +} + +#define REG_SET_SEEN(b1) \ +({ \ + reg_set_seen(jit, b1); \ +}) + +#define REG_SEEN(b1) jit->seen_reg[reg2hex[(b1)]] + +/* + * EMIT macros for code generation + */ + +#define _EMIT2(op) \ +({ \ + if (jit->prg_buf) \ + *(u16 *) (jit->prg_buf + jit->prg) = op; \ + jit->prg += 2; \ +}) -#define SEEN_DATAREF 1 /* might call external helpers */ -#define SEEN_XREG 2 /* ebx is used */ -#define SEEN_MEM 4 /* use mem[] for temporary storage */ -#define SEEN_RET0 8 /* pc_ret0 points to a valid return 0 */ -#define SEEN_LITERAL 16 /* code uses literals */ -#define SEEN_LOAD_WORD 32 /* code uses sk_load_word */ -#define SEEN_LOAD_HALF 64 /* code uses sk_load_half */ -#define SEEN_LOAD_BYTE 128 /* code uses sk_load_byte */ -#define SEEN_LOAD_BMSH 256 /* code uses sk_load_byte_msh */ -#define SEEN_LOAD_IWORD 512 /* code uses sk_load_word_ind */ -#define SEEN_LOAD_IHALF 1024 /* code uses sk_load_half_ind */ -#define SEEN_LOAD_IBYTE 2048 /* code uses sk_load_byte_ind */ - -#define EMIT2(op) \ -({ \ - if (jit->prg + 2 <= jit->mid) \ - *(u16 *) jit->prg = op; \ - jit->prg += 2; \ +#define EMIT2(op, b1, b2) \ +({ \ + _EMIT2(op | reg(b1, b2)); \ + REG_SET_SEEN(b1); \ + REG_SET_SEEN(b2); \ }) -#define EMIT4(op) \ -({ \ - if (jit->prg + 4 <= jit->mid) \ - *(u32 *) jit->prg = op; \ - jit->prg += 4; \ +#define _EMIT4(op) \ +({ \ + if (jit->prg_buf) \ + *(u32 *) (jit->prg_buf + jit->prg) = op; \ + jit->prg += 4; \ }) -#define EMIT4_DISP(op, disp) \ -({ \ - unsigned int __disp = (disp) & 0xfff; \ - EMIT4(op | __disp); \ +#define EMIT4(op, b1, b2) \ +({ \ + _EMIT4(op | reg(b1, b2)); \ + REG_SET_SEEN(b1); \ + REG_SET_SEEN(b2); \ }) -#define EMIT4_IMM(op, imm) \ -({ \ - unsigned int __imm = (imm) & 0xffff; \ - EMIT4(op | __imm); \ +#define EMIT4_RRF(op, b1, b2, b3) \ +({ \ + _EMIT4(op | reg_high(b3) << 8 | reg(b1, b2)); \ + REG_SET_SEEN(b1); \ + REG_SET_SEEN(b2); \ + REG_SET_SEEN(b3); \ }) -#define EMIT4_PCREL(op, pcrel) \ -({ \ - long __pcrel = ((pcrel) >> 1) & 0xffff; \ - EMIT4(op | __pcrel); \ +#define _EMIT4_DISP(op, disp) \ +({ \ + unsigned int __disp = (disp) & 0xfff; \ + _EMIT4(op | __disp); \ }) -#define EMIT6(op1, op2) \ -({ \ - if (jit->prg + 6 <= jit->mid) { \ - *(u32 *) jit->prg = op1; \ - *(u16 *) (jit->prg + 4) = op2; \ - } \ - jit->prg += 6; \ +#define EMIT4_DISP(op, b1, b2, disp) \ +({ \ + _EMIT4_DISP(op | reg_high(b1) << 16 | \ + reg_high(b2) << 8, disp); \ + REG_SET_SEEN(b1); \ + REG_SET_SEEN(b2); \ }) -#define EMIT6_DISP(op1, op2, disp) \ -({ \ - unsigned int __disp = (disp) & 0xfff; \ - EMIT6(op1 | __disp, op2); \ +#define EMIT4_IMM(op, b1, imm) \ +({ \ + unsigned int __imm = (imm) & 0xffff; \ + _EMIT4(op | reg_high(b1) << 16 | __imm); \ + REG_SET_SEEN(b1); \ }) -#define EMIT6_IMM(op, imm) \ -({ \ - unsigned int __imm = (imm); \ - EMIT6(op | (__imm >> 16), __imm & 0xffff); \ +#define EMIT4_PCREL(op, pcrel) \ +({ \ + long __pcrel = ((pcrel) >> 1) & 0xffff; \ + _EMIT4(op | __pcrel); \ }) -#define EMIT_CONST(val) \ -({ \ - unsigned int ret; \ - ret = (unsigned int) (jit->lit - jit->base_ip); \ - jit->seen |= SEEN_LITERAL; \ - if (jit->lit + 4 <= jit->end) \ - *(u32 *) jit->lit = val; \ - jit->lit += 4; \ - ret; \ +#define _EMIT6(op1, op2) \ +({ \ + if (jit->prg_buf) { \ + *(u32 *) (jit->prg_buf + jit->prg) = op1; \ + *(u16 *) (jit->prg_buf + jit->prg + 4) = op2; \ + } \ + jit->prg += 6; \ }) -#define EMIT_FN_CONST(bit, fn) \ -({ \ - unsigned int ret; \ - ret = (unsigned int) (jit->lit - jit->base_ip); \ - if (jit->seen & bit) { \ - jit->seen |= SEEN_LITERAL; \ - if (jit->lit + 8 <= jit->end) \ - *(void **) jit->lit = fn; \ - jit->lit += 8; \ - } \ - ret; \ +#define _EMIT6_DISP(op1, op2, disp) \ +({ \ + unsigned int __disp = (disp) & 0xfff; \ + _EMIT6(op1 | __disp, op2); \ }) -static void bpf_jit_fill_hole(void *area, unsigned int size) +#define EMIT6_DISP(op1, op2, b1, b2, b3, disp) \ +({ \ + _EMIT6_DISP(op1 | reg(b1, b2) << 16 | \ + reg_high(b3) << 8, op2, disp); \ + REG_SET_SEEN(b1); \ + REG_SET_SEEN(b2); \ + REG_SET_SEEN(b3); \ +}) + +#define _EMIT6_DISP_LH(op1, op2, disp) \ +({ \ + unsigned int __disp_h = ((u32)disp) & 0xff000; \ + unsigned int __disp_l = ((u32)disp) & 0x00fff; \ + _EMIT6(op1 | __disp_l, op2 | __disp_h >> 4); \ +}) + +#define EMIT6_DISP_LH(op1, op2, b1, b2, b3, disp) \ +({ \ + _EMIT6_DISP_LH(op1 | reg(b1, b2) << 16 | \ + reg_high(b3) << 8, op2, disp); \ + REG_SET_SEEN(b1); \ + REG_SET_SEEN(b2); \ + REG_SET_SEEN(b3); \ +}) + +#define EMIT6_PCREL(op1, op2, b1, b2, i, off, mask) \ +({ \ + /* Branch instruction needs 6 bytes */ \ + int rel = (addrs[i + off + 1] - (addrs[i + 1] - 6)) / 2;\ + _EMIT6(op1 | reg(b1, b2) << 16 | rel, op2 | mask); \ + REG_SET_SEEN(b1); \ + REG_SET_SEEN(b2); \ +}) + +#define _EMIT6_IMM(op, imm) \ +({ \ + unsigned int __imm = (imm); \ + _EMIT6(op | (__imm >> 16), __imm & 0xffff); \ +}) + +#define EMIT6_IMM(op, b1, imm) \ +({ \ + _EMIT6_IMM(op | reg_high(b1) << 16, imm); \ + REG_SET_SEEN(b1); \ +}) + +#define EMIT_CONST_U32(val) \ +({ \ + unsigned int ret; \ + ret = jit->lit - jit->base_ip; \ + jit->seen |= SEEN_LITERAL; \ + if (jit->prg_buf) \ + *(u32 *) (jit->prg_buf + jit->lit) = (u32) val; \ + jit->lit += 4; \ + ret; \ +}) + +#define EMIT_CONST_U64(val) \ +({ \ + unsigned int ret; \ + ret = jit->lit - jit->base_ip; \ + jit->seen |= SEEN_LITERAL; \ + if (jit->prg_buf) \ + *(u64 *) (jit->prg_buf + jit->lit) = (u64) val; \ + jit->lit += 8; \ + ret; \ +}) + +#define EMIT_ZERO(b1) \ +({ \ + /* llgfr %dst,%dst (zero extend to 64 bit) */ \ + EMIT4(0xb9160000, b1, b1); \ + REG_SET_SEEN(b1); \ +}) + +/* + * Fill whole space with illegal instructions + */ +static void jit_fill_hole(void *area, unsigned int size) { - /* Fill whole space with illegal instructions */ memset(area, 0, size); } -static void bpf_jit_prologue(struct bpf_jit *jit) +/* + * Save registers from "rs" (register start) to "re" (register end) on stack + */ +static void save_regs(struct bpf_jit *jit, u32 rs, u32 re) +{ + u32 off = 72 + (rs - 6) * 8; + + if (rs == re) + /* stg %rs,off(%r15) */ + _EMIT6(0xe300f000 | rs << 20 | off, 0x0024); + else + /* stmg %rs,%re,off(%r15) */ + _EMIT6_DISP(0xeb00f000 | rs << 20 | re << 16, 0x0024, off); +} + +/* + * Restore registers from "rs" (register start) to "re" (register end) on stack + */ +static void restore_regs(struct bpf_jit *jit, u32 rs, u32 re) +{ + u32 off = 72 + (rs - 6) * 8; + + if (jit->seen & SEEN_STACK) + off += STK_OFF; + + if (rs == re) + /* lg %rs,off(%r15) */ + _EMIT6(0xe300f000 | rs << 20 | off, 0x0004); + else + /* lmg %rs,%re,off(%r15) */ + _EMIT6_DISP(0xeb00f000 | rs << 20 | re << 16, 0x0004, off); +} + +/* + * Return first seen register (from start) + */ +static int get_start(struct bpf_jit *jit, int start) { - /* Save