diff options
author | Steven Whitehouse <swhiteho@redhat.com> | 2006-04-03 09:08:57 -0400 |
---|---|---|
committer | Steven Whitehouse <swhiteho@redhat.com> | 2006-04-03 09:08:57 -0400 |
commit | 76467874b83835129dc454e3a7a8e5d1186101b0 (patch) | |
tree | 162129f0c36c35be4aa323cf00626db0e804c3fc /fs | |
parent | 8628de0583504138551a05ad44ca388467f0f552 (diff) | |
parent | 6246b6128bbe34d0752f119cf7c5111c85fe481d (diff) |
Merge branch 'master'
Diffstat (limited to 'fs')
66 files changed, 1305 insertions, 710 deletions
diff --git a/fs/Makefile b/fs/Makefile index 414484ac48c..2c22e282c77 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -10,7 +10,7 @@ obj-y := open.o read_write.o file_table.o buffer.o bio.o super.o \ ioctl.o readdir.o select.o fifo.o locks.o dcache.o inode.o \ attr.o bad_inode.o file.o filesystems.o namespace.o aio.o \ seq_file.o xattr.o libfs.o fs-writeback.o mpage.o direct-io.o \ - ioprio.o pnode.o drop_caches.o splice.o + ioprio.o pnode.o drop_caches.o splice.o sync.o obj-$(CONFIG_INOTIFY) += inotify.o obj-$(CONFIG_EPOLL) += eventpoll.o diff --git a/fs/char_dev.c b/fs/char_dev.c index 4e1b849f912..f3418f7a6e9 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -15,6 +15,7 @@ #include <linux/module.h> #include <linux/smp_lock.h> #include <linux/devfs_fs_kernel.h> +#include <linux/seq_file.h> #include <linux/kobject.h> #include <linux/kobj_map.h> @@ -27,8 +28,6 @@ static struct kobj_map *cdev_map; -#define MAX_PROBE_HASH 255 /* random */ - static DEFINE_MUTEX(chrdevs_lock); static struct char_device_struct { @@ -39,93 +38,29 @@ static struct char_device_struct { char name[64]; struct file_operations *fops; struct cdev *cdev; /* will die */ -} *chrdevs[MAX_PROBE_HASH]; +} *chrdevs[CHRDEV_MAJOR_HASH_SIZE]; /* index in the above */ static inline int major_to_index(int major) { - return major % MAX_PROBE_HASH; -} - -struct chrdev_info { - int index; - struct char_device_struct *cd; -}; - -void *get_next_chrdev(void *dev) -{ - struct chrdev_info *info; - - if (dev == NULL) { - info = kmalloc(sizeof(*info), GFP_KERNEL); - if (!info) - goto out; - info->index=0; - info->cd = chrdevs[info->index]; - if (info->cd) - goto out; - } else { - info = dev; - } - - while (info->index < ARRAY_SIZE(chrdevs)) { - if (info->cd) - info->cd = info->cd->next; - if (info->cd) - goto out; - /* - * No devices on this chain, move to the next - */ - info->index++; - info->cd = (info->index < ARRAY_SIZE(chrdevs)) ? - chrdevs[info->index] : NULL; - if (info->cd) - goto out; - } - -out: - return info; -} - -void *acquire_chrdev_list(void) -{ - mutex_lock(&chrdevs_lock); - return get_next_chrdev(NULL); -} - -void release_chrdev_list(void *dev) -{ - mutex_unlock(&chrdevs_lock); - kfree(dev); + return major % CHRDEV_MAJOR_HASH_SIZE; } +#ifdef CONFIG_PROC_FS -int count_chrdev_list(void) +void chrdev_show(struct seq_file *f, off_t offset) { struct char_device_struct *cd; - int i, count; - - count = 0; - for (i = 0; i < ARRAY_SIZE(chrdevs) ; i++) { - for (cd = chrdevs[i]; cd; cd = cd->next) - count++; + if (offset < CHRDEV_MAJOR_HASH_SIZE) { + mutex_lock(&chrdevs_lock); + for (cd = chrdevs[offset]; cd; cd = cd->next) + seq_printf(f, "%3d %s\n", cd->major, cd->name); + mutex_unlock(&chrdevs_lock); } - - return count; } -int get_chrdev_info(void *dev, int *major, char **name) -{ - struct chrdev_info *info = dev; - - if (info->cd == NULL) - return 1; - - *major = info->cd->major; - *name = info->cd->name; - return 0; -} +#endif /* CONFIG_PROC_FS */ /* * Register a single major with a specified minor range. diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index cb68efba35d..8a2de038882 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -1,3 +1,21 @@ +Version 1.42 +------------ +Fix slow oplock break when mounted to different servers at the same time and +the tids match and we try to find matching fid on wrong server. + +Version 1.41 +------------ +Fix NTLMv2 security (can be enabled in /proc/fs/cifs) so customers can +configure stronger authentication. Fix sfu symlinks so they can +be followed (not just recognized). Fix wraparound of bcc on +read responses when buffer size over 64K and also fix wrap of +max smb buffer size when CIFSMaxBufSize over 64K. Fix oops in +cifs_user_read and cifs_readpages (when EAGAIN on send of smb +on socket is returned over and over). Add POSIX (advisory) byte range +locking support (requires server with newest CIFS UNIX Extensions +to the protocol implemented). Slow down negprot slightly in port 139 +RFC1001 case to give session_init time on buggy servers. + Version 1.40 ------------ Use fsuid (fsgid) more consistently instead of uid (gid). Improve performance diff --git a/fs/cifs/Makefile b/fs/cifs/Makefile index 7384947a0f9..58c77254a23 100644 --- a/fs/cifs/Makefile +++ b/fs/cifs/Makefile @@ -3,4 +3,4 @@ # obj-$(CONFIG_CIFS) += cifs.o -cifs-objs := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o link.o misc.o netmisc.o smbdes.o smbencrypt.o transport.o asn1.o md4.o md5.o cifs_unicode.o nterr.o xattr.o cifsencrypt.o fcntl.o readdir.o ioctl.o +cifs-objs := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o link.o misc.o netmisc.o smbdes.o smbencrypt.o transport.o asn1.o md4.o md5.o cifs_unicode.o nterr.o xattr.o cifsencrypt.o fcntl.o readdir.o ioctl.o ntlmssp.o diff --git a/fs/cifs/README b/fs/cifs/README index b0070d1b149..b2b4d080376 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -422,6 +422,13 @@ A partial list of the supported mount options follows: nomapchars Do not translate any of these seven characters (default). nocase Request case insensitive path name matching (case sensitive is the default if the server suports it). + posixpaths If CIFS Unix extensions are supported, attempt to + negotiate posix path name support which allows certain + characters forbidden in typical CIFS filenames, without + requiring remapping. (default) + noposixpaths If CIFS Unix extensions are supported, do not request + posix path name support (this may cause servers to + reject creatingfile with certain reserved characters). nobrl Do not send byte range lock requests to the server. This is necessary for certain applications that break with cifs style mandatory byte range locks (and most diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index a2c24858d40..e7d63737e65 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -1,7 +1,7 @@ /* * fs/cifs/cifsencrypt.c * - * Copyright (C) International Business Machines Corp., 2005 + * Copyright (C) International Business Machines Corp., 2005,2006 * Author(s): Steve French (sfrench@us.ibm.com) * * This library is free software; you can redistribute it and/or modify @@ -36,7 +36,8 @@ extern void mdfour(unsigned char *out, unsigned char *in, int n); extern void E_md4hash(const unsigned char *passwd, unsigned char *p16); -static int cifs_calculate_signature(const struct smb_hdr * cifs_pdu, const char * key, char * signature) +static int cifs_calculate_signature(const struct smb_hdr * cifs_pdu, + const char * key, char * signature) { struct MD5Context context; @@ -56,9 +57,6 @@ int cifs_sign_smb(struct smb_hdr * cifs_pdu, struct TCP_Server_Info * server, int rc = 0; char smb_signature[20]; - /* BB remember to initialize sequence number elsewhere and initialize mac_signing key elsewhere BB */ - /* BB remember to add code to save expected sequence number in midQ entry BB */ - if((cifs_pdu == NULL) || (server == NULL)) return -EINVAL; @@ -85,20 +83,33 @@ int cifs_sign_smb(struct smb_hdr * cifs_pdu, struct TCP_Server_Info * server, static int cifs_calc_signature2(const struct kvec * iov, int n_vec, const char * key, char * signature) { - struct MD5Context context; - - if((iov == NULL) || (signature == NULL)) - return -EINVAL; + struct MD5Context context; + int i; - MD5Init(&context); - MD5Update(&context,key,CIFS_SESSION_KEY_SIZE+16); + if((iov == NULL) || (signature == NULL)) + return -EINVAL; -/* MD5Update(&context,cifs_pdu->Protocol,cifs_pdu->smb_buf_length); */ /* BB FIXME BB */ + MD5Init(&context); + MD5Update(&context,key,CIFS_SESSION_KEY_SIZE+16); + for(i=0;i<n_vec;i++) { + if(iov[i].iov_base == NULL) { + cERROR(1,("null iovec entry")); + return -EIO; + } else if(iov[i].iov_len == 0) + break; /* bail out if we are sent nothing to sign */ + /* The first entry includes a length field (which does not get + signed that occupies the first 4 bytes before the header */ + if(i==0) { + if (iov[0].iov_len <= 8 ) /* cmd field at offset 9 */ + break; /* nothing to sign or corrupt header */ + MD5Update(&context,iov[0].iov_base+4, iov[0].iov_len-4); + } else + MD5Update(&context,iov[i].iov_base, iov[i].iov_len); + } - MD5Final(signature,&context); + MD5Final(signature,&context); - return -EOPNOTSUPP; -/* return 0; */ + return 0; } @@ -259,4 +270,5 @@ void CalcNTLMv2_response(const struct cifsSesInfo * ses,char * v2_session_respon /* hmac_md5_update(v2_session_response+16)client thing,8,&context); */ /* BB fix */ hmac_md5_final(v2_session_response,&context); + cifs_dump_mem("v2_sess_rsp: ", v2_session_response, 32); /* BB removeme BB */ } diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 4bbc544857b..d4b713e5aff 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -93,13 +93,10 @@ cifs_read_super(struct super_block *sb, void *data, int rc = 0; sb->s_flags |= MS_NODIRATIME; /* and probably even noatime */ - sb->s_fs_info = kmalloc(sizeof(struct cifs_sb_info),GFP_KERNEL); + sb->s_fs_info = kzalloc(sizeof(struct cifs_sb_info),GFP_KERNEL); cifs_sb = CIFS_SB(sb); if(cifs_sb == NULL) return -ENOMEM; - else - memset(cifs_sb,0,sizeof(struct cifs_sb_info)); - rc = cifs_mount(sb, cifs_sb, data, devname); diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index 74f405ae4da..4e829dc672a 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -99,5 +99,5 @@ extern ssize_t cifs_getxattr(struct dentry *, const char *, void *, size_t); extern ssize_t cifs_listxattr(struct dentry *, char *, size_t); extern int cifs_ioctl (struct inode * inode, struct file * filep, unsigned int command, unsigned long arg); -#define CIFS_VERSION "1.40" +#define CIFS_VERSION "1.42" #endif /* _CIFSFS_H */ diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 7bed27601ce..006eb33bff5 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -1,7 +1,7 @@ /* * fs/cifs/cifsglob.h * - * Copyright (C) International Business Machines Corp., 2002,2005 + * Copyright (C) International Business Machines Corp., 2002,2006 * Author(s): Steve French (sfrench@us.ibm.com) * * This library is free software; you can redistribute it and/or modify @@ -430,6 +430,15 @@ struct dir_notify_req { #define CIFS_LARGE_BUFFER 2 #define CIFS_IOVEC 4 /* array of response buffers */ +/* Type of session setup needed */ +#define CIFS_PLAINTEXT 0 +#define CIFS_LANMAN 1 +#define CIFS_NTLM 2 +#define CIFS_NTLMSSP_NEG 3 +#define CIFS_NTLMSSP_AUTH 4 +#define CIFS_SPNEGO_INIT 5 +#define CIFS_SPNEGO_TARG 6 + /* ***************************************************************** * All constants go here diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index cc2471094ca..b2233ac05bd 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -859,7 +859,10 @@ typedef struct smb_com_lock_req { LOCKING_ANDX_RANGE Locks[1]; } __attribute__((packed)) LOCK_REQ; - +/* lock type */ +#define CIFS_RDLCK 0 +#define CIFS_WRLCK 1 +#define CIFS_UNLCK 2 typedef struct cifs_posix_lock { __le16 lock_type; /* 0 = Read, 1 = Write, 2 = Unlock */ __le16 lock_flags; /* 1 = Wait (only valid for setlock) */ @@ -1786,7 +1789,13 @@ typedef struct { #define CIFS_UNIX_POSIX_ACL_CAP 0x00000002 /* support getfacl/setfacl */ #define CIFS_UNIX_XATTR_CAP 0x00000004 /* support new namespace */ #define CIFS_UNIX_EXTATTR_CAP 0x00000008 /* support chattr/chflag */ -#define CIFS_UNIX_POSIX_PATHNAMES_CAP 0x00000010 /* Use POSIX pathnames on the wire. */ +#define CIFS_UNIX_POSIX_PATHNAMES_CAP 0x00000010 /* Allow POSIX path chars */ +#ifdef CONFIG_CIFS_POSIX +#define CIFS_UNIX_CAP_MASK 0x0000001b +#else +#define CIFS_UNIX_CAP_MASK 0x00000013 +#endif /* CONFIG_CIFS_POSIX */ + #define CIFS_POSIX_EXTENSIONS 0x00000010 /* support for new QFSInfo */ diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 7b25463d3c1..2879ba343ca 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -1,7 +1,7 @@ /* * fs/cifs/cifsproto.h * - * Copyright (c) International Business Machines Corp., 2002,2005 + * Copyright (c) International Business Machines Corp., 2002,2006 * Author(s): Steve French (sfrench@us.ibm.com) * * This library is free software; you can redistribute it and/or modify @@ -64,6 +64,14 @@ extern int map_smb_to_linux_error(struct smb_hdr *smb); extern void header_assemble(struct smb_hdr *, char /* command */ , const struct cifsTconInfo *, int /* length of fixed section (word count) in two byte units */); +#ifdef CONFIG_CIFS_EXPERIMENTAL +extern int small_smb_init_no_tc(const int smb_cmd, const int wct, + struct cifsSesInfo *ses, + void ** request_buf); +extern int CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, + const int stage, int * pNTLMv2_flg, + const struct nls_table *nls_cp); +#endif extern __u16 GetNextMid(struct TCP_Server_Info *server); extern struct oplock_q_entry * AllocOplockQEntry(struct inode *, u16, struct cifsTconInfo *); @@ -257,7 +265,10 @@ extern int CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, const __u64 offset, const __u32 numUnlock, const __u32 numLock, const __u8 lockType, const int waitFlag); - +extern int CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, + const __u16 smb_file_id, const int get_flag, + const __u64 len, const __u64 offset, + const __u16 lock_type, const int waitFlag); extern int CIFSSMBTDis(const int xid, struct cifsTconInfo *tcon); extern int CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses); diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index a243fe2792d..d705500aa28 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1,7 +1,7 @@ /* * fs/cifs/cifssmb.c * - * Copyright (C) International Business Machines Corp., 2002,2005 + * Copyright (C) International Business Machines Corp., 2002,2006 * Author(s): Steve French (sfrench@us.ibm.com) * * Contains the routines for constructing the SMB PDUs themselves @@ -186,7 +186,35 @@ small_smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, cifs_stats_inc(&tcon->num_smbs_sent); return rc; -} +} + +#ifdef CONFIG_CIFS_EXPERIMENTAL +int +small_smb_init_no_tc(const int smb_command, const int wct, + struct cifsSesInfo *ses, void **request_buf) +{ + int rc; + struct smb_hdr * buffer; + + rc = small_smb_init(smb_command, wct, NULL, request_buf); + if(rc) + return rc; + + buffer = (struct smb_hdr *)*request_buf; + buffer->Mid = GetNextMid(ses->server); + if (ses->capabilities & CAP_UNICODE) + buffer->Flags2 |= SMBFLG2_UNICODE; + if (ses->capabilities & CAP_STATUS32) + buffer->Flags2 |= SMBFLG2_ERR_STATUS; + + /* uid, tid can stay at zero as set in header assemble */ + + /* BB add support for turning on the signing when + this function is used after 1st of session setup requests */ + + return rc; +} +#endif /* CONFIG_CIFS_EXPERIMENTAL */ /* If the return code is zero, this function must fill in request_buf pointer */ static int @@ -1042,7 +1070,7 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, } } - cifs_small_buf_release(pSMB); +/* cifs_small_buf_release(pSMB); */ /* Freed earlier now in SendReceive2 */ if(*buf) { if(resp_buf_type == CIFS_SMALL_BUFFER) cifs_small_buf_release(iov[0].iov_base); @@ -1246,7 +1274,7 @@ CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, *nbytes += le16_to_cpu(pSMBr->Count); } - cifs_small_buf_release(pSMB); +/* cifs_small_buf_release(pSMB); */ /* Freed earlier now in SendReceive2 */ if(resp_buf_type == CIFS_SMALL_BUFFER) cifs_small_buf_release(iov[0].iov_base); else if(resp_buf_type == CIFS_LARGE_BUFFER) @@ -1325,6 +1353,85 @@ CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, } int +CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, + const __u16 smb_file_id, const int get_flag, const __u64 len, + const __u64 lkoffset, const __u16 lock_type, const int waitFlag) +{ + struct smb_com_transaction2_sfi_req *pSMB = NULL; + struct smb_com_transaction2_sfi_rsp *pSMBr = NULL; + char *data_offset; + struct cifs_posix_lock *parm_data; + int rc = 0; + int bytes_returned = 0; + __u16 params, param_offset, offset, byte_count, count; + + cFYI(1, ("Posix Lock")); + rc = small_smb_init(SMB_COM_TRANSACTION2, 15, tcon, (void **) &pSMB); + + if (rc) + return rc; + + pSMBr = (struct smb_com_transaction2_sfi_rsp *)pSMB; + + params = 6; + pSMB->MaxSetupCount = 0; + pSMB->Reserved = 0; + pSMB->Flags = 0; + pSMB->Timeout = 0; + pSMB->Reserved2 = 0; + param_offset = offsetof(struct smb_com_transaction2_sfi_req, Fid) - 4; + offset = param_offset + params; + + data_offset = (char *) (&pSMB->hdr.Protocol) + offset; + + count = sizeof(struct cifs_posix_lock); + pSMB->MaxParameterCount = cpu_to_le16(2); + pSMB->MaxDataCount = cpu_to_le16(1000); /* BB find max SMB PDU from sess */ + pSMB->SetupCount = 1; + pSMB->Reserved3 = 0; + if(get_flag) + pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_FILE_INFORMATION); + else + pSMB->SubCommand = cpu_to_le16(TRANS2_SET_FILE_INFORMATION); + byte_count = 3 /* pad */ + params + count; + pSMB->DataCount = cpu_to_le16(count); + pSMB->ParameterCount = cpu_to_le16(params); + pSMB->TotalDataCount = pSMB->DataCount; + pSMB->TotalParameterCount = pSMB->ParameterCount; + pSMB->ParameterOffset = cpu_to_le16(param_offset); + parm_data = (struct cifs_posix_lock *) + (((char *) &pSMB->hdr.Protocol) + offset); + + parm_data->lock_type = cpu_to_le16(lock_type); + if(waitFlag) + parm_data->lock_flags = 1; + parm_data->pid = cpu_to_le32(current->tgid); + parm_data->start = lkoffset; + parm_data->length = len; /* normalize negative numbers */ + + pSMB->DataOffset = cpu_to_le16(offset); + pSMB->Fid = smb_file_id; + pSMB->InformationLevel = cpu_to_le16(SMB_SET_POSIX_LOCK); + pSMB->Reserved4 = 0; + pSMB->hdr.smb_buf_length += byte_count; + pSMB->ByteCount = cpu_to_le16(byte_count); + rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, + (struct smb_hdr *) pSMBr, &bytes_returned, 0); + if (rc) { + cFYI(1, ("Send error in Posix Lock = %d", rc)); + } + + if (pSMB) + cifs_small_buf_release(pSMB); + + /* Note: On -EAGAIN error only caller can retry on handle based calls + since file handle passed in no longer valid */ + + return rc; +} + + +int CIFSSMBClose(const int xid, struct cifsTconInfo *tcon, int smb_file_id) { int rc = 0; @@ -2578,7 +2685,7 @@ qsec_out: cifs_small_buf_release(iov[0].iov_base); else if(buf_type == CIFS_LARGE_BUFFER) cifs_buf_release(iov[0].iov_base); - cifs_small_buf_release(pSMB); +/* cifs_small_buf_release(pSMB); */ /* Freed earlier now in SendReceive2 */ return rc; } @@ -2954,7 +3061,8 @@ findFirstRetry: pSMB->TotalParameterCount = cpu_to_le16(params); pSMB->ParameterCount = pSMB->TotalParameterCount; pSMB->ParameterOffset = cpu_to_le16( - offsetof(struct smb_com_transaction2_ffirst_req, SearchAttributes) - 4); + offsetof(struct smb_com_transaction2_ffirst_req, SearchAttributes) + - 4); pSMB->DataCount = 0; pSMB->DataOffset = 0; pSMB->SetupCount = 1; /* one byte, no need to make endian neutral */ @@ -2977,12 +3085,12 @@ findFirstRetry: (struct smb_hdr *) pSMBr, &bytes_returned, 0); cifs_stats_inc(&tcon->num_ffirst); - if (rc) {/* BB add logic to retry regular search if Unix search rejected unexpectedly by server */ + if (rc) {/* BB add logic to retry regular search if Unix search + rejected unexpectedly by server */ /* BB Add code to handle unsupported level rc */ cFYI(1, ("Error in FindFirst = %d", rc)); - if (pSMB) - cifs_buf_release(pSMB); + cifs_buf_release(pSMB); /* BB eventually could optimize out free and realloc of buf */ /* for this case */ @@ -2998,6 +3106,7 @@ findFirstRetry: psrch_inf->unicode = FALSE; psrch_inf->ntwrk_buf_start = (char *)pSMBr; + psrch_inf->smallBuf = 0; psrch_inf->srch_entries_start = (char *) &pSMBr->hdr.Protocol + le16_to_cpu(pSMBr->t2.DataOffset); @@ -3118,9 +3227,14 @@ int CIFSFindNext(const int xid, struct cifsTconInfo *tcon, parms = (T2_FNEXT_RSP_PARMS *)response_data; response_data = (char *)&pSMBr->hdr.Protocol + le16_to_cpu(pSMBr->t2.DataOffset); - cifs_buf_release(psrch_inf->ntwrk_buf_start); + if(psrch_inf->smallBuf) + cifs_small_buf_release( + psrch_inf->ntwrk_buf_start); + else + cifs_buf_release(psrch_inf->ntwrk_buf_start); psrch_inf->srch_entries_start = response_data; psrch_inf->ntwrk_buf_start = (char *)pSMB; + psrch_inf->smallBuf = 0; if(parms->EndofSearch) psrch_inf->endOfSearch = TRUE; else @@ -3834,6 +3948,7 @@ CIFSSMBSetFSUnixInfo(const int xid, struct cifsTconInfo *tcon, __u64 cap) cFYI(1, ("In SETFSUnixInfo")); SETFSUnixRetry: + /* BB switch to small buf init to save memory */ rc = smb_init(SMB_COM_TRANSACTION2, 15, tcon, (void **) &pSMB, (void **) &pSMBr); if (rc) diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 2a0c1f4ca0a..0b86d5ca901 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1,7 +1,7 @@ /* * fs/cifs/connect.c * - * Copyright (C) International Business Machines Corp., 2002,2005 + * Copyright (C) International Business Machines Corp., 2002,2006 * Author(s): Steve French (sfrench@us.ibm.com) * * This library is free software; you can redistribute it and/or modify @@ -564,7 +564,7 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) dump_smb(smb_buffer, length); - if (checkSMB (smb_buffer, smb_buffer->Mid, total_read+4)) { + if (checkSMB(smb_buffer, smb_buffer->Mid, total_read+4)) { cifs_dump_mem("Bad SMB: ", smb_buffer, 48); continue; } @@ -1476,6 +1476,14 @@ ipv4_connect(struct sockaddr_in *psin_server, struct socket **csocket, rc = smb_send(*csocket, smb_buf, 0x44, (struct sockaddr *)psin_server); kfree(ses_init_buf); + msleep(1); /* RFC1001 layer in at least one server + requires very short break before negprot + presumably because not expecting negprot + to follow so fast. This is a simple + solution that works without + complicating the code and causes no + significant slowing down on mount + for everyone else */ } /* else the negprot may still work without this even though malloc failed */ @@ -1920,27 +1928,34 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, cifs_sb->tcon = tcon; tcon->ses = pSesInfo; - /* do not care if following two calls succeed - informational only */ + /* do not care if following two calls succeed - informational */ CIFSSMBQFSDeviceInfo(xid, tcon); CIFSSMBQFSAttributeInfo(xid, tcon); + if (tcon->ses->capabilities & CAP_UNIX) { if(!CIFSSMBQFSUnixInfo(xid, tcon)) { - if(!volume_info.no_psx_acl) { - if(CIFS_UNIX_POSIX_ACL_CAP & - le64_to_cpu(tcon->fsUnixInfo.Capability)) - cFYI(1,("server negotiated posix acl support")); - sb->s_flags |= MS_POSIXACL; + __u64 cap = + le64_to_cpu(tcon->fsUnixInfo.Capability); + cap &= CIFS_UNIX_CAP_MASK; + if(volume_info.no_psx_acl) + cap &= ~CIFS_UNIX_POSIX_ACL_CAP; + else if(CIFS_UNIX_POSIX_ACL_CAP & cap) { + cFYI(1,("negotiated posix acl support")); + sb->s_flags |= MS_POSIXACL; } - /* Try and negotiate POSIX pathnames if we can. */ - if (volume_info.posix_paths && (CIFS_UNIX_POSIX_PATHNAMES_CAP & - le64_to_cpu(tcon->fsUnixInfo.Capability))) { - if (!CIFSSMBSetFSUnixInfo(xid, tcon, CIFS_UNIX_POSIX_PATHNAMES_CAP)) { - cFYI(1,("negotiated posix pathnames support")); - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_POSIX_PATHS; - } else { - cFYI(1,("posix pathnames support requested but not supported")); - } + if(volume_info.posix_paths == 0) + cap &= ~CIFS_UNIX_POSIX_PATHNAMES_CAP; + else if(cap & CIFS_UNIX_POSIX_PATHNAMES_CAP) { + cFYI(1,("negotiate posix pathnames")); + cifs_sb->mnt_cifs_flags |= + CIFS_MOUNT_POSIX_PATHS; + } + + cFYI(1,("Negotiate caps 0x%x",(int)cap)); + + if (CIFSSMBSetFSUnixInfo(xid, tcon, cap)) { + cFYI(1,("setting capabilities failed")); } } } @@ -2278,6 +2293,8 @@ CIFSSpnegoSessSetup(unsigned int xid, struct cifsSesInfo *ses, smb_buffer->Mid = GetNextMid(ses->server); pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; pSMB->req.AndXCommand = 0xFF; + if(ses->server->maxBuf > 64*1024) + ses->server->maxBuf = (64*1023); pSMB->req.MaxBufferSize = cpu_to_le16(ses->server->maxBuf); pSMB->req.MaxMpxCount = cpu_to_le16(ses->server->maxReq); @@ -2525,7 +2542,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, __u32 negotiate_flags, capabilities; __u16 count; - cFYI(1, ("In NTLMSSP sesssetup (negotiate) ")); + cFYI(1, ("In NTLMSSP sesssetup (negotiate)")); if(ses == NULL) return -EINVAL; domain = ses->domainName; @@ -2575,7 +2592,8 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, SecurityBlob->MessageType = NtLmNegotiate; negotiate_flags = NTLMSSP_NEGOTIATE_UNICODE | NTLMSSP_NEGOTIATE_OEM | - NTLMSSP_REQUEST_TARGET | NTLMSSP_NEGOTIATE_NTLM | 0x80000000 | + NTLMSSP_REQUEST_TARGET | NTLMSSP_NEGOTIATE_NTLM | + NTLMSSP_NEGOTIATE_56 | /* NTLMSSP_NEGOTIATE_ALWAYS_SIGN | */ NTLMSSP_NEGOTIATE_128; if(sign_CIFS_PDUs) negotiate_flags |= NTLMSSP_NEGOTIATE_SIGN; @@ -2588,26 +2606,11 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, SecurityBlob->WorkstationName.Length = 0; SecurityBlob->WorkstationName.MaximumLength = 0; - if (domain == NULL) { - SecurityBlob->DomainName.Buffer = 0; - SecurityBlob->DomainName.Length = 0; - SecurityBlob->DomainName.MaximumLength = 0; - } else { - __u16 len; - negotiate_flags |= NTLMSSP_NEGOTIATE_DOMAIN_SUPPLIED; - strncpy(bcc_ptr, domain, 63); - len = strnlen(domain, 64); - SecurityBlob->DomainName.MaximumLength = - cpu_to_le16(len); - SecurityBlob->DomainName.Buffer = - cpu_to_le32((long) &SecurityBlob-> - DomainString - - (long) &SecurityBlob->Signature); - bcc_ptr += len; - SecurityBlobLength += len; - SecurityBlob->DomainName.Length = - cpu_to_le16(len); - } + /* Domain not sent on first Sesssetup in NTLMSSP, instead it is sent + along with username on auth request (ie the response to challenge) */ + SecurityBlob->DomainName.Buffer = 0; + SecurityBlob->DomainName.Length = 0; + SecurityBlob->DomainName.MaximumLength = 0; if (ses->capabilities & CAP_UNICODE) { if ((long) bcc_ptr % 2) { *bcc_ptr = 0; @@ -2677,7 +2680,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, SecurityBlob2->MessageType)); } else if (ses) { ses->Suid = smb_buffer_response->Uid; /* UID left in le format */ - cFYI(1, ("UID = %d ", ses->Suid)); + cFYI(1, ("UID = %d", ses->Suid)); if ((pSMBr->resp.hdr.WordCount == 3) || ((pSMBr->resp.hdr.WordCount == 4) && (blob_len < @@ -2685,17 +2688,17 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, if (pSMBr->resp.hdr.WordCount == 4) { bcc_ptr += blob_len; - cFYI(1, - ("Security Blob Length %d ", + cFYI(1, ("Security Blob Length %d", blob_len)); } - cFYI(1, ("NTLMSSP Challenge rcvd ")); + cFYI(1, ("NTLMSSP Challenge rcvd")); memcpy(ses->server->cryptKey, SecurityBlob2->Challenge, CIFS_CRYPTO_KEY_SIZE); - if(SecurityBlob2->NegotiateFlags & cpu_to_le32(NTLMSSP_NEGOTIATE_NTLMV2)) + if(SecurityBlob2->NegotiateFlags & + cpu_to_le32(NTLMSSP_NEGOTIATE_NTLMV2)) *pNTLMv2_flag = TRUE; if((SecurityBlob2->NegotiateFlags & @@ -2818,7 +2821,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, bcc_ptr++; } else cFYI(1, - ("Variable field of length %d extends beyond end of smb ", + ("Variable field of length %d extends beyond end of smb", len)); } } else { @@ -2830,7 +2833,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, } } else { cERROR(1, - (" Invalid Word count %d: ", + (" Invalid Word count %d:", smb_buffer_response->WordCount)); rc = -EIO; } @@ -3447,7 +3450,7 @@ int cifs_setup_session(unsigned int xid, struct cifsSesInfo *pSesInfo, if (extended_security && (pSesInfo->capabilities & CAP_EXTENDED_SECURITY) && (pSesInfo->server->secType == NTLMSSP)) { - cFYI(1, ("New style sesssetup ")); + cFYI(1, ("New style sesssetup")); rc = CIFSSpnegoSessSetup(xid, pSesInfo, NULL /* security blob */, 0 /* blob length */, @@ -3455,7 +3458,7 @@ int cifs_setup_session(unsigned int xid, struct cifsSesInfo *pSesInfo, } else if (extended_security && (pSesInfo->capabilities & CAP_EXTENDED_SECURITY) && (pSesInfo->server->secType == RawNTLMSSP)) { - cFYI(1, ("NTLMSSP sesssetup ")); + cFYI(1, ("NTLMSSP sesssetup")); rc = CIFSNTLMSSPNegotiateSessSetup(xid, pSesInfo, &ntlmv2_flag, diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 632561dd9c5..1d0ca3eaaca 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -48,13 +48,14 @@ build_path_from_dentry(struct dentry *direntry) struct dentry *temp; int namelen = 0; char *full_path; - char dirsep = CIFS_DIR_SEP(CIFS_SB(direntry->d_sb)); + char dirsep; if(direntry == NULL) return NULL; /* not much we can do if dentry is freed and we need to reopen the file after it was closed implicitly when the server crashed */ + dirsep = CIFS_DIR_SEP(CIFS_SB(direntry->d_sb)); cifs_bp_rename_retry: for (temp = direntry; !IS_ROOT(temp);) { namelen += (1 + temp->d_name.len); @@ -255,12 +256,10 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, CIFSSMBClose(xid, pTcon, fileHandle); } else if(newinode) { pCifsFile = - kmalloc(sizeof (struct cifsFileInfo), GFP_KERNEL); + kzalloc(sizeof (struct cifsFileInfo), GFP_KERNEL); if(pCifsFile == NULL) goto cifs_create_out; - memset((char *)pCifsFile, 0, - sizeof (struct cifsFileInfo)); pCifsFile->netfid = fileHandle; pCifsFile->pid = current->tgid; pCifsFile->pInode = newinode; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index fb49aef1f2e..5c497c52977 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -555,7 +555,10 @@ int cifs_closedir(struct inode *inode, struct file *file) if (ptmp) { cFYI(1, ("closedir free smb buf in srch struct")); pCFileStruct->srch_inf.ntwrk_buf_start = NULL; - cifs_buf_release(ptmp); + if(pCFileStruct->srch_inf.smallBuf) + cifs_small_buf_release(ptmp); + else + cifs_buf_release(ptmp); } ptmp = pCFileStruct->search_resume_name; if (ptmp) { @@ -574,13 +577,14 @@ int cifs_closedir(struct inode *inode, struct file *file) int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) { int rc, xid; - __u32 lockType = LOCKING_ANDX_LARGE_FILES; __u32 numLock = 0; __u32 numUnlock = 0; __u64 length; int wait_flag = FALSE; struct cifs_sb_info *cifs_sb; struct cifsTconInfo *pTcon; + __u16 netfid; + __u8 lockType = LOCKING_ANDX_LARGE_FILES; length = 1 + pfLock->fl_end - pfLock->fl_start; rc = -EACCES; @@ -592,11 +596,11 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) pfLock->fl_end)); if (pfLock->fl_flags & FL_POSIX) - cFYI(1, ("Posix ")); + cFYI(1, ("Posix")); if (pfLock->fl_flags & FL_FLOCK) - cFYI(1, ("Flock ")); + cFYI(1, ("Flock")); if (pfLock->fl_flags & FL_SLEEP) { - cFYI(1, ("Blocking lock ")); + cFYI(1, ("Blocking lock")); wait_flag = TRUE; } if (pfLock->fl_flags & FL_ACCESS) @@ -612,21 +616,23 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) cFYI(1, ("F_WRLCK ")); numLock = 1; } else if (pfLock->fl_type == F_UNLCK) { - cFYI(1, ("F_UNLCK ")); + cFYI(1, ("F_UNLCK")); numUnlock = 1; + /* Check if unlock includes more than + one lock range */ } else if (pfLock->fl_type == F_RDLCK) { - cFYI(1, ("F_RDLCK ")); + cFYI(1, ("F_RDLCK")); lockType |= LOCKING_ANDX_SHARED_LOCK; numLock = 1; } else if (pfLock->fl_type == F_EXLCK) { - cFYI(1, ("F_EXLCK ")); + cFYI(1, ("F_EXLCK")); numLock = 1; } else if (pfLock->fl_type == F_SHLCK) { - cFYI(1, ("F_SHLCK ")); + cFYI(1, ("F_SHLCK")); lockType |= LOCKING_ANDX_SHARED_LOCK; numLock = 1; } else - cFYI(1, ("Unknown type of lock ")); + cFYI(1, ("Unknown type of lock")); cifs_sb = CIFS_SB(file->f_dentry->d_sb); pTcon = cifs_sb->tcon; @@ -635,27 +641,41 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) FreeXid(xid); return -EBADF; } + netfid = ((struct cifsFileInfo *)file->private_data)->netfid; + + /* BB add code here to normalize offset and length to + account for negative length which we can not accept over the + wire */ if (IS_GETLK(cmd)) { - rc = CIFSSMBLock(xid, pTcon, - ((struct cifsFileInfo *)file-> - private_data)->netfid, - length, - pfLock->fl_start, 0, 1, lockType, - 0 /* wait flag */ ); + if(experimEnabled && + (cifs_sb->tcon->ses->capabilities & CAP_UNIX) && + (CIFS_UNIX_FCNTL_CAP & + le64_to_cpu(cifs_sb->tcon->fsUnixInfo.Capability))) { + int posix_lock_type; + if(lockType & LOCKING_ANDX_SHARED_LOCK) + posix_lock_type = CIFS_RDLCK; + else + posix_lock_type = CIFS_WRLCK; + rc = CIFSSMBPosixLock(xid, pTcon, netfid, 1 /* get */, + length, pfLock->fl_start, + posix_lock_type, wait_flag); + FreeXid(xid); + return rc; + } + + /* BB we could chain these into one lock request BB */ + rc = CIFSSMBLock(xid, pTcon, netfid, length, pfLock->fl_start, + 0, 1, lockType, 0 /* wait flag */ ); if (rc == 0) { - rc = CIFSSMBLock(xid, pTcon, - ((struct cifsFileInfo *) file-> - private_data)->netfid, - length, + rc = CIFSSMBLock(xid, pTcon, netfid, length, pfLock->fl_start, 1 /* numUnlock */ , 0 /* numLock */ , lockType, 0 /* wait flag */ ); pfLock->fl_type = F_UNLCK; if (rc != 0) cERROR(1, ("Error unlocking previously locked " - "range %d during test of lock ", - rc)); + "range %d during test of lock", rc)); rc = 0; } else { @@ -667,12 +687,30 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) FreeXid(xid); return rc; } - - rc = CIFSSMBLock(xid, pTcon, - ((struct cifsFileInfo *) file->private_data)-> - netfid, length, - pfLock->fl_start, numUnlock, numLock, lockType, - wait_flag); + if (experimEnabled && + (cifs_sb->tcon->ses->capabilities & CAP_UNIX) && + (CIFS_UNIX_FCNTL_CAP & + le64_to_cpu(cifs_sb->tcon->fsUnixInfo.Capability))) { + int posix_lock_type; + if(lockType & LOCKING_ANDX_SHARED_LOCK) + posix_lock_type = CIFS_RDLCK; + else + posix_lock_type = CIFS_WRLCK; + + if(numUnlock == 1) + posix_lock_type = CIFS_UNLCK; + else if(numLock == 0) { + /* if no lock or unlock then nothing + to do since we do not know what it is */ + FreeXid(xid); + return -EOPNOTSUPP; + } + rc = CIFSSMBPosixLock(xid, pTcon, netfid, 0 /* set */, + length, pfLock->fl_start, + posix_lock_type, wait_flag); + } else + rc = CIFSSMBLock(xid, pTcon, netfid, length, pfLock->fl_start, + numUnlock, numLock, lockType, wait_flag); if (pfLock->fl_flags & FL_POSIX) posix_lock_file_wait(file, pfLock); FreeXid(xid); diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 598eec9778f..957ddd1571c 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -565,11 +565,14 @@ int cifs_unlink(struct inode *inode, struct dentry *direntry) struct cifsInodeInfo *cifsInode; FILE_BASIC_INFO *pinfo_buf; - cFYI(1, ("cifs_unlink, inode = 0x%p with ", inode)); + cFYI(1, ("cifs_unlink, inode = 0x%p", inode)); xid = GetXid(); - cifs_sb = CIFS_SB(inode->i_sb); + if(inode) + cifs_sb = CIFS_SB(inode->i_sb); + else + cifs_sb = CIFS_SB(direntry->d_sb); pTcon = cifs_sb->tcon; /* Unlink can be called from rename so we can not grab the sem here @@ -609,9 +612,8 @@ int cifs_unlink(struct inode *inode, struct dentry *direntry) } } else if (rc == -EACCES) { /* try only if r/o attribute set in local lookup data? */ - pinfo_buf = kmalloc(sizeof(FILE_BASIC_INFO), GFP_KERNEL); + pinfo_buf = kzalloc(sizeof(FILE_BASIC_INFO), GFP_KERNEL); if (pinfo_buf) { - memset(pinfo_buf, 0, sizeof(FILE_BASIC_INFO)); /* ATTRS set to normal clears r/o bit */ pinfo_buf->Attributes = cpu_to_le32(ATTR_NORMAL); if (!(pTcon->ses->flags & CIFS_SES_NT4)) @@ -693,9 +695,11 @@ int cifs_unlink(struct inode *inode, struct dentry *direntry) when needed */ direntry->d_inode->i_ctime = current_fs_time(inode->i_sb); } - inode->i_ctime = inode->i_mtime = current_fs_time(inode->i_sb); - cifsInode = CIFS_I(inode); - cifsInode->time = 0; /* force revalidate of dir as well */ + if(inode) { + inode->i_ctime = inode->i_mtime = current_fs_time(inode->i_sb); + cifsInode = CIFS_I(inode); + cifsInode->time = 0; /* force revalidate of dir as well */ + } kfree(full_path); FreeXid(xid); @@ -1167,7 +1171,7 @@ int cifs_setattr(struct dentry *direntry, struct iattr *attrs) nfid, npid, FALSE); atomic_dec(&open_file->wrtPending); cFYI(1,("SetFSize for attrs rc = %d", rc)); - if(rc == -EINVAL) { + if((rc == -EINVAL) || (rc == -EOPNOTSUPP)) { int bytes_written; rc = CIFSSMBWrite(xid, pTcon, nfid, 0, attrs->ia_size, @@ -1189,7 +1193,7 @@ int cifs_setattr(struct dentry *direntry, struct iattr *attrs) cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); cFYI(1, ("SetEOF by path (setattrs) rc = %d", rc)); - if(rc == -EINVAL) { + if((rc == -EINVAL) || (rc == -EOPNOTSUPP)) { __u16 netfid; int oplock = FALSE; diff --git a/fs/cifs/link.c b/fs/cifs/link.c index 8d0da7c87c7..9562f5bba65 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -67,7 +67,7 @@ cifs_hardlink(struct dentry *old_file, struct inode *inode, cifs_sb_target->local_nls, cifs_sb_target->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); - if(rc == -EIO) + if((rc == -EIO) || (rc == -EINVAL)) rc = -EOPNOTSUPP; } diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 432ba15e2c2..fafd056426e 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -72,10 +72,9 @@ sesInfoAlloc(void) struct cifsSesInfo *ret_buf; ret_buf = - (struct cifsSesInfo *) kmalloc(sizeof (struct cifsSesInfo), + (struct cifsSesInfo *) kzalloc(sizeof (struct cifsSesInfo), GFP_KERNEL); if (ret_buf) { - memset(ret_buf, 0, sizeof (struct cifsSesInfo)); write_lock(&GlobalSMBSeslock); atomic_inc(&sesInfoAllocCount); ret_buf->status = CifsNew; @@ -110,10 +109,9 @@ tconInfoAlloc(void) { struct cifsTconInfo *ret_buf; ret_buf = - (struct cifsTconInfo *) kmalloc(sizeof (struct cifsTconInfo), + (struct cifsTconInfo *) kzalloc(sizeof (struct cifsTconInfo), GFP_KERNEL); if (ret_buf) { - memset(ret_buf, 0, sizeof (struct cifsTconInfo)); write_lock(&GlobalSMBSeslock); atomic_inc(&tconInfoAllocCount); list_add(&ret_buf->cifsConnectionList, @@ -423,9 +421,7 @@ checkSMB(struct smb_hdr *smb, __u16 mid, int length) { __u32 len = smb->smb_buf_length; __u32 clc_len; /* calculated length */ - cFYI(0, - ("Entering checkSMB with Length: %x, smb_buf_length: %x", - length, len)); + cFYI(0, ("checkSMB Length: 0x%x, smb_buf_length: 0x%x", length, len)); if (((unsigned int)length < 2 + sizeof (struct smb_hdr)) || (len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4)) { if ((unsigned int)length < 2 + sizeof (struct smb_hdr)) { @@ -433,29 +429,36 @@ checkSMB(struct smb_hdr *smb, __u16 mid, int length) sizeof (struct smb_hdr) - 1) && (smb->Status.CifsError != 0)) { smb->WordCount = 0; - return 0; /* some error cases do not return wct and bcc */ + /* some error cases do not return wct and bcc */ + return 0; } else { cERROR(1, ("Length less than smb header size")); } - } if (len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) - cERROR(1, - ("smb_buf_length greater than MaxBufSize")); - cERROR(1, - ("bad smb detected. Illegal length. mid=%d", - smb->Mid)); + cERROR(1, ("smb length greater than MaxBufSize, mid=%d", + smb->Mid)); return 1; } if (checkSMBhdr(smb, mid)) return 1; clc_len = smbCalcSize_LE(smb); - if ((4 + len != clc_len) - || (4 + len != (unsigned int)length)) { - cERROR(1, ("Calculated size 0x%x vs actual length 0x%x", - clc_len, 4 + len)); - cERROR(1, ("bad smb size detected for Mid=%d", smb->Mid)); + + if(4 + len != (unsigned int)length) { + cERROR(1, ("Length read does not match RFC1001 length %d",len)); + return 1; + } + + if (4 + len != clc_len) { + /* check if bcc wrapped around for large read responses */ + if((len > 64 * 1024) && (len > clc_len)) { + /* check if lengths match mod 64K */ + if(((4 + len) & 0xFFFF) == (clc_len & 0xFFFF)) + return 0; /* bcc wrapped */ + } + cFYI(1, ("Calculated size %d vs length %d mismatch for mid %d", + clc_len, 4 + len, smb->Mid)); /* Windows XP can return a few bytes too much, presumably an illegal pad, at the end of byte range lock responses so we allow for that three byte pad, as long as actual @@ -469,8 +472,11 @@ checkSMB(struct smb_hdr *smb, __u16 mid, int length) wct and bcc to minimum size and drop the t2 parms and data */ if((4+len > clc_len) && (len <= clc_len + 512)) return 0; - else + else { + cERROR(1, ("RFC1001 size %d bigger than SMB for Mid=%d", + len, smb->Mid)); return 1; + } } return 0; } diff --git a/fs/cifs/ntlmssp.c b/fs/cifs/ntlmssp.c new file mode 100644 index 00000000000..78866f92574 --- /dev/null +++ b/fs/cifs/ntlmssp.c @@ -0,0 +1,129 @@ +/* + * fs/cifs/ntlmssp.h + * + * Copyright (c) International Business Machines Corp., 2006 + * Author(s): Steve French (sfrench@us.ibm.com) + * + * This library is free software; you can redistribute it and/or modify + * it under the terms of the GNU Lesser General Public License as published + * by the Free Software Foundation; either version 2.1 of the License, or + * (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See + * the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "cifspdu.h" +#include "cifsglob.h" +#include "cifsproto.h" +#include "cifs_unicode.h" +#include "cifs_debug.h" +#include "ntlmssp.h" +#include "nterr.h" + +#ifdef CONFIG_CIFS_EXPERIMENTAL +static __u32 cifs_ssetup_hdr(struct cifsSesInfo *ses, SESSION_SETUP_ANDX *pSMB) +{ + __u32 capabilities = 0; + + /* init fields common to all four types of SessSetup */ + /* note that header is initialized to zero in header_assemble */ + pSMB->req.AndXCommand = 0xFF; + pSMB->req.MaxBufferSize = cpu_to_le16(ses->server->maxBuf); + pSMB->req.MaxMpxCount = cpu_to_le16(ses->server->maxReq); + + /* Now no need to set SMBFLG_CASELESS or obsolete CANONICAL PATH */ + + /* BB verify whether signing required on neg or just on auth frame + (and NTLM case) */ + + capabilities = CAP_LARGE_FILES | CAP_NT_SMBS | CAP_LEVEL_II_OPLOCKS | + CAP_LARGE_WRITE_X | CAP_LARGE_READ_X; + + if(ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) + pSMB->req.hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; + + if (ses->capabilities & CAP_UNICODE) { + pSMB->req.hdr.Flags2 |= SMBFLG2_UNICODE; + capabilities |= CAP_UNICODE; + } + if (ses->capabilities & CAP_STATUS32) { + pSMB->req.hdr.Flags2 |= SMBFLG2_ERR_STATUS; + capabilities |= CAP_STATUS32; + } + if (ses->capabilities & CAP_DFS) { + pSMB->req.hdr.Flags2 |= SMBFLG2_DFS; + capabilities |= CAP_DFS; + } + + /* BB check whether to init vcnum BB */ + return capabilities; +} +int +CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, const int type, + int * pNTLMv2_flg, const struct nls_table *nls_cp) +{ + int rc = 0; + int wct; + struct smb_hdr *smb_buffer; + char *bcc_ptr; + SESSION_SETUP_ANDX *pSMB; + __u32 capabilities; + + if(ses == NULL) + return -EINVAL; + + cFYI(1,("SStp type: %d",type)); + if(type < CIFS_NTLM) { +#ifndef CONFIG_CIFS_WEAK_PW_HASH + /* LANMAN and plaintext are less secure and off by default. + So we make this explicitly be turned on in kconfig (in the + build) and turned on at runtime (changed from the default) + in proc/fs/cifs or via mount parm. Unfortunately this is + needed for old Win (e.g. Win95), some obscure NAS and OS/2 */ + return -EOPNOTSUPP; +#endif + wct = 10; /* lanman 2 style sessionsetup */ + } else if(type < CIFS_NTLMSSP_NEG) + wct = 13; /* old style NTLM sessionsetup */ + else /* same size for negotiate or auth, NTLMSSP or extended security */ + wct = 12; + + rc = small_smb_init_no_tc(SMB_COM_SESSION_SETUP_ANDX, wct, ses, + (void **)&smb_buffer); + if(rc) + return rc; + + pSMB = (SESSION_SETUP_ANDX *)smb_buffer; + + capabilities = cifs_ssetup_hdr(ses, pSMB); + bcc_ptr = pByteArea(smb_buffer); + if(type > CIFS_NTLM) { + pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; + capabilities |= CAP_EXTENDED_SECURITY; + pSMB->req.Capabilities = cpu_to_le32(capabilities); + /* BB set password lengths */ + } else if(type < CIFS_NTLM) /* lanman */ { + /* no capabilities flags in old lanman negotiation */ + /* pSMB->old_req.PasswordLength = */ /* BB fixme BB */ + } else /* type CIFS_NTLM */ { + pSMB->req_no_secext.Capabilities = cpu_to_le32(capabilities); + pSMB->req_no_secext.CaseInsensitivePasswordLength = + cpu_to_le16(CIFS_SESSION_KEY_SIZE); + pSMB->req_no_secext.CaseSensitivePasswordLength = + cpu_to_le16(CIFS_SESSION_KEY_SIZE); + } + + +/* rc = SendReceive2(xid, ses, iov, num_iovecs, &resp_buf_type, 0); */ + /* SMB request buf freed in SendReceive2 */ + + return rc; +} +#endif /* CONFIG_CIFS_EXPERIMENTAL */ diff --git a/fs/cifs/ntlmssp.h b/fs/cifs/ntlmssp.h index 803389b64a2..d39b712a11c 100644 --- a/fs/cifs/ntlmssp.h +++ b/fs/cifs/ntlmssp.h @@ -1,7 +1,7 @@ /* * fs/cifs/ntlmssp.h * - * Copyright (c) International Business Machines Corp., 2002 + * Copyright (c) International Business Machines Corp., 2002,2006 * Author(s): Steve French (sfrench@us.ibm.com) * * This library is free software; you can redistribute it and/or modify diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 488bd0d81dc..2f6e2825571 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -604,7 +604,12 @@ static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, cifsFile->search_resume_name = NULL; if(cifsFile->srch_inf.ntwrk_buf_start) { cFYI(1,("freeing SMB ff cache buf on search rewind")); - cifs_buf_release(cifsFile->srch_inf.ntwrk_buf_start); + if(cifsFile->srch_inf.smallBuf) + cifs_small_buf_release(cifsFile->srch_inf. + ntwrk_buf_start); + else + cifs_buf_release(cifsFile->srch_inf. + ntwrk_buf_start); } rc = initiate_cifs_search(xid,file); if(rc) { diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index b12cb8a7da7..3da80409466 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -309,17 +309,16 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, *pRespBufType = CIFS_NO_BUFFER; /* no response buf yet */ - if (ses == NULL) { - cERROR(1,("Null smb session")); - return -EIO; - } - if(ses->server == NULL) { - cERROR(1,("Null tcp session")); + if ((ses == NULL) || (ses->server == NULL)) { + cifs_small_buf_release(in_buf); + cERROR(1,("Null session")); return -EIO; } - if(ses->server->tcpStatus == CifsExiting) + if(ses->server->tcpStatus == CifsExiting) { + cifs_small_buf_release(in_buf); return -ENOENT; + } /* Ensure that we do not send more than 50 overlapping requests to the same server. We may make this configurable later or @@ -346,6 +345,7 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, } else { if(ses->server->tcpStatus == CifsExiting) { spin_unlock(&GlobalMid_Lock); + cifs_small_buf_release(in_buf); return -ENOENT; } @@ -385,6 +385,7 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, midQ = AllocMidQEntry(in_buf, ses); if (midQ == NULL) { up(&ses->server->tcpSem); + cifs_small_buf_release(in_buf); /* If not lock req, update # of requests on wire to server */ if(long_op < 3) { atomic_dec(&ses->server->inFlight); @@ -408,14 +409,18 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, if(rc < 0) { DeleteMidQEntry(midQ); up(&ses->server->tcpSem); + cifs_small_buf_release(in_buf); /* If not lock req, update # of requests on wire to server */ if(long_op < 3) { atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); } return rc; - } else + } else { up(&ses->server->tcpSem); + cifs_small_buf_release(in_buf); + } + if (long_op == -1) goto cifs_no_response_exit2; else if (long_op == 2) /* writes past end of file can take loong time */ @@ -543,6 +548,7 @@ cifs_no_response_exit2: out_unlock2: up(&ses->server->tcpSem); + cifs_small_buf_release(in_buf); /* If not lock req, update # of requests on wire to server */ if(long_op < 3) { atomic_dec(&ses->server->inFlight); diff --git a/fs/dcache.c b/fs/dcache.c index 19458d39950..940d188e5d1 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1101,6 +1101,32 @@ next: } /** + * d_hash_and_lookup - hash the qstr then search for a dentry + * @dir: Directory to search in + * @name: qstr of name we wish to find + * + * On hash failure or on lookup failure NULL is returned. + */ +struct dentry *d_hash_and_lookup(struct dentry *dir, struct qstr *name) +{ + struct dentry *dentry = NULL; + + /* + * Check for a fs-specific hash function. Note that we must + * calculate the standard hash first, as the d_op->d_hash() + * routine may choose to leave the hash value unchanged. + */ + name->hash = full_name_hash(name->name, name->len); + if (dir->d_op && dir->d_op->d_hash) { + if (dir->d_op->d_hash(dir, name) < 0) + goto out; + } + dentry = d_lookup(dir, name); +out: + return dentry; +} + +/** * d_validate - verify dentry provided from insecure source * @dentry: The dentry alleged to be valid child of @dparent * @dparent: The parent dentry (known to be valid) @@ -1172,11 +1198,11 @@ void d_delete(struct dentry * dentry) spin_lock(&dentry->d_lock); isdir = S_ISDIR(dentry->d_inode->i_mode); if (atomic_read(&dentry->d_count) == 1) { - /* remove this and other inotify debug checks after 2.6.18 */ - dentry->d_flags &= ~DCACHE_INOTIFY_PARENT_WATCHED; - dentry_iput(dentry); fsnotify_nameremove(dentry, isdir); + + /* remove this and other inotify debug checks after 2.6.18 */ + dentry->d_flags &= ~DCACHE_INOTIFY_PARENT_WATCHED; return; } @@ -1616,26 +1642,12 @@ ino_t find_inode_number(struct dentry *dir, struct qstr *name) struct dentry * dentry; ino_t ino = 0; - /* - * Check for a fs-specific hash function. Note that we must - * calculate the standard hash first, as the d_op->d_hash() - * routine may choose to leave the hash value unchanged. - */ - name->hash = full_name_hash(name->name, name->len); - if (dir->d_op && dir->d_op->d_hash) - { - if (dir->d_op->d_hash(dir, name) != 0) - goto out; - } - - dentry = d_lookup(dir, name); - if (dentry) - { + dentry = d_hash_and_lookup(dir, name); + if (dentry) { if (dentry->d_inode) ino = dentry->d_inode->i_ino; dput(dentry); } -out: return ino; } diff --git a/fs/direct-io.c b/fs/direct-io.c index 910a8ed74b5..b05d1b21877 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -929,8 +929,7 @@ do_holes: block_in_page += this_chunk_blocks; dio->blocks_available -= this_chunk_blocks; next_block: - if (dio->block_in_file > dio->final_block_in_request) - BUG(); + BUG_ON(dio->block_in_file > dio->final_block_in_request); if (dio->block_in_file == dio->final_block_in_request) break; } diff --git a/fs/dquot.c b/fs/dquot.c index 6b388692093..81d87a413c6 100644 --- a/fs/dquot.c +++ b/fs/dquot.c @@ -590,8 +590,7 @@ we_slept: atomic_dec(&dquot->dq_count); #ifdef __DQUOT_PARANOIA /* sanity check */ - if (!list_empty(&dquot->dq_free)) - BUG(); + BUG_ON(!list_empty(&dquot->dq_free)); #endif put_dquot_last(dquot); spin_unlock(&dq_list_lock); @@ -666,8 +665,7 @@ we_slept: return NODQUOT; } #ifdef __DQUOT_PARANOIA - if (!dquot->dq_sb) /* Has somebody invalidated entry under us? */ - BUG(); + BUG_ON(!dquot->dq_sb); /* Has somebody invalidated entry under us? */ #endif return dquot; diff --git a/fs/exec.c b/fs/exec.c index 950ebd43cdc..0291a68a362 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -561,7 +561,7 @@ static int exec_mmap(struct mm_struct *mm) arch_pick_mmap_layout(mm); if (old_mm) { up_read(&old_mm->mmap_sem); - if (active_mm != old_mm) BUG(); + BUG_ON(active_mm != old_mm); mmput(old_mm); return 0; } diff --git a/fs/fcntl.c b/fs/fcntl.c index 2a2479196f9..d35cbc6bc11 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -453,8 +453,7 @@ static void send_sigio_to_task(struct task_struct *p, /* Make sure we are called with one of the POLL_* reasons, otherwise we could leak kernel stack into userspace. */ - if ((reason & __SI_MASK) != __SI_POLL) - BUG(); + BUG_ON((reason & __SI_MASK) != __SI_POLL); if (reason - POLL_IN >= NSIGPOLL) si.si_band = ~0L; else diff --git a/fs/freevxfs/vxfs_olt.c b/fs/freevxfs/vxfs_olt.c index 76a0708ae97..04950084790 100644 --- a/fs/freevxfs/vxfs_olt.c +++ b/fs/freevxfs/vxfs_olt.c @@ -42,24 +42,21 @@ static inline void vxfs_get_fshead(struct vxfs_oltfshead *fshp, struct vxfs_sb_info *infp) { - if (infp->vsi_fshino) - BUG(); + BUG_ON(infp->vsi_fshino); infp->vsi_fshino = fshp->olt_fsino[0]; } static inline void vxfs_get_ilist(struct vxfs_oltilist *ilistp, struct vxfs_sb_info *infp) { - if (infp->vsi_iext) - BUG(); + BUG_ON(infp->vsi_iext); infp->vsi_iext = ilistp->olt_iext[0]; } static inline u_long vxfs_oblock(struct super_block *sbp, daddr_t block, u_long bsize) { - if (sbp->s_blocksize % bsize) - BUG(); + BUG_ON(sbp->s_blocksize % bsize); return (block * (sbp->s_blocksize / bsize)); } diff --git a/fs/hfsplus/bnode.c b/fs/hfsplus/bnode.c index 8f07e8fbd03..746abc9ecf7 100644 --- a/fs/hfsplus/bnode.c +++ b/fs/hfsplus/bnode.c @@ -466,8 +466,7 @@ void hfs_bnode_unhash(struct hfs_bnode *node) for (p = &node->tree->node_hash[hfs_bnode_hash(node->this)]; *p && *p != node; p = &(*p)->next_hash) ; - if (!*p) - BUG(); + BUG_ON(!*p); *p = node->next_hash; node->tree->node_hash_cnt--; } @@ -622,8 +621,7 @@ void hfs_bnode_put(struct hfs_bnode *node) dprint(DBG_BNODE_REFS, "put_node(%d:%d): %d\n", node->tree->cnid, node->this, atomic_read(&node->refcnt)); - if (!atomic_read(&node->refcnt)) - BUG(); + BUG_ON(!atomic_read(&node->refcnt)); if (!atomic_dec_and_lock(&node->refcnt, &tree->hash_lock)) return; for (i = 0; i < tree->pages_per_bnode; i++) { diff --git a/fs/hfsplus/btree.c b/fs/hfsplus/btree.c index a67edfa34e9..effa8991999 100644 --- a/fs/hfsplus/btree.c +++ b/fs/hfsplus/btree.c @@ -269,8 +269,7 @@ void hfs_bmap_free(struct hfs_bnode *node) u8 *data, byte, m; dprint(DBG_BNODE_MOD, "btree_free_node: %u\n", node->this); - if (!node->this) - BUG(); + BUG_ON(!node->this); tree = node->tree; nidx = node->this; node = hfs_bnode_find(tree, 0); diff --git a/fs/hppfs/hppfs_kern.c b/fs/hppfs/hppfs_kern.c index 2ba20cdb5ba..5e6363be246 100644 --- a/fs/hppfs/hppfs_kern.c +++ b/fs/hppfs/hppfs_kern.c @@ -216,10 +216,10 @@ static struct dentry *hppfs_lookup(struct inode *ino, struct dentry *dentry, static struct inode_operations hppfs_file_iops = { }; -static ssize_t read_proc(struct file *file, char *buf, ssize_t count, +static ssize_t read_proc(struct file *file, char __user *buf, ssize_t count, loff_t *ppos, int is_user) { - ssize_t (*read)(struct file *, char *, size_t, loff_t *); + ssize_t (*read)(struct file *, char __user *, size_t, loff_t *); ssize_t n; read = file->f_dentry->d_inode->i_fop->read; @@ -236,7 +236,7 @@ static ssize_t read_proc(struct file *file, char *buf, ssize_t count, return n; } -static ssize_t hppfs_read_file(int fd, char *buf, ssize_t count) +static ssize_t hppfs_read_file(int fd, char __user *buf, ssize_t count) { ssize_t n; int cur, err; @@ -274,7 +274,7 @@ static ssize_t hppfs_read_file(int fd, char *buf, ssize_t count) return n; } -static ssize_t hppfs_read(struct file *file, char *buf, size_t count, +static ssize_t hppfs_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct hppfs_private *hppfs = file->private_data; @@ -313,12 +313,12 @@ static ssize_t hppfs_read(struct file *file, char *buf, size_t count, return(count); } -static ssize_t hppfs_write(struct file *file, const char *buf, size_t len, +static ssize_t hppfs_write(struct file *file, const char __user *buf, size_t len, loff_t *ppos) { struct hppfs_private *data = file->private_data; struct file *proc_file = data->proc_file; - ssize_t (*write)(struct file *, const char *, size_t, loff_t *); + ssize_t (*write)(struct file *, const char __user *, size_t, loff_t *); int err; write = proc_file->f_dentry->d_inode->i_fop->write; @@ -658,7 +658,7 @@ static struct super_operations hppfs_sbops = { .statfs = hppfs_statfs, }; -static int hppfs_readlink(struct dentry *dentry, char *buffer, int buflen) +static int hppfs_readlink(struct dentry *dentry, char __user *buffer, int buflen) { struct file *proc_file; struct dentry *proc_dentry; diff --git a/fs/inode.c b/fs/inode.c index 32b7c337502..3a2446a27d2 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -172,8 +172,7 @@ static struct inode *alloc_inode(struct super_block *sb) void destroy_inode(struct inode *inode) { - if (inode_has_buffers(inode)) - BUG(); + BUG_ON(inode_has_buffers(inode)); security_inode_free(inode); if (inode->i_sb->s_op->destroy_inode) inode->i_sb->s_op->destroy_inode(inode); @@ -249,12 +248,9 @@ void clear_inode(struct inode *inode) might_sleep(); invalidate_inode_buffers(inode); - if (inode->i_data.nrpages) - BUG(); - if (!(inode->i_state & I_FREEING)) - BUG(); - if (inode->i_state & I_CLEAR) - BUG(); + BUG_ON(inode->i_data.nrpages); + BUG_ON(!(inode->i_state & I_FREEING)); + BUG_ON(inode->i_state & I_CLEAR); wait_on_inode(inode); DQUOT_DROP(inode); if (inode->i_sb && inode->i_sb->s_op->clear_inode) @@ -1054,8 +1050,7 @@ void generic_delete_inode(struct inode *inode) hlist_del_init(&inode->i_hash); spin_unlock(&inode_lock); wake_up_inode(inode); - if (inode->i_state != I_CLEAR) - BUG(); + BUG_ON(inode->i_state != I_CLEAR); destroy_inode(inode); } diff --git a/fs/jffs2/background.c b/fs/jffs2/background.c index 7b77a954112..ff2a872e80e 100644 --- a/fs/jffs2/background.c +++ b/fs/jffs2/background.c @@ -35,8 +35,7 @@ int jffs2_start_garbage_collect_thread(struct jffs2_sb_info *c) pid_t pid; int ret = 0; - if (c->gc_task) - BUG(); + BUG_ON(c->gc_task); init_completion(&c->gc_thread_start); init_completion(&c->gc_thread_exit); diff --git a/fs/locks.c b/fs/locks.c index 4d9e71d43e7..dda83d6cd48 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -168,18 +168,9 @@ static void locks_release_private(struct file_lock *fl) /* Free a lock which is not in use. */ static void locks_free_lock(struct file_lock *fl) { - if (fl == NULL) { - BUG(); - return; - } - if (waitqueue_active(&fl->fl_wait)) - panic("Attempting to free lock with active wait queue"); - - if (!list_empty(&fl->fl_block)) - panic("Attempting to free lock with active block list"); - - if (!list_empty(&fl->fl_link)) - panic("Attempting to free lock on active lock list"); + BUG_ON(waitqueue_active(&fl->fl_wait)); + BUG_ON(!list_empty(&fl->fl_block)); + BUG_ON(!list_empty(&fl->fl_link)); locks_release_private(fl); kmem_cache_free(filelock_cache, fl); @@ -735,8 +726,9 @@ EXPORT_SYMBOL(posix_locks_deadlock); * at the head of the list, but that's secret knowledge known only to * flock_lock_file and posix_lock_file. */ -static int flock_lock_file(struct file *filp, struct file_lock *new_fl) +static int flock_lock_file(struct file *filp, struct file_lock *request) { + struct file_lock *new_fl = NULL; struct file_lock **before; struct inode * inode = filp->f_dentry->d_inode; int error = 0; @@ -751,17 +743,19 @@ static int flock_lock_file(struct file *filp, struct file_lock *new_fl) continue; if (filp != fl->fl_file) continue; - if (new_fl->fl_type == fl->fl_type) + if (request->fl_type == fl->fl_type) goto out; found = 1; locks_delete_lock(before); break; } - unlock_kernel(); - if (new_fl->fl_type == F_UNLCK) - return 0; + if (request->fl_type == F_UNLCK) + goto out; + new_fl = locks_alloc_lock(); + if (new_fl == NULL) + goto out; /* * If a higher-priority process was blocked on the old file lock, * give it the opportunity to lock the file. @@ -769,26 +763,27 @@ static int flock_lock_file(struct file *filp, struct file_lock *new_fl) if (found) cond_resched(); - lock_kernel(); for_each_lock(inode, before) { struct file_lock *fl = *before; if (IS_POSIX(fl)) break; if (IS_LEASE(fl)) continue; - if (!flock_locks_conflict(new_fl, fl)) + if (!flock_locks_conflict(request, fl)) continue; error = -EAGAIN; - if (new_fl->fl_flags & FL_SLEEP) { - locks_insert_block(fl, new_fl); - } + if (request->fl_flags & FL_SLEEP) + locks_insert_block(fl, request); goto out; } + locks_copy_lock(new_fl, request); locks_insert_lock(&inode->i_flock, new_fl); - error = 0; + new_fl = NULL; out: unlock_kernel(); + if (new_fl) + locks_free_lock(new_fl); return error; } @@ -1569,9 +1564,7 @@ asmlinkage long sys_flock(unsigned int fd, unsigned int cmd) error = flock_lock_file_wait(filp, lock); out_free: - if (list_empty(&lock->fl_link)) { - locks_free_lock(lock); - } + locks_free_lock(lock); out_putf: fput(filp); diff --git a/fs/msdos/namei.c b/fs/msdos/namei.c index 626a367bcd8..5b76ccd19e3 100644 --- a/fs/msdos/namei.c +++ b/fs/msdos/namei.c @@ -12,14 +12,6 @@ #include <linux/msdos_fs.h> #include <linux/smp_lock.h> -/* MS-DOS "device special files" */ -static const unsigned char *reserved_names[] = { - "CON ", "PRN ", "NUL ", "AUX ", - "LPT1 ", "LPT2 ", "LPT3 ", "LPT4 ", - "COM1 ", "COM2 ", "COM3 ", "COM4 ", - NULL -}; - /* Characters that are undesirable in an MS-DOS file name */ static unsigned char bad_chars[] = "*?