registers and create stack frame if necessary */ - if (jit->seen & SEEN_DATAREF) { - /* stmg %r8,%r15,88(%r15) */ - EMIT6(0xeb8ff058, 0x0024); - /* lgr %r14,%r15 */ - EMIT4(0xb90400ef); - /* aghi %r15,<offset> */ - EMIT4_IMM(0xa7fb0000, (jit->seen & SEEN_MEM) ? -112 : -80); - /* stg %r14,152(%r15) */ - EMIT6(0xe3e0f098, 0x0024); - } else if ((jit->seen & SEEN_XREG) && (jit->seen & SEEN_LITERAL)) - /* stmg %r12,%r13,120(%r15) */ - EMIT6(0xebcdf078, 0x0024); - else if (jit->seen & SEEN_XREG) - /* stg %r12,120(%r15) */ - EMIT6(0xe3c0f078, 0x0024); - else if (jit->seen & SEEN_LITERAL) - /* stg %r13,128(%r15) */ - EMIT6(0xe3d0f080, 0x0024); + int i; + for (i = start; i <= 15; i++) { + if (jit->seen_reg[i]) + return i; + } + return 0; +} + +/* + * Return last seen register (from start) (gap >= 2) + */ +static int get_end(struct bpf_jit *jit, int start) +{ + int i; + + for (i = start; i < 15; i++) { + if (!jit->seen_reg[i] && !jit->seen_reg[i + 1]) + return i - 1; + } + return jit->seen_reg[15] ? 15 : 14; +} + +#define REGS_SAVE 1 +#define REGS_RESTORE 0 +/* + * Save and restore clobbered registers (6-15) on stack. + * We save/restore registers in chunks with gap >= 2 registers. + */ +static void save_restore_regs(struct bpf_jit *jit, int op) +{ + + int re = 6, rs; + + do { + rs = get_start(jit, re); + if (!rs) + break; + re = get_end(jit, rs + 1); + if (op == REGS_SAVE) + save_regs(jit, rs, re); + else + restore_regs(jit, rs, re); + re++; + } while (re <= 15); +} + +/* + * Emit function prologue + * + * Save registers and create stack frame if necessary. + * See stack frame layout desription in "bpf_jit.h"! + */ +static void bpf_jit_prologue(struct bpf_jit *jit) +{ + /* Save registers */ + save_restore_regs(jit, REGS_SAVE); /* Setup literal pool */ if (jit->seen & SEEN_LITERAL) { /* basr %r13,0 */ - EMIT2(0x0dd0); + EMIT2(0x0d00, REG_L, REG_0); jit->base_ip = jit->prg; } - jit->off_load_word = EMIT_FN_CONST(SEEN_LOAD_WORD, sk_load_word); - jit->off_load_half = EMIT_FN_CONST(SEEN_LOAD_HALF, sk_load_half); - jit->off_load_byte = EMIT_FN_CONST(SEEN_LOAD_BYTE, sk_load_byte); - jit->off_load_bmsh = EMIT_FN_CONST(SEEN_LOAD_BMSH, sk_load_byte_msh); - jit->off_load_iword = EMIT_FN_CONST(SEEN_LOAD_IWORD, sk_load_word_ind); - jit->off_load_ihalf = EMIT_FN_CONST(SEEN_LOAD_IHALF, sk_load_half_ind); - jit->off_load_ibyte = EMIT_FN_CONST(SEEN_LOAD_IBYTE, sk_load_byte_ind); - - /* Filter needs to access skb data */ - if (jit->seen & SEEN_DATAREF) { - /* l %r11,<len>(%r2) */ - EMIT4_DISP(0x58b02000, offsetof(struct sk_buff, len)); - /* s %r11,<data_len>(%r2) */ - EMIT4_DISP(0x5bb02000, offsetof(struct sk_buff, data_len)); - /* lg %r10,<data>(%r2) */ - EMIT6_DISP(0xe3a02000, 0x0004, - offsetof(struct sk_buff, data)); + /* Setup stack and backchain */ + if (jit->seen & SEEN_STACK) { + /* lgr %bfp,%r15 (BPF frame pointer) */ + EMIT4(0xb9040000, BPF_REG_FP, REG_15); + /* aghi %r15,-STK_OFF */ + EMIT4_IMM(0xa70b0000, REG_15, -STK_OFF); + if (jit->seen & SEEN_FUNC) + /* stg %bfp,152(%r15) (backchain) */ + EMIT6_DISP_LH(0xe3000000, 0x0024, BPF_REG_FP, REG_0, + REG_15, 152); } + /* + * For SKB access %b1 contains the SKB pointer. For "bpf_jit.S" + * we store the SKB header length on the stack and the SKB data + * pointer in REG_SKB_DATA. + */ + if (jit->seen & SEEN_SKB) { + /* Header length: llgf %w1,<len>(%b1) */ + EMIT6_DISP_LH(0xe3000000, 0x0016, REG_W1, REG_0, BPF_REG_1, + offsetof(struct sk_buff, len)); + /* s %w1,<data_len>(%b1) */ + EMIT4_DISP(0x5b000000, REG_W1, BPF_REG_1, + offsetof(struct sk_buff, data_len)); + /* stg %w1,ST_OFF_HLEN(%r0,%r15) */ + EMIT6_DISP_LH(0xe3000000, 0x0024, REG_W1, REG_0, REG_15, + STK_OFF_HLEN); + /* lg %skb_data,data_off(%b1) */ + EMIT6_DISP_LH(0xe3000000, 0x0004, REG_SKB_DATA, REG_0, + BPF_REG_1, offsetof(struct sk_buff, data)); + } + /* BPF compatibility: clear A (%b7) and X (%b8) registers */ + if (REG_SEEN(BPF_REG_7)) + /* lghi %b7,0 */ + EMIT4_IMM(0xa7090000, BPF_REG_7, 0); + if (REG_SEEN(BPF_REG_8)) + /* lghi %b8,0 */ + EMIT4_IMM(0xa7090000, BPF_REG_8, 0); } +/* + * Function epilogue + */ static void bpf_jit_epilogue(struct bpf_jit *jit) { /* Return 0 */ if (jit->seen & SEEN_RET0) { jit->ret0_ip = jit->prg; - /* lghi %r2,0 */ - EMIT4(0xa7290000); + /* lghi %b0,0 */ + EMIT4_IMM(0xa7090000, BPF_REG_0, 0); } jit->exit_ip = jit->prg; + /* Load exit code: lgr %r2,%b0 */ + EMIT4(0xb9040000, REG_2, BPF_REG_0); /* Restore registers */ - if (jit->seen & SEEN_DATAREF) - /* lmg %r8,%r15,<offset>(%r15) */ - EMIT6_DISP(0xeb8ff000, 0x0004, - (jit->seen & SEEN_MEM) ? 200 : 168); - else if ((jit->seen & SEEN_XREG) && (jit->seen & SEEN_LITERAL)) - /* lmg %r12,%r13,120(%r15) */ - EMIT6(0xebcdf078, 0x0004); - else if (jit->seen & SEEN_XREG) - /* lg %r12,120(%r15) */ - EMIT6(0xe3c0f078, 0x0004); - else if (jit->seen & SEEN_LITERAL) - /* lg %r13,128(%r15) */ - EMIT6(0xe3d0f080, 0x0004); + save_restore_regs(jit, REGS_RESTORE); /* br %r14 */ - EMIT2(0x07fe); + _EMIT2(0x07fe); } /* - * make sure we dont leak kernel information to user + * Compile one eBPF instruction into s390x code + * + * NOTE: Use noinline because for gcov (-fprofile-arcs) gcc allocates a lot of + * stack space for the large switch statement. */ -static void bpf_jit_noleaks(struct bpf_jit *jit, struct sock_filter *filter) +static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, int i) { - /* Clear temporary memory if (seen & SEEN_MEM) */ - if (jit->seen & SEEN_MEM) - /* xc 0(64,%r15),0(%r15) */ - EMIT6(0xd73ff000, 0xf000); - /* Clear X if (seen & SEEN_XREG) */ - if (jit->seen & SEEN_XREG) - /* lhi %r12,0 */ - EMIT4(0xa7c80000); - /* Clear A if the first register does not set it. */ - switch (filter[0].code) { - case BPF_LD | BPF_W | BPF_ABS: - case BPF_LD | BPF_H | BPF_ABS: - case BPF_LD | BPF_B | BPF_ABS: - case BPF_LD | BPF_W | BPF_LEN: - case BPF_LD | BPF_W | BPF_IND: - case BPF_LD | BPF_H | BPF_IND: - case BPF_LD | BPF_B | BPF_IND: - case BPF_LD | BPF_IMM: - case BPF_LD | BPF_MEM: - case BPF_MISC | BPF_TXA: - case BPF_RET | BPF_K: - /* first instruction sets A register */ - break; - default: /* A = 0 */ - /* lhi %r5,0 */ - EMIT4(0xa7580000); - } -} + struct bpf_insn *insn = &fp->insnsi[i]; + int jmp_off, last, insn_count = 1; + unsigned int func_addr, mask; + u32 dst_reg = insn->dst_reg; + u32 src_reg = insn->src_reg; + u32 *addrs = jit->addrs; + s32 imm = insn->imm; + s16 off = insn->off; -static int bpf_jit_insn(struct bpf_jit *jit, struct sock_filter *filter, - unsigned int *addrs, int i, int last) -{ - unsigned int K; - int offset; - unsigned int mask; - u16 code; - - K = filter->k; - code = bpf_anc_helper(filter); - - switch (code) { - case BPF_ALU | BPF_ADD | BPF_X: /* A += X */ - jit->seen |= SEEN_XREG; - /* ar %r5,%r12 */ - EMIT2(0x1a5c); - break; - case BPF_ALU | BPF_ADD | BPF_K: /* A += K */ - if (!K) + switch (insn->code) { + /* + * BPF_MOV + */ + case BPF_ALU | BPF_MOV | BPF_X: /* dst = (u32) src */ + /* llgfr %dst,%src */ + EMIT4(0xb9160000, dst_reg, src_reg); + break; + case BPF_ALU64 | BPF_MOV | BPF_X: /* dst = src */ + /* lgr %dst,%src */ + EMIT4(0xb9040000, dst_reg, src_reg); + break; + case BPF_ALU | BPF_MOV | BPF_K: /* dst = (u32) imm */ + /* llilf %dst,imm */ + EMIT6_IMM(0xc00f0000, dst_reg, imm); + break; + case BPF_ALU64 | BPF_MOV | BPF_K: /* dst = imm */ + /* lgfi %dst,imm */ + EMIT6_IMM(0xc0010000, dst_reg, imm); + break; + /* + * BPF_LD 64 + */ + case BPF_LD | BPF_IMM | BPF_DW: /* dst = (u64) imm */ + { + /* 16 byte instruction that uses two 'struct bpf_insn' */ + u64 imm64; + + imm64 = (u64)(u32) insn[0].imm | ((u64)(u32) insn[1].imm) << 32; + /* lg %dst,<d(imm)>(%l) */ + EMIT6_DISP_LH(0xe3000000, 0x0004, dst_reg, REG_0, REG_L, + EMIT_CONST_U64(imm64)); + insn_count = 2; + break; + } + /* + * BPF_ADD + */ + case BPF_ALU | BPF_ADD | BPF_X: /* dst = (u32) dst + (u32) src */ + /* ar %dst,%src */ + EMIT2(0x1a00, dst_reg, src_reg); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_ADD | BPF_X: /* dst = dst + src */ + /* agr %dst,%src */ + EMIT4(0xb9080000, dst_reg, src_reg); + break; + case BPF_ALU | BPF_ADD | BPF_K: /* dst = (u32) dst + (u32) imm */ + if (!imm) break; - if (K <= 16383) - /* ahi %r5,<K> */ - EMIT4_IMM(0xa75a0000, K); - else if (test_facility(21)) - /* alfi %r5,<K> */ - EMIT6_IMM(0xc25b0000, K); - else - /* a %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x5a50d000, EMIT_CONST(K)); + /* alfi %dst,imm */ + EMIT6_IMM(0xc20b0000, dst_reg, imm); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_ADD | BPF_K: /* dst = dst + imm */ + if (!imm) + break; + /* agfi %dst,imm */ + EMIT6_IMM(0xc2080000, dst_reg, imm); + break; + /* + * BPF_SUB + */ + case BPF_ALU | BPF_SUB | BPF_X: /* dst = (u32) dst - (u32) src */ + /* sr %dst,%src */ + EMIT2(0x1b00, dst_reg, src_reg); + EMIT_ZERO(dst_reg); break; - case BPF_ALU | BPF_SUB | BPF_X: /* A -= X */ - jit->seen |= SEEN_XREG; - /* sr %r5,%r12 */ - EMIT2(0x1b5c); + case BPF_ALU64 | BPF_SUB | BPF_X: /* dst = dst - src */ + /* sgr %dst,%src */ + EMIT4(0xb9090000, dst_reg, src_reg); break; - case BPF_ALU | BPF_SUB | BPF_K: /* A -= K */ - if (!K) + case BPF_ALU | BPF_SUB | BPF_K: /* dst = (u32) dst - (u32) imm */ + if (!imm) break; - if (K <= 16384) - /* ahi %r5,-K */ - EMIT4_IMM(0xa75a0000, -K); - else if (test_facility(21)) - /* alfi %r5,-K */ - EMIT6_IMM(0xc25b0000, -K); - else - /* s %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x5b50d000, EMIT_CONST(K)); - break; - case BPF_ALU | BPF_MUL | BPF_X: /* A *= X */ - jit->seen |= SEEN_XREG; - /* msr %r5,%r12 */ - EMIT4(0xb252005c); - break; - case BPF_ALU | BPF_MUL | BPF_K: /* A *= K */ - if (K <= 16383) - /* mhi %r5,K */ - EMIT4_IMM(0xa75c0000, K); - else if (test_facility(34)) - /* msfi %r5,<K> */ - EMIT6_IMM(0xc2510000, K); - else - /* ms %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x7150d000, EMIT_CONST(K)); + /* alfi %dst,-imm */ + EMIT6_IMM(0xc20b0000, dst_reg, -imm); + EMIT_ZERO(dst_reg); break; - case BPF_ALU | BPF_DIV | BPF_X: /* A /= X */ - jit->seen |= SEEN_XREG | SEEN_RET0; - /* ltr %r12,%r12 */ - EMIT2(0x12cc); - /* jz <ret0> */ - EMIT4_PCREL(0xa7840000, (jit->ret0_ip - jit->prg)); - /* lhi %r4,0 */ - EMIT4(0xa7480000); - /* dlr %r4,%r12 */ - EMIT4(0xb997004c); - break; - case BPF_ALU | BPF_DIV | BPF_K: /* A /= K */ - if (K == 1) + case BPF_ALU64 | BPF_SUB | BPF_K: /* dst = dst - imm */ + if (!imm) + break; + /* agfi %dst,-imm */ + EMIT6_IMM(0xc2080000, dst_reg, -imm); + break; + /* + * BPF_MUL + */ + case BPF_ALU | BPF_MUL | BPF_X: /* dst = (u32) dst * (u32) src */ + /* msr %dst,%src */ + EMIT4(0xb2520000, dst_reg, src_reg); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_MUL | BPF_X: /* dst = dst * src */ + /* msgr %dst,%src */ + EMIT4(0xb90c0000, dst_reg, src_reg); + break; + case BPF_ALU | BPF_MUL | BPF_K: /* dst = (u32) dst * (u32) imm */ + if (imm == 1) + break; + /* msfi %r5,imm */ + EMIT6_IMM(0xc2010000, dst_reg, imm); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_MUL | BPF_K: /* dst = dst * imm */ + if (imm == 1) break; - /* lhi %r4,0 */ - EMIT4(0xa7480000); - /* dl %r4,<d(K)>(%r13) */ - EMIT6_DISP(0xe340d000, 0x0097, EMIT_CONST(K)); - break; - case BPF_ALU | BPF_MOD | BPF_X: /* A %= X */ - jit->seen |= SEEN_XREG | SEEN_RET0; - /* ltr %r12,%r12 */ - EMIT2(0x12cc); + /* msgfi %dst,imm */ + EMIT6_IMM(0xc2000000, dst_reg, imm); + break; + /* + * BPF_DIV / BPF_MOD + */ + case BPF_ALU | BPF_DIV | BPF_X: /* dst = (u32) dst / (u32) src */ + case BPF_ALU | BPF_MOD | BPF_X: /* dst = (u32) dst % (u32) src */ + { + int rc_reg = BPF_OP(insn->code) == BPF_DIV ? REG_W1 : REG_W0; + + jit->seen |= SEEN_RET0; + /* ltr %src,%src (if src == 0 goto fail) */ + EMIT2(0x1200, src_reg, src_reg); /* jz <ret0> */ - EMIT4_PCREL(0xa7840000, (jit->ret0_ip - jit->prg)); - /* lhi %r4,0 */ - EMIT4(0xa7480000); - /* dlr %r4,%r12 */ - EMIT4(0xb997004c); - /* lr %r5,%r4 */ - EMIT2(0x1854); - break; - case BPF_ALU | BPF_MOD | BPF_K: /* A %= K */ - if (K == 1) { - /* lhi %r5,0 */ - EMIT4(0xa7580000); + EMIT4_PCREL(0xa7840000, jit->ret0_ip - jit->prg); + /* lhi %w0,0 */ + EMIT4_IMM(0xa7080000, REG_W0, 0); + /* lr %w1,%dst */ + EMIT2(0x1800, REG_W1, dst_reg); + /* dlr %w0,%src */ + EMIT4(0xb9970000, REG_W0, src_reg); + /* llgfr %dst,%rc */ + EMIT4(0xb9160000, dst_reg, rc_reg); + break; + } + case BPF_ALU64 | BPF_DIV | BPF_X: /* dst = dst / src */ + case BPF_ALU64 | BPF_MOD | BPF_X: /* dst = dst % src */ + { + int rc_reg = BPF_OP(insn->code) == BPF_DIV ? REG_W1 : REG_W0; + + jit->seen |= SEEN_RET0; + /* ltgr %src,%src (if src == 0 goto fail) */ + EMIT4(0xb9020000, src_reg, src_reg); + /* jz <ret0> */ + EMIT4_PCREL(0xa7840000, jit->ret0_ip - jit->prg); + /* lghi %w0,0 */ + EMIT4_IMM(0xa7090000, REG_W0, 0); + /* lgr %w1,%dst */ + EMIT4(0xb9040000, REG_W1, dst_reg); + /* dlgr %w0,%dst */ + EMIT4(0xb9870000, REG_W0, src_reg); + /* lgr %dst,%rc */ + EMIT4(0xb9040000, dst_reg, rc_reg); + break; + } + case BPF_ALU | BPF_DIV | BPF_K: /* dst = (u32) dst / (u32) imm */ + case BPF_ALU | BPF_MOD | BPF_K: /* dst = (u32) dst % (u32) imm */ + { + int rc_reg = BPF_OP(insn->code) == BPF_DIV ? REG_W1 : REG_W0; + + if (imm == 1) { + if (BPF_OP(insn->code) == BPF_MOD) + /* lhgi %dst,0 */ + EMIT4_IMM(0xa7090000, dst_reg, 0); break; } - /* lhi %r4,0 */ - EMIT4(0xa7480000); - /* dl %r4,<d(K)>(%r13) */ - EMIT6_DISP(0xe340d000, 0x0097, EMIT_CONST(K)); - /* lr %r5,%r4 */ - EMIT2(0x1854); - break; - case BPF_ALU | BPF_AND | BPF_X: /* A &= X */ - jit->seen |= SEEN_XREG; - /* nr %r5,%r12 */ - EMIT2(0x145c); - break; - case BPF_ALU | BPF_AND | BPF_K: /* A &= K */ - if (test_facility(21)) - /* nilf %r5,<K> */ - EMIT6_IMM(0xc05b0000, K); - else - /* n %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x5450d000, EMIT_CONST(K)); - break; - case BPF_ALU | BPF_OR | BPF_X: /* A |= X */ - jit->seen |= SEEN_XREG; - /* or %r5,%r12 */ - EMIT2(0x165c); - break; - case BPF_ALU | BPF_OR | BPF_K: /* A |= K */ - if (test_facility(21)) - /* oilf %r5,<K> */ - EMIT6_IMM(0xc05d0000, K); - else - /* o %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x5650d000, EMIT_CONST(K)); + /* lhi %w0,0 */ + EMIT4_IMM(0xa7080000, REG_W0, 0); + /* lr %w1,%dst */ + EMIT2(0x1800, REG_W1, dst_reg); + /* dl %w0,<d(imm)>(%l) */ + EMIT6_DISP_LH(0xe3000000, 0x0097, REG_W0, REG_0, REG_L, + EMIT_CONST_U32(imm)); + /* llgfr %dst,%rc */ + EMIT4(0xb9160000, dst_reg, rc_reg); + break; + } + case BPF_ALU64 | BPF_DIV | BPF_K: /* dst = dst / imm */ + case BPF_ALU64 | BPF_MOD | BPF_K: /* dst = dst % imm */ + { + int rc_reg = BPF_OP(insn->code) == BPF_DIV ? REG_W1 : REG_W0; + + if (imm == 1) { + if (BPF_OP(insn->code) == BPF_MOD) + /* lhgi %dst,0 */ + EMIT4_IMM(0xa7090000, dst_reg, 0); + break; + } + /* lghi %w0,0 */ + EMIT4_IMM(0xa7090000, REG_W0, 0); + /* lgr %w1,%dst */ + EMIT4(0xb9040000, REG_W1, dst_reg); + /* dlg %w0,<d(imm)>(%l) */ + EMIT6_DISP_LH(0xe3000000, 0x0087, REG_W0, REG_0, REG_L, + EMIT_CONST_U64(imm)); + /* lgr %dst,%rc */ + EMIT4(0xb9040000, dst_reg, rc_reg); + break; + } + /* + * BPF_AND + */ + case BPF_ALU | BPF_AND | BPF_X: /* dst = (u32) dst & (u32) src */ + /* nr %dst,%src */ + EMIT2(0x1400, dst_reg, src_reg); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_AND | BPF_X: /* dst = dst & src */ + /* ngr %dst,%src */ + EMIT4(0xb9800000, dst_reg, src_reg); + break; + case BPF_ALU | BPF_AND | BPF_K: /* dst = (u32) dst & (u32) imm */ + /* nilf %dst,imm */ + EMIT6_IMM(0xc00b0000, dst_reg, imm); + EMIT_ZERO(dst_reg); break; - case BPF_ANC | SKF_AD_ALU_XOR_X: /* A ^= X; */ - case BPF_ALU | BPF_XOR | BPF_X: - jit->seen |= SEEN_XREG; - /* xr %r5,%r12 */ - EMIT2(0x175c); + case BPF_ALU64 | BPF_AND | BPF_K: /* dst = dst & imm */ + /* ng %dst,<d(imm)>(%l) */ + EMIT6_DISP_LH(0xe3000000, 0x0080, dst_reg, REG_0, REG_L, + EMIT_CONST_U64(imm)); break; - case BPF_ALU | BPF_XOR | BPF_K: /* A ^= K */ - if (!K) + /* + * BPF_OR + */ + case BPF_ALU | BPF_OR | BPF_X: /* dst = (u32) dst | (u32) src */ + /* or %dst,%src */ + EMIT2(0x1600, dst_reg, src_reg); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_OR | BPF_X: /* dst = dst | src */ + /* ogr %dst,%src */ + EMIT4(0xb9810000, dst_reg, src_reg); + break; + case BPF_ALU | BPF_OR | BPF_K: /* dst = (u32) dst | (u32) imm */ + /* oilf %dst,imm */ + EMIT6_IMM(0xc00d0000, dst_reg, imm); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_OR | BPF_K: /* dst = dst | imm */ + /* og %dst,<d(imm)>(%l) */ + EMIT6_DISP_LH(0xe3000000, 0x0081, dst_reg, REG_0, REG_L, + EMIT_CONST_U64(imm)); + break; + /* + * BPF_XOR + */ + case BPF_ALU | BPF_XOR | BPF_X: /* dst = (u32) dst ^ (u32) src */ + /* xr %dst,%src */ + EMIT2(0x1700, dst_reg, src_reg); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_XOR | BPF_X: /* dst = dst ^ src */ + /* xgr %dst,%src */ + EMIT4(0xb9820000, dst_reg, src_reg); + break; + case BPF_ALU | BPF_XOR | BPF_K: /* dst = (u32) dst ^ (u32) imm */ + if (!imm) break; - /* x %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x5750d000, EMIT_CONST(K)); + /* xilf %dst,imm */ + EMIT6_IMM(0xc0070000, dst_reg, imm); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_XOR | BPF_K: /* dst = dst ^ imm */ + /* xg %dst,<d(imm)>(%l) */ + EMIT6_DISP_LH(0xe3000000, 0x0082, dst_reg, REG_0, REG_L, + EMIT_CONST_U64(imm)); + break; + /* + * BPF_LSH + */ + case BPF_ALU | BPF_LSH | BPF_X: /* dst = (u32) dst << (u32) src */ + /* sll %dst,0(%src) */ + EMIT4_DISP(0x89000000, dst_reg, src_reg, 0); + EMIT_ZERO(dst_reg); break; - case BPF_ALU | BPF_LSH | BPF_X: /* A <<= X; */ - jit->seen |= SEEN_XREG; - /* sll %r5,0(%r12) */ - EMIT4(0x8950c000); + case BPF_ALU64 | BPF_LSH | BPF_X: /* dst = dst << src */ + /* sllg %dst,%dst,0(%src) */ + EMIT6_DISP_LH(0xeb000000, 0x000d, dst_reg, dst_reg, src_reg, 0); break; - case BPF_ALU | BPF_LSH | BPF_K: /* A <<= K */ - if (K == 0) + case BPF_ALU | BPF_LSH | BPF_K: /* dst = (u32) dst << (u32) imm */ + if (imm == 0) break; - /* sll %r5,K */ - EMIT4_DISP(0x89500000, K); + /* sll %dst,imm(%r0) */ + EMIT4_DISP(0x89000000, dst_reg, REG_0, imm); + EMIT_ZERO(dst_reg); break; - case BPF_ALU | BPF_RSH | BPF_X: /* A >>= X; */ - jit->seen |= SEEN_XREG; - /* srl %r5,0(%r12) */ - EMIT4(0x8850c000); + case BPF_ALU64 | BPF_LSH | BPF_K: /* dst = dst << imm */ + if (imm == 0) + break; + /* sllg %dst,%dst,imm(%r0) */ + EMIT6_DISP_LH(0xeb000000, 0x000d, dst_reg, dst_reg, REG_0, imm); + break; + /* + * BPF_RSH + */ + case BPF_ALU | BPF_RSH | BPF_X: /* dst = (u32) dst >> (u32) src */ + /* srl %dst,0(%src) */ + EMIT4_DISP(0x88000000, dst_reg, src_reg, 0); + EMIT_ZERO(dst_reg); break; - case BPF_ALU | BPF_RSH | BPF_K: /* A >>= K; */ - if (K == 0) + case BPF_ALU64 | BPF_RSH | BPF_X: /* dst = dst >> src */ + /* srlg %dst,%dst,0(%src) */ + EMIT6_DISP_LH(0xeb000000, 0x000c, dst_reg, dst_reg, src_reg, 0); + break; + case BPF_ALU | BPF_RSH | BPF_K: /* dst = (u32) dst >> (u32) imm */ + if (imm == 0) break; - /* srl %r5,K */ - EMIT4_DISP(0x88500000, K); - break; - case BPF_ALU | BPF_NEG: /* A = -A */ - /* lcr %r5,%r5 */ - EMIT2(0x1355); - break; - case BPF_JMP | BPF_JA: /* ip += K */ - offset = addrs[i + K] + jit->start - jit->prg; - EMIT4_PCREL(0xa7f40000, offset); - break; - case BPF_JMP | BPF_JGT | BPF_K: /* ip += (A > K) ? jt : jf */ - mask = 0x200000; /* jh */ - goto kbranch; - case BPF_JMP | BPF_JGE | BPF_K: /* ip += (A >= K) ? jt : jf */ - mask = 0xa00000; /* jhe */ - goto kbranch; - case BPF_JMP | BPF_JEQ | BPF_K: /* ip += (A == K) ? jt : jf */ - mask = 0x800000; /* je */ -kbranch: /* Emit compare if the branch targets are different */ - if (filter->jt != filter->jf) { - if (test_facility(21)) - /* clfi %r5,<K> */ - EMIT6_IMM(0xc25f0000, K); - else - /* cl %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x5550d000, EMIT_CONST(K)); - } -branch: if (filter->jt == filter->jf) { - if (filter->jt == 0) - break; - /* j <jt> */ - offset = addrs[i + filter->jt] + jit->start - jit->prg; - EMIT4_PCREL(0xa7f40000, offset); + /* srl %dst,imm(%r0) */ + EMIT4_DISP(0x88000000, dst_reg, REG_0, imm); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_RSH | BPF_K: /* dst = dst >> imm */ + if (imm == 0) break; - } - if (filter->jt != 0) { - /* brc <mask>,<jt> */ - offset = addrs[i + filter->jt] + jit->start - jit->prg; - EMIT4_PCREL(0xa7040000 | mask, offset); - } - if (filter->jf != 0) { - /* brc <mask^15>,<jf> */ - offset = addrs[i + filter->jf] + jit->start - jit->prg; - EMIT4_PCREL(0xa7040000 | (mask ^ 0xf00000), offset); - } + /* srlg %dst,%dst,imm(%r0) */ + EMIT6_DISP_LH(0xeb000000, 0x000c, dst_reg, dst_reg, REG_0, imm); break; - case BPF_JMP | BPF_JSET | BPF_K: /* ip += (A & K) ? jt : jf */ - mask = 0x700000; /* jnz */ - /* Emit test if the branch targets are different */ - if (filter->jt != filter->jf) { - if (K > 65535) { - /* lr %r4,%r5 */ - EMIT2(0x1845); - /* n %r4,<d(K)>(%r13) */ - EMIT4_DISP(0x5440d000, EMIT_CONST(K)); - } else - /* tmll %r5,K */ - EMIT4_IMM(0xa7510000, K); - } - goto branch; - case BPF_JMP | BPF_JGT | BPF_X: /* ip += (A > X) ? jt : jf */ - mask = 0x200000; /* jh */ - goto xbranch; - case BPF_JMP | BPF_JGE | BPF_X: /* ip += (A >= X) ? jt : jf */ - mask = 0xa00000; /* jhe */ - goto xbranch; - case BPF_JMP | BPF_JEQ | BPF_X: /* ip += (A == X) ? jt : jf */ - mask = 0x800000; /* je */ -xbranch: /* Emit compare if the branch targets are different */ - if (filter->jt != filter->jf) { - jit->seen |= SEEN_XREG; - /* clr %r5,%r12 */ - EMIT2(0x155c); - } - goto branch; - case BPF_JMP | BPF_JSET | BPF_X: /* ip += (A & X) ? jt : jf */ - mask = 0x700000; /* jnz */ - /* Emit test if the branch targets are different */ - if (filter->jt != filter->jf) { - jit->seen |= SEEN_XREG; - /* lr %r4,%r5 */ - EMIT2(0x1845); - /* nr %r4,%r12 */ - EMIT2(0x144c); + /* + * BPF_ARSH + */ + case BPF_ALU64 | BPF_ARSH | BPF_X: /* ((s64) dst) >>= src */ + /* srag %dst,%dst,0(%src) */ + EMIT6_DISP_LH(0xeb000000, 0x000a, dst_reg, dst_reg, src_reg, 0); + break; + case BPF_ALU64 | BPF_ARSH | BPF_K: /* ((s64) dst) >>= imm */ + if (imm == 0) + break; + /* srag %dst,%dst,imm(%r0) */ + EMIT6_DISP_LH(0xeb000000, 0x000a, dst_reg, dst_reg, REG_0, imm); + break; + /* + * BPF_NEG + */ + case BPF_ALU | BPF_NEG: /* dst = (u32) -dst */ + /* lcr %dst,%dst */ + EMIT2(0x1300, dst_reg, dst_reg); + EMIT_ZERO(dst_reg); + break; + case BPF_ALU64 | BPF_NEG: /* dst = -dst */ + /* lcgr %dst,%dst */ + EMIT4(0xb9130000, dst_reg, dst_reg); + break; + /* + * BPF_FROM_BE/LE + */ + case BPF_ALU | BPF_END | BPF_FROM_BE: + /* s390 is big endian, therefore only clear high order bytes */ + switch (imm) { + case 16: /* dst = (u16) cpu_to_be16(dst) */ + /* llghr %dst,%dst */ + EMIT4(0xb9850000, dst_reg, dst_reg); + break; + case 32: /* dst = (u32) cpu_to_be32(dst) */ + /* llgfr %dst,%dst */ + EMIT4(0xb9160000, dst_reg, dst_reg); + break; + case 64: /* dst = (u64) cpu_to_be64(dst) */ + break; } - goto branch; - case BPF_LD | BPF_W | BPF_ABS: /* A = *(u32 *) (skb->data+K) */ - jit->seen |= SEEN_DATAREF | SEEN_RET0 | SEEN_LOAD_WORD; - offset = jit->off_load_word; - goto load_abs; - case BPF_LD | BPF_H | BPF_ABS: /* A = *(u16 *) (skb->data+K) */ - jit->seen |= SEEN_DATAREF | SEEN_RET0 | SEEN_LOAD_HALF; - offset = jit->off_load_half; - goto load_abs; - case BPF_LD | BPF_B | BPF_ABS: /* A = *(u8 *) (skb->data+K) */ - jit->seen |= SEEN_DATAREF | SEEN_RET0 | SEEN_LOAD_BYTE; - offset = jit->off_load_byte; -load_abs: if ((int) K < 0) - goto out; -call_fn: /* lg %r1,<d(function)>(%r13) */ - EMIT6_DISP(0xe310d000, 0x0004, offset); - /* l %r3,<d(K)>(%r13) */ - EMIT4_DISP(0x5830d000, EMIT_CONST(K)); - /* basr %r8,%r1 */ - EMIT2(0x0d81); - /* jnz <ret0> */ - EMIT4_PCREL(0xa7740000, (jit->ret0_ip - jit->prg)); break; - case BPF_LD | BPF_W | BPF_IND: /* A = *(u32 *) (skb->data+K+X) */ - jit->seen |= SEEN_DATAREF | SEEN_RET0 | SEEN_LOAD_IWORD; - offset = jit->off_load_iword; - goto call_fn; - case BPF_LD | BPF_H | BPF_IND: /* A = *(u16 *) (skb->data+K+X) */ - jit->seen |= SEEN_DATAREF | SEEN_RET0 | SEEN_LOAD_IHALF; - offset = jit->off_load_ihalf; - goto call_fn; - case BPF_LD | BPF_B | BPF_IND: /* A = *(u8 *) (skb->data+K+X) */ - jit->seen |= SEEN_DATAREF | SEEN_RET0 | SEEN_LOAD_IBYTE; - offset = jit->off_load_ibyte; - goto call_fn; - case BPF_LDX | BPF_B | BPF_MSH: - /* X = (*(u8 *)(skb->data+K) & 0xf) << 2 */ - jit->seen |= SEEN_RET0; - if ((int) K < 0) { - /* j <ret0> */ - EMIT4_PCREL(0xa7f40000, (jit->ret0_ip - jit->prg)); + case BPF_ALU | BPF_END | BPF_FROM_LE: + switch (imm) { + case 16: /* dst = (u16) cpu_to_le16(dst) */ + /* lrvr %dst,%dst */ + EMIT4(0xb91f0000, dst_reg, dst_reg); + /* srl %dst,16(%r0) */ + EMIT4_DISP(0x88000000, dst_reg, REG_0, 16); + /* llghr %dst,%dst */ + EMIT4(0xb9850000, dst_reg, dst_reg); + break; + case 32: /* dst = (u32) cpu_to_le32(dst) */ + /* lrvr %dst,%dst */ + EMIT4(0xb91f0000, dst_reg, dst_reg); + /* llgfr %dst,%dst */ + EMIT4(0xb9160000, dst_reg, dst_reg); + break; + case 64: /* dst = (u64) cpu_to_le64(dst) */ + /* lrvgr %dst,%dst */ + EMIT4(0xb90f0000, dst_reg, dst_reg); break; } - jit->seen |= SEEN_DATAREF | SEEN_LOAD_BMSH; - offset = jit->off_load_bmsh; - goto call_fn; - case BPF_LD | BPF_W | BPF_LEN: /* A = skb->len; */ - BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, len) != 4); - /* l %r5,<d(len)>(%r2) */ - EMIT4_DISP(0x58502000, offsetof(struct sk_buff, len)); - break; - case BPF_LDX | BPF_W | BPF_LEN: /* X = skb->len; */ - jit->seen |= SEEN_XREG; - /* l %r12,<d(len)>(%r2) */ - EMIT4_DISP(0x58c02000, offsetof(struct sk_buff, len)); - break; - case BPF_LD | BPF_IMM: /* A = K */ - if (K <= 16383) - /* lhi %r5,K */ - EMIT4_IMM(0xa7580000, K); - else if (test_facility(21)) - /* llilf %r5,<K> */ - EMIT6_IMM(0xc05f0000, K); - else - /* l %r5,<d(K)>(%r13) */ - EMIT4_DISP(0x5850d000, EMIT_CONST(K)); - break; - case BPF_LDX | BPF_IMM: /* X = K */ - jit->seen |= SEEN_XREG; - if (K <= 16383) - /* lhi %r12,<K> */ - EMIT4_IMM(0xa7c80000, K); - else if (test_facility(21)) - /* llilf %r12,<K> */ - EMIT6_IMM(0xc0cf0000, K); - else - /* l %r12,<d(K)>(%r13) */ - EMIT4_DISP(0x58c0d000, EMIT_CONST(K)); break; - case BPF_LD | BPF_MEM: /* A = mem[K] */ + /* + * BPF_ST(X) + */ + case BPF_STX | BPF_MEM | BPF_B: /* *(u8 *)(dst + off) = src_reg */ + /* stcy %src,off(%dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0072, src_reg, dst_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_STX | BPF_MEM | BPF_H: /* (u16 *)(dst + off) = src */ + /* sthy %src,off(%dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0070, src_reg, dst_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_STX | BPF_MEM | BPF_W: /* *(u32 *)(dst + off) = src */ + /* sty %src,off(%dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0050, src_reg, dst_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_STX | BPF_MEM | BPF_DW: /* (u64 *)(dst + off) = src */ + /* stg %src,off(%dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0024, src_reg, dst_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_ST | BPF_MEM | BPF_B: /* *(u8 *)(dst + off) = imm */ + /* lhi %w0,imm */ + EMIT4_IMM(0xa7080000, REG_W0, (u8) imm); + /* stcy %w0,off(dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0072, REG_W0, dst_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_ST | BPF_MEM | BPF_H: /* (u16 *)(dst + off) = imm */ + /* lhi %w0,imm */ + EMIT4_IMM(0xa7080000, REG_W0, (u16) imm); + /* sthy %w0,off(dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0070, REG_W0, dst_reg, REG_0, off); jit->seen |= SEEN_MEM; - /* l %r5,<K>(%r15) */ - EMIT4_DISP(0x5850f000, - (jit->seen & SEEN_DATAREF) ? 