<>|\""; static unsigned char bad_if_strict_pc[] = "+=,; "; @@ -40,7 +32,6 @@ static int msdos_format_name(const unsigned char *name, int len, */ { unsigned char *walk; - const unsigned char **reserved; unsigned char c; int space; @@ -127,11 +118,7 @@ static int msdos_format_name(const unsigned char *name, int len, } while (walk - res < MSDOS_NAME) *walk++ = ' '; - if (!opts->atari) - /* GEMDOS is less stupid and has no reserved names */ - for (reserved = reserved_names; *reserved; reserved++) - if (!strncmp(res, *reserved, 8)) - return -EINVAL; + return 0; } diff --git a/fs/namei.c b/fs/namei.c index 22f6e8d16aa..96723ae83c8 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1254,7 +1254,7 @@ out: return dentry; } -struct dentry * lookup_hash(struct nameidata *nd) +static struct dentry *lookup_hash(struct nameidata *nd) { return __lookup_hash(&nd->last, nd->dentry, nd); } @@ -2697,7 +2697,6 @@ EXPORT_SYMBOL(follow_up); EXPORT_SYMBOL(get_write_access); /* binfmt_aout */ EXPORT_SYMBOL(getname); EXPORT_SYMBOL(lock_rename); -EXPORT_SYMBOL(lookup_hash); EXPORT_SYMBOL(lookup_one_len); EXPORT_SYMBOL(page_follow_link_light); EXPORT_SYMBOL(page_put_link); diff --git a/fs/pipe.c b/fs/pipe.c index 109a102c150..795df987cd3 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -95,6 +95,8 @@ static void anon_pipe_buf_release(struct pipe_inode_info *info, struct pipe_buff { struct page *page = buf->page; + buf->flags &= ~PIPE_BUF_FLAG_STOLEN; + /* * If nobody else uses this page, and we don't already have a * temporary page, let's keep track of it as a one-deep @@ -124,7 +126,7 @@ static void anon_pipe_buf_unmap(struct pipe_inode_info *info, struct pipe_buffer static int anon_pipe_buf_steal(struct pipe_inode_info *info, struct pipe_buffer *buf) { - buf->stolen = 1; + buf->flags |= PIPE_BUF_FLAG_STOLEN; return 0; } diff --git a/fs/proc/base.c b/fs/proc/base.c index 8f1f49ceebe..a3a3eecef68 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -534,12 +534,15 @@ static int proc_oom_score(struct task_struct *task, char *buffer) /* If the process being read is separated by chroot from the reading process, * don't let the reader access the threads. + * + * note: this does dput(root) and mntput(vfsmnt) on exit. */ static int proc_check_chroot(struct dentry *root, struct vfsmount *vfsmnt) { struct dentry *de, *base; struct vfsmount *our_vfsmnt, *mnt; int res = 0; + read_lock(¤t->fs->lock); our_vfsmnt = mntget(current->fs->rootmnt); base = dget(current->fs->root); @@ -549,11 +552,11 @@ static int proc_check_chroot(struct dentry *root, struct vfsmount *vfsmnt) de = root; mnt = vfsmnt; - while (vfsmnt != our_vfsmnt) { - if (vfsmnt == vfsmnt->mnt_parent) + while (mnt != our_vfsmnt) { + if (mnt == mnt->mnt_parent) goto out; - de = vfsmnt->mnt_mountpoint; - vfsmnt = vfsmnt->mnt_parent; + de = mnt->mnt_mountpoint; + mnt = mnt->mnt_parent; } if (!is_subdir(de, base)) @@ -564,7 +567,7 @@ exit: dput(base); mntput(our_vfsmnt); dput(root); - mntput(mnt); + mntput(vfsmnt); return res; out: spin_unlock(&vfsmount_lock); diff --git a/fs/proc/proc_misc.c b/fs/proc/proc_misc.c index ef5a3323f4b..5c10ea15742 100644 --- a/fs/proc/proc_misc.c +++ b/fs/proc/proc_misc.c @@ -249,144 +249,60 @@ static int cpuinfo_open(struct inode *inode, struct file *file) return seq_open(file, &cpuinfo_op); } -enum devinfo_states { - CHR_HDR, - CHR_LIST, - BLK_HDR, - BLK_LIST, - DEVINFO_DONE -}; - -struct devinfo_state { - void *chrdev; - void *blkdev; - unsigned int num_records; - unsigned int cur_record; - enum devinfo_states state; +static struct file_operations proc_cpuinfo_operations = { + .open = cpuinfo_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, }; -static void *devinfo_start(struct seq_file *f, loff_t *pos) +static int devinfo_show(struct seq_file *f, void *v) { - struct devinfo_state *info = f->private; + int i = *(loff_t *) v; - if (*pos) { - if ((info) && (*pos <= info->num_records)) - return info; - return NULL; + if (i < CHRDEV_MAJOR_HASH_SIZE) { + if (i == 0) + seq_printf(f, "Character devices:\n"); + chrdev_show(f, i); + } else { + i -= CHRDEV_MAJOR_HASH_SIZE; + if (i == 0) + seq_printf(f, "\nBlock devices:\n"); + blkdev_show(f, i); } - info = kmalloc(sizeof(*info), GFP_KERNEL); - f->private = info; - info->chrdev = acquire_chrdev_list(); - info->blkdev = acquire_blkdev_list(); - info->state = CHR_HDR; - info->num_records = count_chrdev_list(); - info->num_records += count_blkdev_list(); - info->num_records += 2; /* Character and Block headers */ - *pos = 1; - info->cur_record = *pos; - return info; + return 0; } -static void *devinfo_next(struct seq_file *f, void *v, loff_t *pos) +static void *devinfo_start(struct seq_file *f, loff_t *pos) { - int idummy; - char *ndummy; - struct devinfo_state *info = f->private; - - switch (info->state) { - case CHR_HDR: - info->state = CHR_LIST; - (*pos)++; - /*fallthrough*/ - case CHR_LIST: - if (get_chrdev_info(info->chrdev,&idummy,&ndummy)) { - /* - * The character dev list is complete - */ - info->state = BLK_HDR; - } else { - info->chrdev = get_next_chrdev(info->chrdev); - } - (*pos)++; - break; - case BLK_HDR: - info->state = BLK_LIST; - (*pos)++; - /*fallthrough*/ - case BLK_LIST: - if (get_blkdev_info(info->blkdev,&idummy,&ndummy)) { - /* - * The block dev list is complete - */ - info->state = DEVINFO_DONE; - } else { - info->blkdev = get_next_blkdev(info->blkdev); - } - (*pos)++; - break; - case DEVINFO_DONE: - (*pos)++; - info->cur_record = *pos; - info = NULL; - break; - default: - break; - } - if (info) - info->cur_record = *pos; - return info; + if (*pos < (BLKDEV_MAJOR_HASH_SIZE + CHRDEV_MAJOR_HASH_SIZE)) + return pos; + return NULL; } -static void devinfo_stop(struct seq_file *f, void *v) +static void *devinfo_next(struct seq_file *f, void *v, loff_t *pos) { - struct devinfo_state *info = f->private; - - if (info) { - release_chrdev_list(info->chrdev); - release_blkdev_list(info->blkdev); - f->private = NULL; - kfree(info); - } + (*pos)++; + if (*pos >= (BLKDEV_MAJOR_HASH_SIZE + CHRDEV_MAJOR_HASH_SIZE)) + return NULL; + return pos; } -static int devinfo_show(struct seq_file *f, void *arg) -{ - int major; - char *name; - struct devinfo_state *info = f->private; - - switch(info->state) { - case CHR_HDR: - seq_printf(f,"Character devices:\n"); - /* fallthrough */ - case CHR_LIST: - if (!get_chrdev_info(info->chrdev,&major,&name)) - seq_printf(f,"%3d %s\n",major,name); - break; - case BLK_HDR: - seq_printf(f,"\nBlock devices:\n"); - /* fallthrough */ - case BLK_LIST: - if (!get_blkdev_info(info->blkdev,&major,&name)) - seq_printf(f,"%3d %s\n",major,name); - break; - default: - break; - } - - return 0; +static void devinfo_stop(struct seq_file *f, void *v) +{ + /* Nothing to do */ } -static struct seq_operations devinfo_op = { - .start = devinfo_start, - .next = devinfo_next, - .stop = devinfo_stop, - .show = devinfo_show, +static struct seq_operations devinfo_ops = { + .start = devinfo_start, + .next = devinfo_next, + .stop = devinfo_stop, + .show = devinfo_show }; -static int devinfo_open(struct inode *inode, struct file *file) +static int devinfo_open(struct inode *inode, struct file *filp) { - return seq_open(file, &devinfo_op); + return seq_open(filp, &devinfo_ops); } static struct file_operations proc_devinfo_operations = { @@ -396,13 +312,6 @@ static struct file_operations proc_devinfo_operations = { .release = seq_release, }; -static struct file_operations proc_cpuinfo_operations = { - .open = cpuinfo_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - extern struct seq_operations vmstat_op; static int vmstat_open(struct inode *inode, struct file *file) { diff --git a/fs/select.c b/fs/select.c index b3a3a1326af..071660fa7b0 100644 --- a/fs/select.c +++ b/fs/select.c @@ -314,7 +314,7 @@ static int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, int ret, size, max_fdset; struct fdtable *fdt; /* Allocate small arguments on the stack to save memory and be faster */ - char stack_fds[SELECT_STACK_ALLOC]; + long stack_fds[SELECT_STACK_ALLOC/sizeof(long)]; ret = -EINVAL; if (n < 0) @@ -639,8 +639,10 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) struct poll_list *walk; struct fdtable *fdt; int max_fdset; - /* Allocate small arguments on the stack to save memory and be faster */ - char stack_pps[POLL_STACK_ALLOC]; + /* Allocate small arguments on the stack to save memory and be + faster - use long to make sure the buffer is aligned properly + on 64 bit archs to avoid unaligned access */ + long stack_pps[POLL_STACK_ALLOC/sizeof(long)]; struct poll_list *stack_pp = NULL; /* Do a sanity check on nfds ... */ diff --git a/fs/smbfs/file.c b/fs/smbfs/file.c index c56bd99a970..ed9a24d19d7 100644 --- a/fs/smbfs/file.c +++ b/fs/smbfs/file.c @@ -178,11 +178,9 @@ smb_writepage(struct page *page, struct writeback_control *wbc) unsigned offset = PAGE_CACHE_SIZE; int err; - if (!mapping) - BUG(); + BUG_ON(!mapping); inode = mapping->host; - if (!inode) - BUG(); + BUG_ON(!inode); end_index = inode->i_size >> PAGE_CACHE_SHIFT; diff --git a/fs/splice.c b/fs/splice.c index 7c2bbf18d7a..bfa42a277bb 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -22,7 +22,10 @@ #include <linux/pipe_fs_i.h> #include <linux/mm_inline.h> #include <linux/swap.h> +#include <linux/writeback.h> +#include <linux/buffer_head.h> #include <linux/module.h> +#include <linux/syscalls.h> /* * Passed to the actors @@ -34,28 +37,37 @@ struct splice_desc { loff_t pos; /* file position */ }; +/* + * Attempt to steal a page from a pipe buffer. This should perhaps go into + * a vm helper function, it's already simplified quite a bit by the + * addition of remove_mapping(). If success is returned, the caller may + * attempt to reuse this page for another destination. + */ static int page_cache_pipe_buf_steal(struct pipe_inode_info *info, struct pipe_buffer *buf) { struct page *page = buf->page; + struct address_space *mapping = page_mapping(page); WARN_ON(!PageLocked(page)); WARN_ON(!PageUptodate(page)); - if (!remove_mapping(page_mapping(page), page)) - return 1; + /* + * At least for ext2 with nobh option, we need to wait on writeback + * completing on this page, since we'll remove it from the pagecache. + * Otherwise truncate wont wait on the page, allowing the disk + * blocks to be reused by someone else before we actually wrote our + * data to them. fs corruption ensues. + */ + wait_on_page_writeback(page); - if (PageLRU(page)) { - struct zone *zone = page_zone(page); + if (PagePrivate(page)) + try_to_release_page(page, mapping_gfp_mask(mapping)); - spin_lock_irq(&zone->lru_lock); - BUG_ON(!PageLRU(page)); - __ClearPageLRU(page); - del_page_from_lru(zone, page); - spin_unlock_irq(&zone->lru_lock); - } + if (!remove_mapping(mapping, page)) + return 1; - buf->stolen = 1; + buf->flags |= PIPE_BUF_FLAG_STOLEN | PIPE_BUF_FLAG_LRU; return 0; } @@ -64,7 +76,7 @@ static void page_cache_pipe_buf_release(struct pipe_inode_info *info, { page_cache_release(buf->page); buf->page = NULL; - buf->stolen = 0; + buf->flags &= ~(PIPE_BUF_FLAG_STOLEN | PIPE_BUF_FLAG_LRU); } static void *page_cache_pipe_buf_map(struct file *file, @@ -91,8 +103,7 @@ static void *page_cache_pipe_buf_map(struct file *file, static void page_cache_pipe_buf_unmap(struct pipe_inode_info *info, struct pipe_buffer *buf) { - if (!buf->stolen) - unlock_page(buf->page); + unlock_page(buf->page); kunmap(buf->page); } @@ -104,9 +115,13 @@ static struct pipe_buf_operations page_cache_pipe_buf_ops = { .steal = page_cache_pipe_buf_steal, }; +/* + * Pipe output worker. This sets up our pipe format with the page cache + * pipe buffer operations. Otherwise very similar to the regular pipe_writev(). + */ static ssize_t move_to_pipe(struct inode *inode, struct page **pages, int nr_pages, unsigned long offset, - unsigned long len) + unsigned long len, unsigned int flags) { struct pipe_inode_info *info; int ret, do_wakeup, i; @@ -159,6 +174,12 @@ static ssize_t move_to_pipe(struct inode *inode, struct page **pages, break; } + if (flags & SPLICE_F_NONBLOCK) { + if (!ret) + ret = -EAGAIN; + break; + } + if (signal_pending(current)) { if (!ret) ret = -ERESTARTSYS; @@ -191,7 +212,7 @@ static ssize_t move_to_pipe(struct inode *inode, struct page **pages, } static int __generic_file_splice_read(struct file *in, struct inode *pipe, - size_t len) + size_t len, unsigned int flags) { struct address_space *mapping = in->f_mapping; unsigned int offset, nr_pages; @@ -231,9 +252,9 @@ static int __generic_file_splice_read(struct file *in, struct inode *pipe, * fill shadow[] with pages at the right locations, so we only * have to fill holes */ - memset(shadow, 0, i * sizeof(struct page *)); - for (j = 0, pidx = index; j < i; pidx++, j++) - shadow[pages[j]->index - pidx] = pages[j]; + memset(shadow, 0, nr_pages * sizeof(struct page *)); + for (j = 0; j < i; j++) + shadow[pages[j]->index - index] = pages[j]; /* * now fill in the holes @@ -279,9 +300,19 @@ static int __generic_file_splice_read(struct file *in, struct inode *pipe, * Now we splice them into the pipe.. */ splice_them: - return move_to_pipe(pipe, pages, i, offset, len); + return move_to_pipe(pipe, pages, i, offset, len, flags); } +/** + * generic_file_splice_read - splice data from file to a pipe + * @in: file to splice from + * @pipe: pipe to splice to + * @len: number of bytes to splice + * @flags: splice modifier flags + * + * Will read pages from given file and fill them into a pipe. + * + */ ssize_t generic_file_splice_read(struct file *in, struct inode *pipe, size_t len, unsigned int flags) { @@ -291,7 +322,7 @@ ssize_t generic_file_splice_read(struct file *in, struct inode *pipe, ret = 0; spliced = 0; while (len) { - ret = __generic_file_splice_read(in, pipe, len); + ret = __generic_file_splice_read(in, pipe, len, flags); if (ret <= 0) break; @@ -299,6 +330,11 @@ ssize_t generic_file_splice_read(struct file *in, struct inode *pipe, in->f_pos += ret; len -= ret; spliced += ret; + + if (!(flags & SPLICE_F_NONBLOCK)) + continue; + ret = -EAGAIN; + break; } if (spliced) @@ -307,8 +343,11 @@ ssize_t generic_file_splice_read(struct file *in, struct inode *pipe, return ret; } +EXPORT_SYMBOL(generic_file_splice_read); + /* - * Send 'len' bytes to socket from 'file' at position 'pos' using sendpage(). + * Send 'sd->len' bytes to socket from 'sd->file' at position 'sd->pos' + * using sendpage(). */ static int pipe_to_sendpage(struct pipe_inode_info *info, struct pipe_buffer *buf, struct splice_desc *sd) @@ -318,6 +357,7 @@ static int pipe_to_sendpage(struct pipe_inode_info *info, unsigned int offset; ssize_t ret; void *ptr; + int more; /* * sub-optimal, but we are limited by the pipe ->map. we don't @@ -330,9 +370,9 @@ static int pipe_to_sendpage(struct pipe_inode_info *info, return PTR_ERR(ptr); offset = pos & ~PAGE_CACHE_MASK; + more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len; - ret = file->f_op->sendpage(file, buf->page, offset, sd->len, &pos, - sd->len < sd->total_len); + ret = file->f_op->sendpage(file, buf->page, offset, sd->len, &pos,more); buf->ops->unmap(info, buf); if (ret == sd->len) @@ -354,16 +394,19 @@ static int pipe_to_sendpage(struct pipe_inode_info *info, * - Destination page does not exist, we can add the pipe page to * the page cache and avoid the copy. * - * For now we just do the slower thing and always copy pages over, it's - * easier than migrating pages from the pipe to the target file. For the - * case of doing file | file splicing, the migrate approach had some LRU - * nastiness... + * If asked to move pages to the output file (SPLICE_F_MOVE is set in + * sd->flags), we attempt to migrate pages from the pipe to the output + * file address space page cache. This is possible if no one else has + * the pipe page referenced outside of the pipe and page cache. If + * SPLICE_F_MOVE isn't set, or we cannot move the page, we simply create + * a new page in the output file page cache and fill/dirty that. */ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf, struct splice_desc *sd) { struct file *file = sd->file; struct address_space *mapping = file->f_mapping; + gfp_t gfp_mask = mapping_gfp_mask(mapping); unsigned int offset; struct page *page; pgoff_t index; @@ -384,18 +427,23 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf, * reuse buf page, if SPLICE_F_MOVE is set */ if (sd->flags & SPLICE_F_MOVE) { + /* + * If steal succeeds, buf->page is now pruned from the vm + * side (LRU and page cache) and we can reuse it. + */ if (buf->ops->steal(info, buf)) goto find_page; page = buf->page; - if (add_to_page_cache_lru(page, mapping, index, - mapping_gfp_mask(mapping))) + if (add_to_page_cache(page, mapping, index, gfp_mask)) goto find_page; + + if (!(buf->flags & PIPE_BUF_FLAG_LRU)) + lru_cache_add(page); } else { find_page: ret = -ENOMEM; - page = find_or_create_page(mapping, index, - mapping_gfp_mask(mapping)); + page = find_or_create_page(mapping, index, gfp_mask); if (!page) goto out; @@ -432,10 +480,13 @@ find_page: } ret = mapping->a_ops->prepare_write(file, page, 0, sd->len); - if (ret) + if (ret == AOP_TRUNCATED_PAGE) { + page_cache_release(page); + goto find_page; + } else if (ret) goto out; - if (!buf->stolen) { + if (!(buf->flags & PIPE_BUF_FLAG_STOLEN)) { char *dst = kmap_atomic(page, KM_USER0); memcpy(dst + offset, src + buf->offset, sd->len); @@ -444,16 +495,18 @@ find_page: } ret = mapping->a_ops->commit_write(file, page, 0, sd->len); - if (ret < 0) + if (ret == AOP_TRUNCATED_PAGE) { + page_cache_release(page); + goto find_page; + } else if (ret) goto out; - set_page_dirty(page); - ret = write_one_page(page, 0); + balance_dirty_pages_ratelimited(mapping); out: - if (ret < 0) - unlock_page(page); - if (!buf->stolen) + if (!(buf->flags & PIPE_BUF_FLAG_STOLEN)) { page_cache_release(page); + unlock_page(page); + } buf->ops->unmap(info, buf); return ret; } @@ -461,6 +514,11 @@ out: typedef int (splice_actor)(struct pipe_inode_info *, struct pipe_buffer *, struct splice_desc *); +/* + * Pipe input worker. Most of this logic works like a regular pipe, the + * key here is the 'actor' worker passed in that actually moves the data + * to the wanted destination. See pipe_to_file/pipe_to_sendpage above. + */ static ssize_t move_from_pipe(struct inode *inode, struct file *out, size_t len, unsigned int flags, splice_actor *actor) @@ -527,6 +585,12 @@ static ssize_t move_from_pipe(struct inode *inode, struct file *out, break; } + if (flags & SPLICE_F_NONBLOCK) { + if (!ret) + ret = -EAGAIN; + break; + } + if (signal_pending(current)) { if (!ret) ret = -ERESTARTSYS; @@ -556,21 +620,67 @@ static ssize_t move_from_pipe(struct inode *inode, struct file *out, } +/** + * generic_file_splice_write - splice data from a pipe to a file + * @inode: pipe inode + * @out: file to write to + * @len: number of bytes to splice + * @flags: splice modifier flags + * + * Will either move or copy pages (determined by @flags options) from + * the given pipe inode to the given file. + * + */ ssize_t generic_file_splice_write(struct inode *inode, struct file *out, size_t len, unsigned int flags) { - return move_from_pipe(inode, out, len, flags, pipe_to_file); + struct address_space *mapping = out->f_mapping; + ssize_t ret = move_from_pipe(inode, out, len, flags, pipe_to_file); + + /* + * if file or inode is SYNC and we actually wrote some data, sync it + */ + if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(mapping->host)) + && ret > 0) { + struct inode *inode = mapping->host; + int err; + + mutex_lock(&inode->i_mutex); + err = generic_osync_inode(mapping->host, mapping, + OSYNC_METADATA|OSYNC_DATA); + mutex_unlock(&inode->i_mutex); + + if (err) + ret = err; + } + + return ret; } +EXPORT_SYMBOL(generic_file_splice_write); + +/** + * generic_splice_sendpage - splice data from a pipe to a socket + * @inode: pipe inode + * @out: socket to write to + * @len: number of bytes to splice + * @flags: splice modifier flags + * + * Will send @len bytes from the pipe to a network socket. No data copying + * is involved. + * + */ ssize_t generic_splice_sendpage(struct inode *inode, struct file *out, size_t len, unsigned int flags) { return move_from_pipe(inode, out, len, flags, pipe_to_sendpage); } -EXPORT_SYMBOL(generic_file_splice_write); -EXPORT_SYMBOL(generic_file_splice_read); +EXPORT_SYMBOL(generic_splice_sendpage); +/* + * Attempt to initiate a splice from pipe to file. + */ static long do_splice_from(struct inode *pipe, struct file *out, size_t len, unsigned int flags) { @@ -591,6 +701,9 @@ static long do_splice_from(struct inode *pipe, struct file *out, size_t len, return out->f_op->splice_write(pipe, out, len, flags); } +/* + * Attempt to initiate a splice from a file to a pipe. + */ static long do_splice_to(struct file *in, struct inode *pipe, size_t len, unsigned int flags) { @@ -619,6 +732,9 @@ static long do_splice_to(struct file *in, struct inode *pipe, size_t len, return in->f_op->splice_read(in, pipe, len, flags); } +/* + * Determine where to splice to/from. + */ static long do_splice(struct file *in, struct file *out, size_t len, unsigned int flags) { diff --git a/fs/sync.c b/fs/sync.c new file mode 100644 index 00000000000..8616006d209 --- /dev/null +++ b/fs/sync.c @@ -0,0 +1,164 @@ +/* + * High-level sync()-related operations + */ + +#include <linux/kernel.h> +#include <linux/file.h> +#include <linux/fs.h> +#include <linux/module.h> +#include <linux/writeback.h> +#include <linux/syscalls.h> +#include <linux/linkage.h> +#include <linux/pagemap.h> + +#define VALID_FLAGS (SYNC_FILE_RANGE_WAIT_BEFORE|SYNC_FILE_RANGE_WRITE| \ + SYNC_FILE_RANGE_WAIT_AFTER) + +/* + * sys_sync_file_range() permits finely controlled syncing over a segment of + * a file in the range offset .. (offset+nbytes-1) inclusive. If nbytes is + * zero then sys_sync_file_range() will operate from offset out to EOF. + * + * The flag bits are: + * + * SYNC_FILE_RANGE_WAIT_BEFORE: wait upon writeout of all pages in the range + * before performing the write. + * + * SYNC_FILE_RANGE_WRITE: initiate writeout of all those dirty pages in the + * range which are not presently under writeback. + * + * SYNC_FILE_RANGE_WAIT_AFTER: wait upon writeout of all pages in the range + * after performing the write. + * + * Useful combinations of the flag bits are: + * + * SYNC_FILE_RANGE_WAIT_BEFORE|SYNC_FILE_RANGE_WRITE: ensures that all pages + * in the range which were dirty on entry to sys_sync_file_range() are placed + * under writeout. This is a start-write-for-data-integrity operation. + * + * SYNC_FILE_RANGE_WRITE: start writeout of all dirty pages in the range which + * are not presently under writeout. This is an asynchronous flush-to-disk + * operation. Not suitable for data integrity operations. + * + * SYNC_FILE_RANGE_WAIT_BEFORE (or SYNC_FILE_RANGE_WAIT_AFTER): wait for + * completion of writeout of all pages in the range. This will be used after an + * earlier SYNC_FILE_RANGE_WAIT_BEFORE|SYNC_FILE_RANGE_WRITE operation to wait + * for that operation to complete and to return the result. + * + * SYNC_FILE_RANGE_WAIT_BEFORE|SYNC_FILE_RANGE_WRITE|SYNC_FILE_RANGE_WAIT_AFTER: + * a traditional sync() operation. This is a write-for-data-integrity operation + * which will ensure that all pages in the range which were dirty on entry to + * sys_sync_file_range() are committed to disk. + * + * + * SYNC_FILE_RANGE_WAIT_BEFORE and SYNC_FILE_RANGE_WAIT_AFTER will detect any + * I/O errors or ENOSPC conditions and will return those to the caller, after + * clearing the EIO and ENOSPC flags in the address_space. + * + * It should be noted that none of these operations write out the file's + * metadata. So unless the application is strictly performing overwrites of + * already-instantiated disk blocks, there are no guarantees here that the data + * will be available after a crash. + */ +asmlinkage long sys_sync_file_range(int fd, loff_t offset, loff_t nbytes, + int flags) +{ + int ret; + struct file *file; + loff_t endbyte; /* inclusive */ + int fput_needed; + umode_t i_mode; + + ret = -EINVAL; + if (flags & ~VALID_FLAGS) + goto out; + + endbyte = offset + nbytes; + + if ((s64)offset < 0) + goto out; + if ((s64)endbyte < 0) + goto out; + if (endbyte < offset) + goto out; + + if (sizeof(pgoff_t) == 4) { + if (offset >= (0x100000000ULL << PAGE_CACHE_SHIFT)) { + /* + * The range starts outside a 32 bit machine's + * pagecache addressing capabilities. Let it "succeed" + */ + ret = 0; + goto out; + } + if (endbyte >= (0x100000000ULL << PAGE_CACHE_SHIFT)) { + /* + * Out to EOF + */ + nbytes = 0; + } + } + + if (nbytes == 0) + endbyte = -1; + else + endbyte--; /* inclusive */ + + ret = -EBADF; + file = fget_light(fd, &fput_needed); + if (!file) + goto out; + + i_mode = file->f_dentry->d_inode->i_mode; + ret = -ESPIPE; + if (!S_ISREG(i_mode) && !S_ISBLK(i_mode) && !S_ISDIR(i_mode) && + !S_ISLNK(i_mode)) + goto out_put; + + ret = do_sync_file_range(file, offset, endbyte, flags); +out_put: + fput_light(file, fput_needed); +out: + return ret; +} + +/* + * `endbyte' is inclusive + */ +int do_sync_file_range(struct file *file, loff_t offset, loff_t endbyte, + int flags) +{ + int ret; + struct address_space *mapping; + + mapping = file->f_mapping; + if (!mapping) { + ret = -EINVAL; + goto out; + } + + ret = 0; + if (flags & SYNC_FILE_RANGE_WAIT_BEFORE) { + ret = wait_on_page_writeback_range(mapping, + offset >> PAGE_CACHE_SHIFT, + endbyte >> PAGE_CACHE_SHIFT); + if (ret < 0) + goto out; + } + + if (flags & SYNC_FILE_RANGE_WRITE) { + ret = __filemap_fdatawrite_range(mapping, offset, endbyte, + WB_SYNC_NONE); + if (ret < 0) + goto out; + } + + if (flags & SYNC_FILE_RANGE_WAIT_AFTER) { + ret = wait_on_page_writeback_range(mapping, + offset >> PAGE_CACHE_SHIFT, + endbyte >> PAGE_CACHE_SHIFT); + } +out: + return ret; +} +EXPORT_SYMBOL_GPL(do_sync_file_range); diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index f26880a4785..6cfdc9a8777 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -50,7 +50,7 @@ static struct sysfs_dirent * sysfs_new_dirent(struct sysfs_dirent * parent_sd, return sd; } -/** +/* * * Return -EEXIST if there is already a sysfs element with the same name for * the same parent. diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 830f76fa098..f1cb1ddde51 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -183,7 +183,7 @@ fill_write_buffer(struct sysfs_buffer * buffer, const char __user * buf, size_t return -ENOMEM; if (count >= PAGE_SIZE) - count = PAGE_SIZE; + count = PAGE_SIZE - 1; error = copy_from_user(buffer->page,buf,count); buffer->needs_read_fill = 1; return error ? -EFAULT : count; diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index 4c29ac41ac3..f0b347bd12c 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -175,8 +175,7 @@ const unsigned char * sysfs_get_name(struct sysfs_dirent *sd) struct bin_attribute * bin_attr; struct sysfs_symlink * sl; - if (!sd || !sd->s_element) - BUG(); + BUG_ON(!sd || !sd->s_element); switch (sd->s_type) { case SYSFS_DIR: diff --git a/fs/sysv/dir.c b/fs/sysv/dir.c index 8c66e9270dd..d7074341ee8 100644 --- a/fs/sysv/dir.c +++ b/fs/sysv/dir.c @@ -253,8 +253,7 @@ int sysv_delete_entry(struct sysv_dir_entry *de, struct page *page) lock_page(page); err = mapping->a_ops->prepare_write(NULL, page, from, to); - if (err) - BUG(); + BUG_ON(err); de->inode = 0; err = dir_commit_chunk(page, from, to); dir_put_page(page); @@ -353,8 +352,7 @@ void sysv_set_link(struct sysv_dir_entry *de, struct page *page, lock_page(page); err = page->mapping->a_ops->prepare_write(NULL, page, from, to); - if (err) - BUG(); + BUG_ON(err); de->inode = cpu_to_fs16(SYSV_SB(inode->i_sb), inode->i_ino); err = dir_commit_chunk(page, from, to); dir_put_page(page); diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 81e0e8459af..2983afd5e7f 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -312,12 +312,10 @@ static int udf_get_block(struct inode *inode, sector_t block, struct buffer_head err = 0; bh = inode_getblk(inode, block, &err, &phys, &new); - if (bh) - BUG(); + BUG_ON(bh); if (err) goto abort; - if (!phys) - BUG(); + BUG_ON(!phys); if (new) set_buffer_new(bh_result); diff --git a/fs/vfat/namei.c b/fs/vfat/namei.c index ef46939c0c1..a56cec3be5f 100644 --- a/fs/vfat/namei.c +++ b/fs/vfat/namei.c @@ -185,24 +185,6 @@ static int vfat_valid_longname(const unsigned char *name, unsigned int len) return -EINVAL; if (len >= 256) return -ENAMETOOLONG; - - /* MS-DOS "device special files" */ - if (len == 3 || (len > 3 && name[3] == '.')) { /* basename == 3 */ - if (!strnicmp(name, "aux", 3) || - !strnicmp(name, "con", 3) || - !strnicmp(name, "nul", 3) || - !strnicmp(name, "prn", 3)) - return -EINVAL; - } - if (len == 4 || (len > 4 && name[4] == '.')) { /* basename == 4 */ - /* "com1", "com2", ... */ - if ('1' <= name[3] && name[3] <= '9') { - if (!strnicmp(name, "com", 3) || - !strnicmp(name, "lpt", 3)) - return -EINVAL; - } - } - return 0; } diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index 85997b1205f..ae4c4754ed3 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -69,7 +69,6 @@ __xfs_file_read( return rval; } - STATIC ssize_t xfs_file_aio_read( struct kiocb *iocb, @@ -90,7 +89,6 @@ xfs_file_aio_read_invis( return __xfs_file_read(iocb, buf, IO_ISAIO|IO_INVIS, count, pos); } - STATIC inline ssize_t __xfs_file_write( struct kiocb *iocb, @@ -113,7 +111,6 @@ __xfs_file_write( return rval; } - STATIC ssize_t xfs_file_aio_write( struct kiocb *iocb, @@ -134,7 +131,6 @@ xfs_file_aio_write_invis( return __xfs_file_write(iocb, buf, IO_ISAIO|IO_INVIS, count, pos); } - STATIC inline ssize_t __xfs_file_readv( struct file *file, @@ -179,7 +175,6 @@ xfs_file_readv_invis( return __xfs_file_readv(file, iov, IO_INVIS, nr_segs, ppos); } - STATIC inline ssize_t __xfs_file_writev( struct file *file, @@ -204,7 +199,6 @@ __xfs_file_writev( return rval; } - STATIC ssize_t xfs_file_writev( struct file *file, @@ -228,7 +222,7 @@ xfs_file_writev_invis( STATIC ssize_t xfs_file_sendfile( struct file *filp, - loff_t *ppos, + loff_t *pos, size_t count, read_actor_t actor, void *target) @@ -236,10 +230,80 @@ xfs_file_sendfile( vnode_t *vp = vn_from_inode(filp->f_dentry->d_inode); ssize_t rval; - VOP_SENDFILE(vp, filp, ppos, 0, count, actor, target, NULL, rval); + VOP_SENDFILE(vp, filp, pos, 0, count, actor, target, NULL, rval); return rval; } +STATIC ssize_t +xfs_file_sendfile_invis( + struct file *filp, + loff_t *pos, + size_t count, + read_actor_t actor, + void *target) +{ + vnode_t *vp = vn_from_inode(filp->f_dentry->d_inode); + ssize_t rval; + + VOP_SENDFILE(vp, filp, pos, IO_INVIS, count, actor, target, NULL, rval); + return rval; +} + +STATIC ssize_t +xfs_file_splice_read( + struct file *infilp, + struct inode *pipe, + size_t len, + unsigned int flags) +{ + vnode_t *vp = vn_from_inode(infilp->f_dentry->d_inode); + ssize_t rval; + + VOP_SPLICE_READ(vp, infilp, pipe, len, flags, 0, NULL, rval); + return rval; +} + +STATIC ssize_t +xfs_file_splice_read_invis( + struct file *infilp, + struct inode *pipe, + size_t len, + unsigned int flags) +{ + vnode_t *vp = vn_from_inode(infilp->f_dentry->d_inode); + ssize_t rval; + + VOP_SPLICE_READ(vp, infilp, pipe, len, flags, IO_INVIS, NULL, rval); + return rval; +} + +STATIC ssize_t +xfs_file_splice_write( + struct inode *pipe, + struct file *outfilp, + size_t len, + unsigned int flags) +{ + vnode_t *vp = vn_from_inode(outfilp->f_dentry->d_inode); + ssize_t rval; + + VOP_SPLICE_WRITE(vp, pipe, outfilp, len, flags, 0, NULL, rval); + return rval; +} + +STATIC ssize_t +xfs_file_splice_write_invis( + struct inode *pipe, + struct file *outfilp, + size_t len, + unsigned int flags) +{ + vnode_t *vp = vn_from_inode(outfilp->f_dentry->d_inode); + ssize_t rval; + + VOP_SPLICE_WRITE(vp, pipe, outfilp, len, flags, IO_INVIS, NULL, rval); + return rval; +} STATIC int xfs_file_open( @@ -251,13 +315,10 @@ xfs_file_open( if (!