160 + K*4 : K*4); - break; - case BPF_LDX | BPF_MEM: /* X = mem[K] */ - jit->seen |= SEEN_XREG | SEEN_MEM; - /* l %r12,<K>(%r15) */ - EMIT4_DISP(0x58c0f000, - (jit->seen & SEEN_DATAREF) ? 160 + K*4 : K*4); - break; - case BPF_MISC | BPF_TAX: /* X = A */ - jit->seen |= SEEN_XREG; - /* lr %r12,%r5 */ - EMIT2(0x18c5); - break; - case BPF_MISC | BPF_TXA: /* A = X */ - jit->seen |= SEEN_XREG; - /* lr %r5,%r12 */ - EMIT2(0x185c); - break; - case BPF_RET | BPF_K: - if (K == 0) { - jit->seen |= SEEN_RET0; - if (last) - break; - /* j <ret0> */ - EMIT4_PCREL(0xa7f40000, jit->ret0_ip - jit->prg); - } else { - if (K <= 16383) - /* lghi %r2,K */ - EMIT4_IMM(0xa7290000, K); - else - /* llgf %r2,<K>(%r13) */ - EMIT6_DISP(0xe320d000, 0x0016, EMIT_CONST(K)); - /* j <exit> */ - if (last && !(jit->seen & SEEN_RET0)) - break; - EMIT4_PCREL(0xa7f40000, jit->exit_ip - jit->prg); - } break; - case BPF_RET | BPF_A: - /* llgfr %r2,%r5 */ - EMIT4(0xb9160025); + case BPF_ST | BPF_MEM | BPF_W: /* *(u32 *)(dst + off) = imm */ + /* llilf %w0,imm */ + EMIT6_IMM(0xc00f0000, REG_W0, (u32) imm); + /* sty %w0,off(%dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0050, REG_W0, dst_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_ST | BPF_MEM | BPF_DW: /* *(u64 *)(dst + off) = imm */ + /* lgfi %w0,imm */ + EMIT6_IMM(0xc0010000, REG_W0, imm); + /* stg %w0,off(%dst) */ + EMIT6_DISP_LH(0xe3000000, 0x0024, REG_W0, dst_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + /* + * BPF_STX XADD (atomic_add) + */ + case BPF_STX | BPF_XADD | BPF_W: /* *(u32 *)(dst + off) += src */ + /* laal %w0,%src,off(%dst) */ + EMIT6_DISP_LH(0xeb000000, 0x00fa, REG_W0, src_reg, + dst_reg, off); + jit->seen |= SEEN_MEM; + break; + case BPF_STX | BPF_XADD | BPF_DW: /* *(u64 *)(dst + off) += src */ + /* laalg %w0,%src,off(%dst) */ + EMIT6_DISP_LH(0xeb000000, 0x00ea, REG_W0, src_reg, + dst_reg, off); + jit->seen |= SEEN_MEM; + break; + /* + * BPF_LDX + */ + case BPF_LDX | BPF_MEM | BPF_B: /* dst = *(u8 *)(ul) (src + off) */ + /* llgc %dst,0(off,%src) */ + EMIT6_DISP_LH(0xe3000000, 0x0090, dst_reg, src_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_LDX | BPF_MEM | BPF_H: /* dst = *(u16 *)(ul) (src + off) */ + /* llgh %dst,0(off,%src) */ + EMIT6_DISP_LH(0xe3000000, 0x0091, dst_reg, src_reg, REG_0, off); + jit->seen |= SEEN_MEM; + break; + case BPF_LDX | BPF_MEM | BPF_W: /* dst = *(u32 *)(ul) (src + off) */ + /* llgf %dst,off(%src) */ + jit->seen |= SEEN_MEM; + EMIT6_DISP_LH(0xe3000000, 0x0016, dst_reg, src_reg, REG_0, off); + break; + case BPF_LDX | BPF_MEM | BPF_DW: /* dst = *(u64 *)(ul) (src + off) */ + /* lg %dst,0(off,%src) */ + jit->seen |= SEEN_MEM; + EMIT6_DISP_LH(0xe3000000, 0x0004, dst_reg, src_reg, REG_0, off); + break; + /* + * BPF_JMP / CALL + */ + case BPF_JMP | BPF_CALL: + { + /* + * b0 = (__bpf_call_base + imm)(b1, b2, b3, b4, b5) + */ + const u64 func = (u64)__bpf_call_base + imm; + + REG_SET_SEEN(BPF_REG_5); + jit->seen |= SEEN_FUNC; + /* lg %w1,<d(imm)>(%l) */ + EMIT6_DISP(0xe3000000, 0x0004, REG_W1, REG_0, REG_L, + EMIT_CONST_U64(func)); + /* basr %r14,%w1 */ + EMIT2(0x0d00, REG_14, REG_W1); + /* lgr %b0,%r2: load return value into %b0 */ + EMIT4(0xb9040000, BPF_REG_0, REG_2); + break; + } + case BPF_JMP | BPF_EXIT: /* return b0 */ + last = (i == fp->len - 1) ? 1 : 0; + if (last && !(jit->seen & SEEN_RET0)) + break; /* j <exit> */ EMIT4_PCREL(0xa7f40000, jit->exit_ip - jit->prg); break; - case BPF_ST: /* mem[K] = A */ - jit->seen |= SEEN_MEM; - /* st %r5,<K>(%r15) */ - EMIT4_DISP(0x5050f000, - (jit->seen & SEEN_DATAREF) ? 160 + K*4 : K*4); - break; - case BPF_STX: /* mem[K] = X : mov %ebx,off8(%rbp) */ - jit->seen |= SEEN_XREG | SEEN_MEM; - /* st %r12,<K>(%r15) */ - EMIT4_DISP(0x50c0f000, - (jit->seen & SEEN_DATAREF) ? 160 + K*4 : K*4); - break; - case BPF_ANC | SKF_AD_PROTOCOL: /* A = ntohs(skb->protocol); */ - BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, protocol) != 2); - /* lhi %r5,0 */ - EMIT4(0xa7580000); - /* icm %r5,3,<d(protocol)>(%r2) */ - EMIT4_DISP(0xbf532000, offsetof(struct sk_buff, protocol)); - break; - case BPF_ANC | SKF_AD_IFINDEX: /* if (!skb->dev) return 0; - * A = skb->dev->ifindex */ - BUILD_BUG_ON(FIELD_SIZEOF(struct net_device, ifindex) != 4); - jit->seen |= SEEN_RET0; - /* lg %r1,<d(dev)>(%r2) */ - EMIT6_DISP(0xe3102000, 0x0004, offsetof(struct sk_buff, dev)); - /* ltgr %r1,%r1 */ - EMIT4(0xb9020011); - /* jz <ret0> */ - EMIT4_PCREL(0xa7840000, jit->ret0_ip - jit->prg); - /* l %r5,<d(ifindex)>(%r1) */ - EMIT4_DISP(0x58501000, offsetof(struct net_device, ifindex)); - break; - case BPF_ANC | SKF_AD_MARK: /* A = skb->mark */ - BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, mark) != 4); - /* l %r5,<d(mark)>(%r2) */ - EMIT4_DISP(0x58502000, offsetof(struct sk_buff, mark)); - break; - case BPF_ANC | SKF_AD_QUEUE: /* A = skb->queue_mapping */ - BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, queue_mapping) != 2); - /* lhi %r5,0 */ - EMIT4(0xa7580000); - /* icm %r5,3,<d(queue_mapping)>(%r2) */ - EMIT4_DISP(0xbf532000, offsetof(struct sk_buff, queue_mapping)); - break; - case BPF_ANC | SKF_AD_HATYPE: /* if (!skb->dev) return 0; - * A = skb->dev->type */ - BUILD_BUG_ON(FIELD_SIZEOF(struct net_device, type) != 2); - jit->seen |= SEEN_RET0; - /* lg %r1,<d(dev)>(%r2) */ - EMIT6_DISP(0xe3102000, 0x0004, offsetof(struct sk_buff, dev)); - /* ltgr %r1,%r1 */ - EMIT4(0xb9020011); - /* jz <ret0> */ - EMIT4_PCREL(0xa7840000, jit->ret0_ip - jit->prg); - /* lhi %r5,0 */ - EMIT4(0xa7580000); - /* icm %r5,3,<d(type)>(%r1) */ - EMIT4_DISP(0xbf531000, offsetof(struct net_device, type)); - break; - case BPF_ANC | SKF_AD_RXHASH: /* A = skb->hash */ - BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, hash) != 4); - /* l %r5,<d(hash)>(%r2) */ - EMIT4_DISP(0x58502000, offsetof(struct sk_buff, hash)); - break; - case BPF_ANC | SKF_AD_VLAN_TAG: - case BPF_ANC | SKF_AD_VLAN_TAG_PRESENT: - BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, vlan_tci) != 2); - BUILD_BUG_ON(VLAN_TAG_PRESENT != 0x1000); - /* lhi %r5,0 */ - EMIT4(0xa7580000); - /* icm %r5,3,<d(vlan_tci)>(%r2) */ - EMIT4_DISP(0xbf532000, offsetof(struct sk_buff, vlan_tci)); - if (code == (BPF_ANC | SKF_AD_VLAN_TAG)) { - /* nill %r5,0xefff */ - EMIT4_IMM(0xa5570000, ~VLAN_TAG_PRESENT); - } else { - /* nill %r5,0x1000 */ - EMIT4_IMM(0xa5570000, VLAN_TAG_PRESENT); - /* srl %r5,12 */ - EMIT4_DISP(0x88500000, 12); - } + /* + * Branch relative (number of skipped instructions) to offset on + * condition. + * + * Condition code to mask mapping: + * + * CC | Description | Mask + * ------------------------------ + * 0 | Operands equal | 8 + * 1 | First operand low | 4 + * 2 | First operand high | 2 + * 3 | Unused | 1 + * + * For s390x relative branches: ip = ip + off_bytes + * For BPF relative branches: insn = insn + off_insns + 1 + * + * For example for s390x with offset 0 we jump to the branch + * instruction itself (loop) and for BPF with offset 0 we + * branch to the instruction behind the branch. + */ + case BPF_JMP | BPF_JA: /* if (true) */ + mask = 0xf000; /* j */ + goto branch_oc; + case BPF_JMP | BPF_JSGT | BPF_K: /* ((s64) dst > (s64) imm) */ + mask = 0x2000; /* jh */ + goto branch_ks; + case BPF_JMP | BPF_JSGE | BPF_K: /* ((s64) dst >= (s64) imm) */ + mask = 0xa000; /* jhe */ + goto branch_ks; + case BPF_JMP | BPF_JGT | BPF_K: /* (dst_reg > imm) */ + mask = 0x2000; /* jh */ + goto branch_ku; + case BPF_JMP | BPF_JGE | BPF_K: /* (dst_reg >= imm) */ + mask = 0xa000; /* jhe */ + goto branch_ku; + case BPF_JMP | BPF_JNE | BPF_K: /* (dst_reg != imm) */ + mask = 0x7000; /* jne */ + goto branch_ku; + case BPF_JMP | BPF_JEQ | BPF_K: /* (dst_reg == imm) */ + mask = 0x8000; /* je */ + goto branch_ku; + case BPF_JMP | BPF_JSET | BPF_K: /* (dst_reg & imm) */ + mask = 0x7000; /* jnz */ + /* lgfi %w1,imm (load sign extend imm) */ + EMIT6_IMM(0xc0010000, REG_W1, imm); + /* ngr %w1,%dst */ + EMIT4(0xb9800000, REG_W1, dst_reg); + goto branch_oc; + + case BPF_JMP | BPF_JSGT | BPF_X: /* ((s64) dst > (s64) src) */ + mask = 0x2000; /* jh */ + goto branch_xs; + case BPF_JMP | BPF_JSGE | BPF_X: /* ((s64) dst >= (s64) src) */ + mask = 0xa000; /* jhe */ + goto branch_xs; + case BPF_JMP | BPF_JGT | BPF_X: /* (dst > src) */ + mask = 0x2000; /* jh */ + goto branch_xu; + case BPF_JMP | BPF_JGE | BPF_X: /* (dst >= src) */ + mask = 0xa000; /* jhe */ + goto branch_xu; + case BPF_JMP | BPF_JNE | BPF_X: /* (dst != src) */ + mask = 0x7000; /* jne */ + goto branch_xu; + case BPF_JMP | BPF_JEQ | BPF_X: /* (dst == src) */ + mask = 0x8000; /* je */ + goto branch_xu; + case BPF_JMP | BPF_JSET | BPF_X: /* (dst & src) */ + mask = 0x7000; /* jnz */ + /* ngrk %w1,%dst,%src */ + EMIT4_RRF(0xb9e40000, REG_W1, dst_reg, src_reg); + goto branch_oc; +branch_ks: + /* lgfi %w1,imm (load sign extend imm) */ + EMIT6_IMM(0xc0010000, REG_W1, imm); + /* cgrj %dst,%w1,mask,off */ + EMIT6_PCREL(0xec000000, 0x0064, dst_reg, REG_W1, i, off, mask); + break; +branch_ku: + /* lgfi %w1,imm (load sign extend imm) */ + EMIT6_IMM(0xc0010000, REG_W1, imm); + /* clgrj %dst,%w1,mask,off */ + EMIT6_PCREL(0xec000000, 0x0065, dst_reg, REG_W1, i, off, mask); + break; +branch_xs: + /* cgrj %dst,%src,mask,off */ + EMIT6_PCREL(0xec000000, 0x0064, dst_reg, src_reg, i, off, mask); + break; +branch_xu: + /* clgrj %dst,%src,mask,off */ + EMIT6_PCREL(0xec000000, 0x0065, dst_reg, src_reg, i, off, mask); + break; +branch_oc: + /* brc mask,jmp_off (branch instruction needs 4 bytes) */ + jmp_off = addrs[i + off + 1] - (addrs[i + 1] - 4); + EMIT4_PCREL(0xa7040000 | mask << 8, jmp_off); break; - case BPF_ANC | SKF_AD_PKTTYPE: - /* lhi %r5,0 */ - EMIT4(0xa7580000); - /* ic %r5,<d(pkt_type_offset)>(%r2) */ - EMIT4_DISP(0x43502000, PKT_TYPE_OFFSET()); - /* srl %r5,5 */ - EMIT4_DISP(0x88500000, 5); - break; - case BPF_ANC | SKF_AD_CPU: /* A = smp_processor_id() */ -#ifdef CONFIG_SMP - /* l %r5,<d(cpu_nr)> */ - EMIT4_DISP(0x58500000, offsetof(struct _lowcore, cpu_nr)); -#else - /* lhi %r5,0 */ - EMIT4(0xa7580000); -#endif + /* + * BPF_LD + */ + case BPF_LD | BPF_ABS | BPF_B: /* b0 = *(u8 *) (skb->data+imm) */ + case BPF_LD | BPF_IND | BPF_B: /* b0 = *(u8 *) (skb->data+imm+src) */ + if ((BPF_MODE(insn->code) == BPF_ABS) && (imm >= 0)) + func_addr = __pa(sk_load_byte_pos); + else + func_addr = __pa(sk_load_byte); + goto call_fn; + case BPF_LD | BPF_ABS | BPF_H: /* b0 = *(u16 *) (skb->data+imm) */ + case BPF_LD | BPF_IND | BPF_H: /* b0 = *(u16 *) (skb->data+imm+src) */ + if ((BPF_MODE(insn->code) == BPF_ABS) && (imm >= 0)) + func_addr = __pa(sk_load_half_pos); + else + func_addr = __pa(sk_load_half); + goto call_fn; + case BPF_LD | BPF_ABS | BPF_W: /* b0 = *(u32 *) (skb->data+imm) */ + case BPF_LD | BPF_IND | BPF_W: /* b0 = *(u32 *) (skb->data+imm+src) */ + if ((BPF_MODE(insn->code) == BPF_ABS) && (imm >= 0)) + func_addr = __pa(sk_load_word_pos); + else + func_addr = __pa(sk_load_word); + goto call_fn; +call_fn: + jit->seen |= SEEN_SKB | SEEN_RET0 | SEEN_FUNC; + REG_SET_SEEN(REG_14); /* Return address of possible func call */ + + /* + * Implicit input: + * BPF_REG_6 (R7) : skb pointer + * REG_SKB_DATA (R12): skb data pointer + * + * Calculated input: + * BPF_REG_2 (R3) : offset of byte(s) to fetch in skb + * BPF_REG_5 (R6) : return address + * + * Output: + * BPF_REG_0 (R14): data read from skb + * + * Scratch registers (BPF_REG_1-5) + */ + + /* Call function: llilf %w1,func_addr */ + EMIT6_IMM(0xc00f0000, REG_W1, func_addr); + + /* Offset: lgfi %b2,imm */ + EMIT6_IMM(0xc0010000, BPF_REG_2, imm); + if (BPF_MODE(insn->code) == BPF_IND) + /* agfr %b2,%src (%src is s32 here) */ + EMIT4(0xb9180000, BPF_REG_2, src_reg); + + /* basr %b5,%w1 (%b5 is call saved) */ + EMIT2(0x0d00, BPF_REG_5, REG_W1); + + /* + * Note: For fast access we jump directly after the + * jnz instruction from bpf_jit.S + */ + /* jnz <ret0> */ + EMIT4_PCREL(0xa7740000, jit->ret0_ip - jit->prg); break; default: /* too complex, give up */ - goto out; + pr_err("Unknown opcode %02x\n", insn->code); + return -1; + } + return insn_count; +} + +/* + * Compile eBPF program into s390x code + */ +static int bpf_jit_prog(struct bpf_jit *jit, struct bpf_prog *fp) +{ + int i, insn_count; + + jit->lit = jit->lit_start; + jit->prg = 0; + + bpf_jit_prologue(jit); + for (i = 0; i < fp->len; i += insn_count) { + insn_count = bpf_jit_insn(jit, fp, i); + if (insn_count < 0) + return -1; + jit->addrs[i + 1] = jit->prg; /* Next instruction address */ } - addrs[i] = jit->prg - jit->start; + bpf_jit_epilogue(jit); + + jit->lit_start = jit->prg; + jit->size = jit->lit; + jit->size_prg = jit->prg; return 0; -out: - return -1; } +/* + * Classic BPF function stub. BPF programs will be converted into + * eBPF and then bpf_int_jit_compile() will be called. + */ void bpf_jit_compile(struct bpf_prog *fp) { - struct bpf_binary_header *header = NULL; - unsigned long size, prg_len, lit_len; - struct bpf_jit jit, cjit; - unsigned