(filp->f_flags & O_LARGEFILE) && i_size_read(inode) > MAX_NON_LFS) return -EFBIG; - - ASSERT(vp); VOP_OPEN(vp, NULL, error); return -error; } - STATIC int xfs_file_release( struct inode *inode, @@ -271,7 +332,6 @@ xfs_file_release( return -error; } - STATIC int xfs_file_fsync( struct file *filp, @@ -285,21 +345,11 @@ xfs_file_fsync( if (datasync) flags |= FSYNC_DATA; - - ASSERT(vp); VOP_FSYNC(vp, flags, NULL, (xfs_off_t)0, (xfs_off_t)-1, error); return -error; } -/* - * xfs_file_readdir maps to VOP_READDIR(). - * We need to build a uio, cred, ... - */ - -#define nextdp(dp) ((struct xfs_dirent *)((char *)(dp) + (dp)->d_reclen)) - #ifdef CONFIG_XFS_DMAPI - STATIC struct page * xfs_vm_nopage( struct vm_area_struct *area, @@ -319,10 +369,8 @@ xfs_vm_nopage( return filemap_nopage(area, address, type); } - #endif /* CONFIG_XFS_DMAPI */ - STATIC int xfs_file_readdir( struct file *filp, @@ -330,7 +378,7 @@ xfs_file_readdir( filldir_t filldir) { int error = 0; - vnode_t *vp; + vnode_t *vp = vn_from_inode(filp->f_dentry->d_inode); uio_t uio; iovec_t iov; int eof = 0; @@ -340,9 +388,6 @@ xfs_file_readdir( xfs_off_t start_offset, curr_offset; xfs_dirent_t *dbp = NULL; - vp = vn_from_inode(filp->f_dentry->d_inode); - ASSERT(vp); - /* Try fairly hard to get memory */ do { if ((read_buf = (caddr_t)kmalloc(rlen, GFP_KERNEL))) @@ -387,7 +432,7 @@ xfs_file_readdir( } size -= dbp->d_reclen; curr_offset = (loff_t)dbp->d_off /* & 0x7fffffff */; - dbp = nextdp(dbp); + dbp = (xfs_dirent_t *)((char *)dbp + dbp->d_reclen); } } done: @@ -402,7 +447,6 @@ done: return -error; } - STATIC int xfs_file_mmap( struct file *filp, @@ -457,11 +501,10 @@ xfs_file_ioctl_invis( unsigned int cmd, unsigned long arg) { - int error; struct inode *inode = filp->f_dentry->d_inode; vnode_t *vp = vn_from_inode(inode); + int error; - ASSERT(vp); VOP_IOCTL(vp, inode, filp, IO_INVIS, cmd, (void __user *)arg, error); VMODIFY(vp); @@ -537,6 +580,8 @@ const struct file_operations xfs_file_operations = { .aio_read = xfs_file_aio_read, .aio_write = xfs_file_aio_write, .sendfile = xfs_file_sendfile, + .splice_read = xfs_file_splice_read, + .splice_write = xfs_file_splice_write, .unlocked_ioctl = xfs_file_ioctl, #ifdef CONFIG_COMPAT .compat_ioctl = xfs_file_compat_ioctl, @@ -558,7 +603,9 @@ const struct file_operations xfs_invis_file_operations = { .writev = xfs_file_writev_invis, .aio_read = xfs_file_aio_read_invis, .aio_write = xfs_file_aio_write_invis, - .sendfile = xfs_file_sendfile, + .sendfile = xfs_file_sendfile_invis, + .splice_read = xfs_file_splice_read_invis, + .splice_write = xfs_file_splice_write_invis, .unlocked_ioctl = xfs_file_ioctl_invis, #ifdef CONFIG_COMPAT .compat_ioctl = xfs_file_compat_invis_ioctl, diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h index 1fe09f2d651..e9fe43d7476 100644 --- a/fs/xfs/linux-2.6/xfs_linux.h +++ b/fs/xfs/linux-2.6/xfs_linux.h @@ -103,6 +103,7 @@ */ #undef HAVE_REFCACHE /* reference cache not needed for NFS in 2.6 */ #define HAVE_SENDFILE /* sendfile(2) exists in 2.6, but not in 2.4 */ +#define HAVE_SPLICE /* a splice(2) exists in 2.6, but not in 2.4 */ #ifdef CONFIG_SMP #define HAVE_PERCPU_SB /* per cpu superblock counters are a 2.6 feature */ #else diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c index 84ddf189389..90cd314acba 100644 --- a/fs/xfs/linux-2.6/xfs_lrw.c +++ b/fs/xfs/linux-2.6/xfs_lrw.c @@ -301,36 +301,23 @@ xfs_sendfile( void *target, cred_t *credp) { + xfs_inode_t *ip = XFS_BHVTOI(bdp); + xfs_mount_t *mp = ip->i_mount; ssize_t ret; - xfs_fsize_t n; - xfs_inode_t *ip; - xfs_mount_t *mp; - vnode_t *vp; - - ip = XFS_BHVTOI(bdp); - vp = BHV_TO_VNODE(bdp); - mp = ip->i_mount; XFS_STATS_INC(xs_read_calls); - - n = XFS_MAXIOFFSET(mp) - *offset; - if ((n <= 0) || (count == 0)) - return 0; - - if (n < count) - count = n; - - if (XFS_FORCED_SHUTDOWN(ip->i_mount)) + if (XFS_FORCED_SHUTDOWN(mp)) return -EIO; xfs_ilock(ip, XFS_IOLOCK_SHARED); - if (DM_EVENT_ENABLED(vp->v_vfsp, ip, DM_EVENT_READ) && + if (DM_EVENT_ENABLED(BHV_TO_VNODE(bdp)->v_vfsp, ip, DM_EVENT_READ) && (!(ioflags & IO_INVIS))) { vrwlock_t locktype = VRWLOCK_READ; int error; - error = XFS_SEND_DATA(mp, DM_EVENT_READ, BHV_TO_VNODE(bdp), *offset, count, + error = XFS_SEND_DATA(mp, DM_EVENT_READ, BHV_TO_VNODE(bdp), + *offset, count, FILP_DELAY_FLAG(filp), &locktype); if (error) { xfs_iunlock(ip, XFS_IOLOCK_SHARED); @@ -340,12 +327,96 @@ xfs_sendfile( xfs_rw_enter_trace(XFS_SENDFILE_ENTER, &ip->i_iocore, (void *)(unsigned long)target, count, *offset, ioflags); ret = generic_file_sendfile(filp, offset, count, actor, target); + if (ret > 0) + XFS_STATS_ADD(xs_read_bytes, ret); xfs_iunlock(ip, XFS_IOLOCK_SHARED); + return ret; +} +ssize_t +xfs_splice_read( + bhv_desc_t *bdp, + struct file *infilp, + struct inode *pipe, + size_t count, + int flags, + int ioflags, + cred_t *credp) +{ + xfs_inode_t *ip = XFS_BHVTOI(bdp); + xfs_mount_t *mp = ip->i_mount; + ssize_t ret; + + XFS_STATS_INC(xs_read_calls); + if (XFS_FORCED_SHUTDOWN(ip->i_mount)) + return -EIO; + + xfs_ilock(ip, XFS_IOLOCK_SHARED); + + if (DM_EVENT_ENABLED(BHV_TO_VNODE(bdp)->v_vfsp, ip, DM_EVENT_READ) && + (!(ioflags & IO_INVIS))) { + vrwlock_t locktype = VRWLOCK_READ; + int error; + + error = XFS_SEND_DATA(mp, DM_EVENT_READ, BHV_TO_VNODE(bdp), + infilp->f_pos, count, + FILP_DELAY_FLAG(infilp), &locktype); + if (error) { + xfs_iunlock(ip, XFS_IOLOCK_SHARED); + return -error; + } + } + xfs_rw_enter_trace(XFS_SPLICE_READ_ENTER, &ip->i_iocore, + pipe, count, infilp->f_pos, ioflags); + ret = generic_file_splice_read(infilp, pipe, count, flags); if (ret > 0) XFS_STATS_ADD(xs_read_bytes, ret); + xfs_iunlock(ip, XFS_IOLOCK_SHARED); + return ret; +} + +ssize_t +xfs_splice_write( + bhv_desc_t *bdp, + struct inode *pipe, + struct file *outfilp, + size_t count, + int flags, + int ioflags, + cred_t *credp) +{ + xfs_inode_t *ip = XFS_BHVTOI(bdp); + xfs_mount_t *mp = ip->i_mount; + ssize_t ret; + + XFS_STATS_INC(xs_write_calls); + if (XFS_FORCED_SHUTDOWN(ip->i_mount)) + return -EIO; + + xfs_ilock(ip, XFS_IOLOCK_EXCL); + + if (DM_EVENT_ENABLED(BHV_TO_VNODE(bdp)->v_vfsp, ip, DM_EVENT_WRITE) && + (!(ioflags & IO_INVIS))) { + vrwlock_t locktype = VRWLOCK_WRITE; + int error; + + error = XFS_SEND_DATA(mp, DM_EVENT_WRITE, BHV_TO_VNODE(bdp), + outfilp->f_pos, count, + FILP_DELAY_FLAG(outfilp), &locktype); + if (error) { + xfs_iunlock(ip, XFS_IOLOCK_EXCL); + return -error; + } + } + xfs_rw_enter_trace(XFS_SPLICE_WRITE_ENTER, &ip->i_iocore, + pipe, count, outfilp->f_pos, ioflags); + ret = generic_file_splice_write(pipe, outfilp, count, flags); + if (ret > 0) + XFS_STATS_ADD(xs_write_bytes, ret); + + xfs_iunlock(ip, XFS_IOLOCK_EXCL); return ret; } @@ -363,7 +434,7 @@ xfs_zero_last_block( xfs_fsize_t end_size) { xfs_fileoff_t last_fsb; - xfs_mount_t *mp; + xfs_mount_t *mp = io->io_mount; int nimaps; int zero_offset; int zero_len; @@ -373,8 +444,6 @@ xfs_zero_last_block( ASSERT(ismrlocked(io->io_lock, MR_UPDATE) != 0); - mp = io->io_mount; - zero_offset = XFS_B_FSB_OFFSET(mp, isize); if (zero_offset == 0) { /* @@ -405,10 +474,9 @@ xfs_zero_last_block( * don't deadlock when the buffer cache calls back to us. */ XFS_IUNLOCK(mp, io, XFS_ILOCK_EXCL| XFS_EXTSIZE_RD); - loff = XFS_FSB_TO_B(mp, last_fsb); + loff = XFS_FSB_TO_B(mp, last_fsb); zero_len = mp->m_sb.sb_blocksize - zero_offset; - error = xfs_iozero(ip, loff + zero_offset, zero_len, end_size); XFS_ILOCK(mp, io, XFS_ILOCK_EXCL|XFS_EXTSIZE_RD); @@ -441,7 +509,7 @@ xfs_zero_eof( xfs_fileoff_t zero_count_fsb; xfs_fileoff_t last_fsb; xfs_extlen_t buf_len_fsb; - xfs_mount_t *mp; + xfs_mount_t *mp = io->io_mount; int nimaps; int error = 0; xfs_bmbt_irec_t imap; @@ -450,8 +518,6 @@ xfs_zero_eof( ASSERT(ismrlocked(io->io_iolock, MR_UPDATE)); ASSERT(offset > isize); - mp = io->io_mount; - /* * First handle zeroing the block on which isize resides. * We only zero a part of that block so it is handled specially. diff --git a/fs/xfs/linux-2.6/xfs_lrw.h b/fs/xfs/linux-2.6/xfs_lrw.h index 38864a88d42..eaa5659713f 100644 --- a/fs/xfs/linux-2.6/xfs_lrw.h +++ b/fs/xfs/linux-2.6/xfs_lrw.h @@ -60,6 +60,8 @@ struct xfs_iomap; #define XFS_IOMAP_ALLOC_ENTER 25 #define XFS_IOMAP_ALLOC_MAP 26 #define XFS_IOMAP_UNWRITTEN 27 +#define XFS_SPLICE_READ_ENTER 28 +#define XFS_SPLICE_WRITE_ENTER 29 extern void xfs_rw_enter_trace(int, struct xfs_iocore *, void *, size_t, loff_t, int); extern void xfs_inval_cached_trace(struct xfs_iocore *, @@ -78,6 +80,7 @@ extern int xfs_bmap(struct bhv_desc *, xfs_off_t, ssize_t, int, struct xfs_iomap *, int *); extern int xfsbdstrat(struct xfs_mount *, struct xfs_buf *); extern int xfs_bdstrat_cb(struct xfs_buf *); +extern int xfs_dev_is_read_only(struct xfs_mount *, char *); extern int xfs_zero_eof(struct vnode *, struct xfs_iocore *, xfs_off_t, xfs_fsize_t, xfs_fsize_t); @@ -90,7 +93,11 @@ extern ssize_t xfs_write(struct bhv_desc *, struct kiocb *, extern ssize_t xfs_sendfile(struct bhv_desc *, struct file *, loff_t *, int, size_t, read_actor_t, void *, struct cred *); - -extern int xfs_dev_is_read_only(struct xfs_mount *, char *); +extern ssize_t xfs_splice_read(struct bhv_desc *, struct file *, + struct inode *, size_t, int, int, + struct cred *); +extern ssize_t xfs_splice_write(struct bhv_desc *, struct inode *, + struct file *, size_t, int, int, + struct cred *); #endif /* __XFS_LRW_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 1884300417e..68f4793e8a1 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -67,7 +67,8 @@ mempool_t *xfs_ioend_pool; STATIC struct xfs_mount_args * xfs_args_allocate( - struct super_block *sb) + struct super_block *sb, + int silent) { struct xfs_mount_args *args; @@ -80,8 +81,8 @@ xfs_args_allocate( args->flags |= XFSMNT_DIRSYNC; if (sb->s_flags & MS_SYNCHRONOUS) args->flags |= XFSMNT_WSYNC; - - /* Default to 32 bit inodes on Linux all the time */ + if (silent) + args->flags |= XFSMNT_QUIET; args->flags |= XFSMNT_32BITINODES; return args; @@ -719,7 +720,7 @@ xfs_fs_remount( char *options) { vfs_t *vfsp = vfs_from_sb(sb); - struct xfs_mount_args *args = xfs_args_allocate(sb); + struct xfs_mount_args *args = xfs_args_allocate(sb, 0); int error; VFS_PARSEARGS(vfsp, options, args, 1, error); @@ -825,7 +826,7 @@ xfs_fs_fill_super( { vnode_t *rootvp; struct vfs *vfsp = vfs_allocate(sb); - struct xfs_mount_args *args = xfs_args_allocate(sb); + struct xfs_mount_args *args = xfs_args_allocate(sb, silent); struct kstatfs statvfs; int error, error2; diff --git a/fs/xfs/linux-2.6/xfs_vnode.h b/fs/xfs/linux-2.6/xfs_vnode.h index 06f5845e956..6f1c79a28f8 100644 --- a/fs/xfs/linux-2.6/xfs_vnode.h +++ b/fs/xfs/linux-2.6/xfs_vnode.h @@ -173,6 +173,12 @@ typedef ssize_t (*vop_write_t)(bhv_desc_t *, struct kiocb *, typedef ssize_t (*vop_sendfile_t)(bhv_desc_t *, struct file *, loff_t *, int, size_t, read_actor_t, void *, struct cred *); +typedef ssize_t (*vop_splice_read_t)(bhv_desc_t *, struct file *, + struct inode *, size_t, int, int, + struct cred *); +typedef ssize_t (*vop_splice_write_t)(bhv_desc_t *, struct inode *, + struct file *, size_t, int, int, + struct cred *); typedef int (*vop_ioctl_t)(bhv_desc_t *, struct inode *, struct file *, int, unsigned int, void __user *); typedef int (*vop_getattr_t)(bhv_desc_t *, struct vattr *, int, @@ -231,6 +237,8 @@ typedef struct vnodeops { vop_read_t vop_read; vop_write_t vop_write; vop_sendfile_t vop_sendfile; + vop_splice_read_t vop_splice_read; + vop_splice_write_t vop_splice_write; vop_ioctl_t vop_ioctl; vop_getattr_t vop_getattr; vop_setattr_t vop_setattr; @@ -276,6 +284,10 @@ typedef struct vnodeops { rv = _VOP_(vop_write, vp)((vp)->v_fbhv,file,iov,segs,offset,ioflags,cr) #define VOP_SENDFILE(vp,f,off,ioflags,cnt,act,targ,cr,rv) \ rv = _VOP_(vop_sendfile, vp)((vp)->v_fbhv,f,off,ioflags,cnt,act,targ,cr) +#define VOP_SPLICE_READ(vp,f,pipe,cnt,fl,iofl,cr,rv) \ + rv = _VOP_(vop_splice_read, vp)((vp)->v_fbhv,f,pipe,cnt,fl,iofl,cr) +#define VOP_SPLICE_WRITE(vp,f,pipe,cnt,fl,iofl,cr,rv) \ + rv = _VOP_(vop_splice_write, vp)((vp)->v_fbhv,f,pipe,cnt,fl,iofl,cr) #define VOP_BMAP(vp,of,sz,rw,b,n,rv) \ rv = _VOP_(vop_bmap, vp)((vp)->v_fbhv,of,sz,rw,b,n) #define VOP_OPEN(vp, cr, rv) \ diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index 73c1e5e80c0..7fb5eca9bd5 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -2624,7 +2624,7 @@ xfs_qm_vop_chown_reserve( { int error; xfs_mount_t *mp; - uint delblks, blkflags; + uint delblks, blkflags, prjflags = 0; xfs_dquot_t *unresudq, *unresgdq, *delblksudq, *delblksgdq; ASSERT(XFS_ISLOCKED_INODE(ip)); @@ -2650,10 +2650,13 @@ xfs_qm_vop_chown_reserve( } } if (XFS_IS_OQUOTA_ON(ip->i_mount) && gdqp) { - if ((XFS_IS_GQUOTA_ON(ip->i_mount) && - ip->i_d.di_gid != be32_to_cpu(gdqp->q_core.d_id)) || - (XFS_IS_PQUOTA_ON(ip->i_mount) && - ip->i_d.di_projid != be32_to_cpu(gdqp->q_core.d_id))) { + if (XFS_IS_PQUOTA_ON(ip->i_mount) && + ip->i_d.di_projid != be32_to_cpu(gdqp->q_core.d_id)) + prjflags = XFS_QMOPT_ENOSPC; + + if (prjflags || + (XFS_IS_GQUOTA_ON(ip->i_mount) && + ip->i_d.di_gid != be32_to_cpu(gdqp->q_core.d_id))) { delblksgdq = gdqp; if (delblks) { ASSERT(ip->i_gdquot); @@ -2664,7 +2667,7 @@ xfs_qm_vop_chown_reserve( if ((error = xfs_trans_reserve_quota_bydquots(tp, ip->i_mount, delblksudq, delblksgdq, ip->i_d.di_nblocks, 1, - flags | blkflags))) + flags | blkflags | prjflags))) return (error); /* @@ -2681,7 +2684,7 @@ xfs_qm_vop_chown_reserve( ASSERT(unresudq || unresgdq); if ((error = xfs_trans_reserve_quota_bydquots(NULL, ip->i_mount, delblksudq, delblksgdq, (xfs_qcnt_t)delblks, 0, - flags | blkflags))) + flags | blkflags | prjflags))) return (error); xfs_trans_reserve_quota_bydquots(NULL, ip->i_mount, unresudq, unresgdq, -((xfs_qcnt_t)delblks), 0, diff --git a/fs/xfs/quota/xfs_trans_dquot.c b/fs/xfs/quota/xfs_trans_dquot.c index d8e131ec0aa..9168918db25 100644 --- a/fs/xfs/quota/xfs_trans_dquot.c +++ b/fs/xfs/quota/xfs_trans_dquot.c @@ -595,12 +595,19 @@ xfs_trans_unreserve_and_mod_dquots( } } +STATIC int +xfs_quota_error(uint flags) +{ + if (flags & XFS_QMOPT_ENOSPC) + return ENOSPC; + return EDQUOT; +} + /* * This reserves disk blocks and inodes against a dquot. * Flags indicate if the dquot is to be locked here and also * if the blk reservation is for RT or regular blocks. * Sending in XFS_QMOPT_FORCE_RES flag skips the quota check. - * Returns EDQUOT if quota is exceeded. */ STATIC int xfs_trans_dqresv( @@ -666,19 +673,15 @@ xfs_trans_dqresv( */ if (hardlimit > 0ULL && (hardlimit <= nblks + *resbcountp)) { - error = EDQUOT; + error = xfs_quota_error(flags); goto error_return; } if (softlimit > 0ULL && (softlimit <= nblks + *resbcountp)) { - /* - * If timer or warnings has expired, - * return EDQUOT - */ if ((timer != 0 && get_seconds() > timer) || (warns != 0 && warns >= warnlimit)) { - error = EDQUOT; + error = xfs_quota_error(flags); goto error_return; } } @@ -695,16 +698,12 @@ xfs_trans_dqresv( if (!softlimit) softlimit = q->qi_isoftlimit; if (hardlimit > 0ULL && count >= hardlimit) { - error = EDQUOT; + error = xfs_quota_error(flags); goto error_return; } else if (softlimit > 0ULL && count >= softlimit) { - /* - * If timer or warnings has expired, - * return EDQUOT - */ if ((timer != 0 && get_seconds() > timer) || (warns != 0 && warns >= warnlimit)) { - error = EDQUOT; + error = xfs_quota_error(flags); goto