int *addrs; - int pass, i; +} + +/* + * Compile eBPF program "fp" + */ +void bpf_int_jit_compile(struct bpf_prog *fp) +{ + struct bpf_binary_header *header; + struct bpf_jit jit; + int pass; if (!bpf_jit_enable) return; - addrs = kcalloc(fp->len, sizeof(*addrs), GFP_KERNEL); - if (addrs == NULL) + memset(&jit, 0, sizeof(jit)); + jit.addrs = kcalloc(fp->len + 1, sizeof(*jit.addrs), GFP_KERNEL); + if (jit.addrs == NULL) return; - memset(&jit, 0, sizeof(cjit)); - memset(&cjit, 0, sizeof(cjit)); - - for (pass = 0; pass < 10; pass++) { - jit.prg = jit.start; - jit.lit = jit.mid; - - bpf_jit_prologue(&jit); - bpf_jit_noleaks(&jit, fp->insns); - for (i = 0; i < fp->len; i++) { - if (bpf_jit_insn(&jit, fp->insns + i, addrs, i, - i == fp->len - 1)) - goto out; - } - bpf_jit_epilogue(&jit); - if (jit.start) { - WARN_ON(jit.prg > cjit.prg || jit.lit > cjit.lit); - if (memcmp(&jit, &cjit, sizeof(jit)) == 0) - break; - } else if (jit.prg == cjit.prg && jit.lit == cjit.lit) { - prg_len = jit.prg - jit.start; - lit_len = jit.lit - jit.mid; - size = prg_len + lit_len; - if (size >= BPF_SIZE_MAX) - goto out; - header = bpf_jit_binary_alloc(size, &jit.start, - 2, bpf_jit_fill_hole); - if (!header) - goto out; - jit.prg = jit.mid = jit.start + prg_len; - jit.lit = jit.end = jit.start + prg_len + lit_len; - jit.base_ip += (unsigned long) jit.start; - jit.exit_ip += (unsigned long) jit.start; - jit.ret0_ip += (unsigned long) jit.start; - } - cjit = jit; + /* + * Three initial passes: + * - 1/2: Determine clobbered registers + * - 3: Calculate program size and addrs arrray + */ + for (pass = 1; pass <= 3; pass++) { + if (bpf_jit_prog(&jit, fp)) + goto free_addrs; } + /* + * Final pass: Allocate and generate program + */ + if (jit.size >= BPF_SIZE_MAX) + goto free_addrs; + header = bpf_jit_binary_alloc(jit.size, &jit.prg_buf, 2, jit_fill_hole); + if (!header) + goto free_addrs; + if (bpf_jit_prog(&jit, fp)) + goto free_addrs; if (bpf_jit_enable > 1) { - bpf_jit_dump(fp->len, jit.end - jit.start, pass, jit.start); - if (jit.start) - print_fn_code(jit.start, jit.mid - jit.start); + bpf_jit_dump(fp->len, jit.size, pass, jit.prg_buf); + if (jit.prg_buf) + print_fn_code(jit.prg_buf, jit.size_prg); } - if (jit.start) { + if (jit.prg_buf) { set_memory_ro((unsigned long)header, header->pages); - fp->bpf_func = (void *) jit.start; + fp->bpf_func = (void *) jit.prg_buf; fp->jited = true; } -out: - kfree(addrs); +free_addrs: + kfree(jit.addrs); } +/* + * Free eBPF program + */ void bpf_jit_free(struct bpf_prog *fp) { unsigned long addr = (unsigned long)fp->bpf_func & PAGE_MASK; diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c index 98336200c7b2..598f023cf8a6 100644 --- a/arch/s390/pci/pci.c +++ b/arch/s390/pci/pci.c @@ -190,6 +190,11 @@ int zpci_fmb_enable_device(struct zpci_dev *zdev) return -ENOMEM; WARN_ON((u64) zdev->fmb & 0xf); + /* reset software counters */ + atomic64_set(&zdev->allocated_pages, 0); + atomic64_set(&zdev->mapped_pages, 0); + atomic64_set(&zdev->unmapped_pages, 0); + args.fmb_addr = virt_to_phys(zdev->fmb); return mod_pci(zdev, ZPCI_MOD_FC_SET_MEASURE, 0, &args); } @@ -822,6 +827,7 @@ int zpci_create_device(struct zpci_dev *zdev) if (rc) goto out; + mutex_init(&zdev->lock); if (zdev->state == ZPCI_FN_STATE_CONFIGURED) { rc = zpci_enable_device(zdev); if (rc) diff --git a/arch/s390/pci/pci_debug.c b/arch/s390/pci/pci_debug.c index c22d4402ae45..4129b0a5fd78 100644 --- a/arch/s390/pci/pci_debug.c +++ b/arch/s390/pci/pci_debug.c @@ -31,12 +31,25 @@ static char *pci_perf_names[] = { "Refresh operations", "DMA read bytes", "DMA write bytes", - /* software counters */ +}; + +static char *pci_sw_names[] = { "Allocated pages", "Mapped pages", "Unmapped pages", }; +static void pci_sw_counter_show(struct seq_file *m) +{ + struct zpci_dev *zdev = m->private; + atomic64_t *counter = &zdev->allocated_pages; + int i; + + for (i = 0; i < ARRAY_SIZE(pci_sw_names); i++, counter++) + seq_printf(m, "%26s:\t%llu\n", pci_sw_names[i], + atomic64_read(counter)); +} + static int pci_perf_show(struct seq_file *m, void *v) { struct zpci_dev *zdev = m->private; @@ -45,7 +58,10 @@ static int pci_perf_show(struct seq_file *m, void *v) if (!zdev) return 0; + + mutex_lock(&zdev->lock); if (!zdev->fmb) { + mutex_unlock(&zdev->lock); seq_puts(m, "FMB statistics disabled\n"); return 0; } @@ -65,12 +81,9 @@ static int pci_perf_show(struct seq_file *m, void *v) for (i = 4; i < 6; i++) seq_printf(m, "%26s:\t%llu\n", pci_perf_names[i], *(stat + i)); - /* software counters */ - for (i = 6; i < ARRAY_SIZE(pci_perf_names); i++) - seq_printf(m, "%26s:\t%llu\n", - pci_perf_names[i], - atomic64_read((atomic64_t *) (stat + i))); + pci_sw_counter_show(m); + mutex_unlock(&zdev->lock); return 0; } @@ -88,19 +101,17 @@ static ssize_t pci_perf_seq_write(struct file *file, const char __user *ubuf, if (rc) return rc; + mutex_lock(&zdev->lock); switch (val) { case 0: rc = zpci_fmb_disable_device(zdev); - if (rc) - return rc; break; case 1: rc = zpci_fmb_enable_device(zdev); - if (rc) - return rc; break; } - return count; + mutex_unlock(&zdev->lock); + return rc ? rc : count; } static int pci_perf_seq_open(struct inode *inode, struct file *filp) diff --git a/arch/s390/pci/pci_dma.c b/arch/s390/pci/pci_dma.c index 4cbb29a4d615..6fd8d5836138 100644 --- a/arch/s390/pci/pci_dma.c +++ b/arch/s390/pci/pci_dma.c @@ -300,7 +300,7 @@ static dma_addr_t s390_dma_map_pages(struct device *dev, struct page *page, flags |= ZPCI_TABLE_PROTECTED; if (!dma_update_trans(zdev, pa, dma_addr, size, flags)) { - atomic64_add(nr_pages, &zdev->fmb->mapped_pages); + atomic64_add(nr_pages, &zdev->mapped_pages); return dma_addr + (offset & ~PAGE_MASK); } @@ -328,7 +328,7 @@ static void s390_dma_unmap_pages(struct device *dev, dma_addr_t dma_addr, zpci_err_hex(&dma_addr, sizeof(dma_addr)); } - atomic64_add(npages, &zdev->fmb->unmapped_pages); + atomic64_add(npages, &zdev->unmapped_pages); iommu_page_index = (dma_addr - zdev->start_dma) >> PAGE_SHIFT; dma_free_iommu(zdev, iommu_page_index, npages); } @@ -357,7 +357,7 @@ static void *s390_dma_alloc(struct device *dev, size_t size, return NULL; } - atomic64_add(size / PAGE_SIZE, &zdev->fmb->allocated_pages); + atomic64_add(size / PAGE_SIZE, &zdev->allocated_pages); if (dma_handle) *dma_handle = map; return (void *) pa; @@ -370,7 +370,7 @@ static void s390_dma_free(struct device *dev, size_t size, struct zpci_dev *zdev = get_zdev(to_pci_dev(dev)); size = PAGE_ALIGN(size); - atomic64_sub(size / PAGE_SIZE, &zdev->fmb->allocated_pages); + atomic64_sub(size / PAGE_SIZE, &zdev->allocated_pages); s390_dma_unmap_pages(dev, dma_handle, size, DMA_BIDIRECTIONAL, NULL); free_pages((unsigned long) pa, get_order(size)); } |