error_return; } } @@ -751,13 +750,14 @@ error_return: /* - * Given a dquot(s), make disk block and/or inode reservations against them. + * Given dquot(s), make disk block and/or inode reservations against them. * The fact that this does the reservation against both the usr and - * grp quotas is important, because this follows a both-or-nothing + * grp/prj quotas is important, because this follows a both-or-nothing * approach. * * flags = XFS_QMOPT_DQLOCK indicate if dquot(s) need to be locked. * XFS_QMOPT_FORCE_RES evades limit enforcement. Used by chown. + * XFS_QMOPT_ENOSPC returns ENOSPC not EDQUOT. Used by pquota. * XFS_TRANS_DQ_RES_BLKS reserves regular disk blocks * XFS_TRANS_DQ_RES_RTBLKS reserves realtime disk blocks * dquots are unlocked on return, if they were not locked by caller. @@ -772,25 +772,27 @@ xfs_trans_reserve_quota_bydquots( long ninos, uint flags) { - int resvd; + int resvd = 0, error; - if (! XFS_IS_QUOTA_ON(mp)) - return (0); + if (!XFS_IS_QUOTA_ON(mp)) + return 0; if (tp && tp->t_dqinfo == NULL) xfs_trans_alloc_dqinfo(tp); ASSERT(flags & XFS_QMOPT_RESBLK_MASK); - resvd = 0; if (udqp) { - if (xfs_trans_dqresv(tp, mp, udqp, nblks, ninos, flags)) - return (EDQUOT); + error = xfs_trans_dqresv(tp, mp, udqp, nblks, ninos, + (flags & ~XFS_QMOPT_ENOSPC)); + if (error) + return error; resvd = 1; } if (gdqp) { - if (xfs_trans_dqresv(tp, mp, gdqp, nblks, ninos, flags)) { + error = xfs_trans_dqresv(tp, mp, gdqp, nblks, ninos, flags); + if (error) { /* * can't do it, so backout previous reservation */ @@ -799,14 +801,14 @@ xfs_trans_reserve_quota_bydquots( xfs_trans_dqresv(tp, mp, udqp, -nblks, -ninos, flags); } - return (EDQUOT); + return error; } } /* * Didn't change anything critical, so, no need to log */ - return (0); + return 0; } @@ -814,8 +816,6 @@ xfs_trans_reserve_quota_bydquots( * Lock the dquot and change the reservation if we can. * This doesn't change the actual usage, just the reservation. * The inode sent in is locked. - * - * Returns 0 on success, EDQUOT or other errors otherwise */ STATIC int xfs_trans_reserve_quota_nblks( @@ -824,20 +824,24 @@ xfs_trans_reserve_quota_nblks( xfs_inode_t *ip, long nblks, long ninos, - uint type) + uint flags) { int error; if (!XFS_IS_QUOTA_ON(mp)) - return (0); + return 0; + if (XFS_IS_PQUOTA_ON(mp)) + flags |= XFS_QMOPT_ENOSPC; ASSERT(ip->i_ino != mp->m_sb.sb_uquotino); ASSERT(ip->i_ino != mp->m_sb.sb_gquotino); ASSERT(XFS_ISLOCKED_INODE_EXCL(ip)); ASSERT(XFS_IS_QUOTA_RUNNING(ip->i_mount)); - ASSERT((type & ~XFS_QMOPT_FORCE_RES) == XFS_TRANS_DQ_RES_RTBLKS || - (type & ~XFS_QMOPT_FORCE_RES) == XFS_TRANS_DQ_RES_BLKS); + ASSERT((flags & ~(XFS_QMOPT_FORCE_RES | XFS_QMOPT_ENOSPC)) == + XFS_TRANS_DQ_RES_RTBLKS || + (flags & ~(XFS_QMOPT_FORCE_RES | XFS_QMOPT_ENOSPC)) == + XFS_TRANS_DQ_RES_BLKS); /* * Reserve nblks against these dquots, with trans as the mediator. @@ -845,8 +849,8 @@ xfs_trans_reserve_quota_nblks( error = xfs_trans_reserve_quota_bydquots(tp, mp, ip->i_udquot, ip->i_gdquot, nblks, ninos, - type); - return (error); + flags); + return error; } /* diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index d384e489705..26939d364bc 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -4719,18 +4719,17 @@ xfs_bmapi( /* * Make a transaction-less quota reservation for * delayed allocation blocks. This number gets - * adjusted later. - * We return EDQUOT if we haven't allocated - * blks already inside this loop; + * adjusted later. We return if we haven't + * allocated blocks already inside this loop. */ - if (XFS_TRANS_RESERVE_QUOTA_NBLKS( + if ((error = XFS_TRANS_RESERVE_QUOTA_NBLKS( mp, NULL, ip, (long)alen, 0, rt ? XFS_QMOPT_RES_RTBLKS : - XFS_QMOPT_RES_REGBLKS)) { + XFS_QMOPT_RES_REGBLKS))) { if (n == 0) { *nmap = 0; ASSERT(cur == NULL); - return XFS_ERROR(EDQUOT); + return error; } break; } diff --git a/fs/xfs/xfs_bmap.h b/fs/xfs/xfs_bmap.h index f83399c89ce..8e0d73d9ccc 100644 --- a/fs/xfs/xfs_bmap.h +++ b/fs/xfs/xfs_bmap.h @@ -353,10 +353,11 @@ xfs_check_nostate_extents( xfs_extnum_t num); /* - * Call xfs_bmap_do_search_extents() to search for the extent - * record containing block bno. If in multi-level in-core extent - * allocation mode, find and extract the target extent buffer, - * otherwise just use the direct extent list. + * Search the extent records for the entry containing block bno. + * If bno lies in a hole, point to the next entry. If bno lies + * past eof, *eofp will be set, and *prevp will contain the last + * entry (null if none). Else, *lastxp will be set to the index + * of the found entry; *gotp will contain the entry. */ xfs_bmbt_rec_t * xfs_bmap_search_multi_extents(struct xfs_ifork *, xfs_fileoff_t, int *, diff --git a/fs/xfs/xfs_clnt.h b/fs/xfs/xfs_clnt.h index 022fff62085..5b7eb81453b 100644 --- a/fs/xfs/xfs_clnt.h +++ b/fs/xfs/xfs_clnt.h @@ -68,6 +68,7 @@ struct xfs_mount_args { * enforcement */ #define XFSMNT_PQUOTAENF 0x00000040 /* IRIX project quota limit * enforcement */ +#define XFSMNT_QUIET 0x00000080 /* don't report mount errors */ #define XFSMNT_NOALIGN 0x00000200 /* don't allocate at * stripe boundaries*/ #define XFSMNT_RETERR 0x00000400 /* return error to user */ diff --git a/fs/xfs/xfs_error.h b/fs/xfs/xfs_error.h index 26b8e709a56..bc43163456e 100644 --- a/fs/xfs/xfs_error.h +++ b/fs/xfs/xfs_error.h @@ -186,4 +186,7 @@ extern void xfs_fs_cmn_err(int level, struct xfs_mount *mp, char *fmt, ...); #define xfs_fs_repair_cmn_err(level, mp, fmt, args...) \ xfs_fs_cmn_err(level, mp, fmt " Unmount and run xfs_repair.", ## args) +#define xfs_fs_mount_cmn_err(f, fmt, args...) \ + ((f & XFS_MFSI_QUIET)? cmn_err(CE_WARN, "XFS: " fmt, ## args) : (void)0) + #endif /* __XFS_ERROR_H__ */ diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 72e7e78bfff..049fabb7f7e 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -213,7 +213,8 @@ xfs_mount_free( STATIC int xfs_mount_validate_sb( xfs_mount_t *mp, - xfs_sb_t *sbp) + xfs_sb_t *sbp, + int flags) { /* * If the log device and data device have the @@ -223,33 +224,29 @@ xfs_mount_validate_sb( * a volume filesystem in a non-volume manner. */ if (sbp->sb_magicnum != XFS_SB_MAGIC) { - cmn_err(CE_WARN, "XFS: bad magic number"); + xfs_fs_mount_cmn_err(flags, "bad magic number"); return XFS_ERROR(EWRONGFS); } if (!XFS_SB_GOOD_VERSION(sbp)) { - cmn_err(CE_WARN, "XFS: bad version"); + xfs_fs_mount_cmn_err(flags, "bad version"); return XFS_ERROR(EWRONGFS); } if (unlikely( sbp->sb_logstart == 0 && mp->m_logdev_targp == mp->m_ddev_targp)) { - cmn_err(CE_WARN, - "XFS: filesystem is marked as having an external log; " - "specify logdev on the\nmount command line."); - XFS_CORRUPTION_ERROR("xfs_mount_validate_sb(1)", - XFS_ERRLEVEL_HIGH, mp, sbp); - return XFS_ERROR(EFSCORRUPTED); + xfs_fs_mount_cmn_err(flags, + "filesystem is marked as having an external log; " + "specify logdev on the\nmount command line."); + return XFS_ERROR(EINVAL); } if (unlikely( sbp->sb_logstart != 0 && mp->m_logdev_targp != mp->m_ddev_targp)) { - cmn_err(CE_WARN, - "XFS: filesystem is marked as having an internal log; " - "don't specify logdev on\nthe mount command line."); - XFS_CORRUPTION_ERROR("xfs_mount_validate_sb(2)", - XFS_ERRLEVEL_HIGH, mp, sbp); - return XFS_ERROR(EFSCORRUPTED); + xfs_fs_mount_cmn_err(flags, + "filesystem is marked as having an internal log; " + "do not specify logdev on\nthe mount command line."); + return XFS_ERROR(EINVAL); } /* @@ -274,9 +271,7 @@ xfs_mount_validate_sb( (sbp->sb_rextsize * sbp->sb_blocksize > XFS_MAX_RTEXTSIZE) || (sbp->sb_rextsize * sbp->sb_blocksize < XFS_MIN_RTEXTSIZE) || (sbp->sb_imax_pct > 100 || sbp->sb_imax_pct < 1))) { - cmn_err(CE_WARN, "XFS: SB sanity check 1 failed"); - XFS_CORRUPTION_ERROR("xfs_mount_validate_sb(3)", - XFS_ERRLEVEL_LOW, mp, sbp); + xfs_fs_mount_cmn_err(flags, "SB sanity check 1 failed"); return XFS_ERROR(EFSCORRUPTED); } @@ -289,9 +284,7 @@ xfs_mount_validate_sb( (xfs_drfsbno_t)sbp->sb_agcount * sbp->sb_agblocks || sbp->sb_dblocks < (xfs_drfsbno_t)(sbp->sb_agcount - 1) * sbp->sb_agblocks + XFS_MIN_AG_BLOCKS)) { - cmn_err(CE_WARN, "XFS: SB sanity check 2 failed"); - XFS_ERROR_REPORT("xfs_mount_validate_sb(4)", - XFS_ERRLEVEL_LOW, mp); + xfs_fs_mount_cmn_err(flags, "SB sanity check 2 failed"); return XFS_ERROR(EFSCORRUPTED); } @@ -307,15 +300,13 @@ xfs_mount_validate_sb( (sbp->sb_dblocks << (sbp->sb_blocklog - BBSHIFT)) > UINT_MAX || (sbp->sb_rblocks << (sbp->sb_blocklog - BBSHIFT)) > UINT_MAX)) { #endif - cmn_err(CE_WARN, - "XFS: File system is too large to be mounted on this system."); + xfs_fs_mount_cmn_err(flags, + "file system too large to be mounted on this system."); return XFS_ERROR(E2BIG); } if (unlikely(sbp->sb_inprogress)) { - cmn_err(CE_WARN, "XFS: file system busy"); - XFS_ERROR_REPORT("xfs_mount_validate_sb(5)", - XFS_ERRLEVEL_LOW, mp); + xfs_fs_mount_cmn_err(flags, "file system busy"); return XFS_ERROR(EFSCORRUPTED); } @@ -323,8 +314,8 @@ xfs_mount_validate_sb( * Version 1 directory format has never worked on Linux. */ if (unlikely(!XFS_SB_VERSION_HASDIRV2(sbp))) { - cmn_err(CE_WARN, - "XFS: Attempted to mount file system using version 1 directory format"); + xfs_fs_mount_cmn_err(flags, + "file system using version 1 directory format"); return XFS_ERROR(ENOSYS); } @@ -332,11 +323,11 @@ xfs_mount_validate_sb( * Until this is fixed only page-sized or smaller data blocks work. */ if (unlikely(sbp->sb_blocksize > PAGE_SIZE)) { - cmn_err(CE_WARN, - "XFS: Attempted to mount file system with blocksize %d bytes", + xfs_fs_mount_cmn_err(flags, + "file system with blocksize %d bytes", sbp->sb_blocksize); - cmn_err(CE_WARN, - "XFS: Only page-sized (%ld) or less blocksizes currently work.", + xfs_fs_mount_cmn_err(flags, + "only pagesize (%ld) or less will currently work.", PAGE_SIZE); return XFS_ERROR(ENOSYS); } @@ -484,7 +475,7 @@ xfs_xlatesb( * Does the initial read of the superblock. */ int -xfs_readsb(xfs_mount_t *mp) +xfs_readsb(xfs_mount_t *mp, int flags) { unsigned int sector_size; unsigned int extra_flags; @@ -506,7 +497,7 @@ xfs_readsb(xfs_mount_t *mp) bp = xfs_buf_read_flags(mp->m_ddev_targp, XFS_SB_DADDR, BTOBB(sector_size), extra_flags); if (!bp || XFS_BUF_ISERROR(bp)) { - cmn_err(CE_WARN, "XFS: SB read failed"); + xfs_fs_mount_cmn_err(flags, "SB read failed"); error = bp ? XFS_BUF_GETERROR(bp) : ENOMEM; goto fail; } @@ -520,9 +511,9 @@ xfs_readsb(xfs_mount_t *mp) sbp = XFS_BUF_TO_SBP(bp); xfs_xlatesb(XFS_BUF_PTR(bp), &(mp->m_sb), 1, XFS_SB_ALL_BITS); - error = xfs_mount_validate_sb(mp, &(mp->m_sb)); + error = xfs_mount_validate_sb(mp, &(mp->m_sb), flags); if (error) { - cmn_err(CE_WARN, "XFS: SB validate failed"); + xfs_fs_mount_cmn_err(flags, "SB validate failed"); goto fail; } @@ -530,8 +521,8 @@ xfs_readsb(xfs_mount_t *mp) * We must be able to do sector-sized and sector-aligned IO. */ if (sector_size > mp->m_sb.sb_sectsize) { - cmn_err(CE_WARN, - "XFS: device supports only %u byte sectors (not %u)", + xfs_fs_mount_cmn_err(flags, + "device supports only %u byte sectors (not %u)", sector_size, mp->m_sb.sb_sectsize); error = ENOSYS; goto fail; @@ -548,7 +539,7 @@ xfs_readsb(xfs_mount_t *mp) bp = xfs_buf_read_flags(mp->m_ddev_targp, XFS_SB_DADDR, BTOBB(sector_size), extra_flags); if (!bp || XFS_BUF_ISERROR(bp)) { - cmn_err(CE_WARN, "XFS: SB re-read failed"); + xfs_fs_mount_cmn_err(flags, "SB re-read failed"); error = bp ? XFS_BUF_GETERROR(bp) : ENOMEM; goto fail; } @@ -678,7 +669,7 @@ xfs_mountfs( int error = 0; if (mp->m_sb_bp == NULL) { - if ((error = xfs_readsb(mp))) { + if ((error = xfs_readsb(mp, mfsi_flags))) { return error; } } diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 66cbee79864..668ad23fd37 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -510,9 +510,12 @@ xfs_preferred_iosize(xfs_mount_t *mp) */ #define XFS_MFSI_SECOND 0x01 /* Secondary mount -- skip stuff */ #define XFS_MFSI_CLIENT 0x02 /* Is a client -- skip lots of stuff */ +/* XFS_MFSI_RRINODES */ #define XFS_MFSI_NOUNLINK 0x08 /* Skip unlinked inode processing in */ /* log recovery */ #define XFS_MFSI_NO_QUOTACHECK 0x10 /* Skip quotacheck processing */ +/* XFS_MFSI_CONVERT_SUNIT */ +#define XFS_MFSI_QUIET 0x40 /* Be silent if mount errors found */ /* * Macros for getting from mount to vfs and back. @@ -581,7 +584,7 @@ extern int xfs_mod_incore_sb_unlocked(xfs_mount_t *, xfs_sb_field_t, extern int xfs_mod_incore_sb_batch(xfs_mount_t *, xfs_mod_sb_t *, uint, int); extern struct xfs_buf *xfs_getsb(xfs_mount_t *, int); -extern int xfs_readsb(xfs_mount_t *mp); +extern int xfs_readsb(xfs_mount_t *, int); extern void xfs_freesb(xfs_mount_t *); extern void xfs_do_force_shutdown(bhv_desc_t *, int, char *, int); extern int xfs_syncsub(xfs_mount_t *, int, int, int *); diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h index 4f6a034de7f..7fbef974bce 100644 --- a/fs/xfs/xfs_quota.h +++ b/fs/xfs/xfs_quota.h @@ -196,10 +196,11 @@ typedef struct xfs_qoff_logformat { #define XFS_QMOPT_QUOTAOFF 0x0000080 /* quotas are being turned off */ #define XFS_QMOPT_UMOUNTING 0x0000100 /* filesys is being unmounted */ #define XFS_QMOPT_DOLOG 0x0000200 /* log buf changes (in quotacheck) */ -#define XFS_QMOPT_DOWARN 0x0000400 /* increase warning cnt if necessary */ +#define XFS_QMOPT_DOWARN 0x0000400 /* increase warning cnt if needed */ #define XFS_QMOPT_ILOCKED 0x0000800 /* inode is already locked (excl) */ -#define XFS_QMOPT_DQREPAIR 0x0001000 /* repair dquot, if damaged. */ +#define XFS_QMOPT_DQREPAIR 0x0001000 /* repair dquot if damaged */ #define XFS_QMOPT_GQUOTA 0x0002000 /* group dquot requested */ +#define XFS_QMOPT_ENOSPC 0x0004000 /* enospc instead of edquot (prj) */ /* * flags to xfs_trans_mod_dquot to indicate which field needs to be diff --git a/fs/xfs/xfs_vfsops.c b/fs/xfs/xfs_vfsops.c index 504d2a80747..f0e09ca1413 100644 --- a/fs/xfs/xfs_vfsops.c +++ b/fs/xfs/xfs_vfsops.c @@ -442,6 +442,9 @@ xfs_mount( p = vfs_bhv_lookup(vfsp, VFS_POSITION_IO); mp->m_io_ops = p ? *(xfs_ioops_t *) vfs_bhv_custom(p) : xfs_iocore_xfs; + if (args->flags & XFSMNT_QUIET) + flags |= XFS_MFSI_QUIET; + /* * Open real time and log devices - order is important. */ @@ -492,7 +495,7 @@ xfs_mount( error = xfs_start_flags(vfsp, args, mp); if (error) goto error1; - error = xfs_readsb(mp); + error = xfs_readsb(mp, flags); if (error) goto error1; error = xfs_finish_flags(vfsp, args, mp); @@ -1697,8 +1700,9 @@ xfs_parseargs( int dsunit, dswidth, vol_dsunit, vol_dswidth; int iosize; - args->flags2 |= XFSMNT2_COMPAT_IOSIZE; args->flags |= XFSMNT_IDELETE; + args->flags |= XFSMNT_BARRIER; + args->flags2 |= XFSMNT2_COMPAT_IOSIZE; if (!options) goto done; @@ -1947,8 +1951,6 @@ xfs_showargs( seq_printf(m, "," MNTOPT_IKEEP); if (!(mp->m_flags & XFS_MOUNT_COMPAT_IOSIZE)) seq_printf(m, "," MNTOPT_LARGEIO); - if (mp->m_flags & XFS_MOUNT_BARRIER) - seq_printf(m, "," MNTOPT_BARRIER); if (!(vfsp->vfs_flag & VFS_32BITINODES)) seq_printf(m, "," MNTOPT_64BITINODE); diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index de49601919c..fa71b305ba5 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -4649,6 +4649,10 @@ vnodeops_t xfs_vnodeops = { #ifdef HAVE_SENDFILE .vop_sendfile = xfs_sendfile, #endif +#ifdef HAVE_SPLICE + .vop_splice_read = xfs_splice_read, + .vop_splice_write = xfs_splice_write, +#endif .vop_write = xfs_write, .vop_ioctl = xfs_ioctl, .vop_getattr = xfs_getattr, |