Mercurial > ecos
changeset 1373:00a9f2cb1081
* cdl/jffs2.cdl:
* src/malloc-ecos.c:
* src/fs-ecos.c: Allocate jffs2_raw_node_ref structs
from pool or malloc depending on a CDL configuration.
| author | asl |
|---|---|
| date | Thu, 20 Nov 2003 21:06:25 +0000 |
| parents | 121121d3d18f |
| children | 0db10af9ebb7 |
| files | packages/fs/jffs2/current/ChangeLog packages/fs/jffs2/current/cdl/jffs2.cdl packages/fs/jffs2/current/src/fs-ecos.c packages/fs/jffs2/current/src/malloc-ecos.c packages/fs/jffs2/current/src/os-ecos.h |
| diffstat | 5 files changed, 1557 insertions(+), 1458 deletions(-) [+] |
line wrap: on
line diff
--- a/packages/fs/jffs2/current/ChangeLog +++ b/packages/fs/jffs2/current/ChangeLog @@ -1,8 +1,20 @@ +2003-11-20 Thomas Koeller <thomas.koeller@baslerweb.com> + + * cdl/jffs2.cdl: + * src/malloc-ecos.c: + * src/fs-ecos.c: Allocate jffs2_raw_node_ref structs + from pool or malloc depending on a CDL configuration. + +2003-11-20 David Woodhouse <dwmw2@infradead.org> + + * Rearrangement of the compression code into a cleaner API + which can be disabled under control from CDL. + 2003-10-14 Thomas Koeller <thomas.koeller@baslerweb.com> - * src/os-ecos.h: Made definition of CONFIG_JFFS2_FS_DEBUG - conditional, so it can be overwritten by a -D compiler - option. + * src/os-ecos.h: Made definition of CONFIG_JFFS2_FS_DEBUG + conditional, so it can be overwritten by a -D compiler + option. 2003-09-23 Thomas Koeller <thomas.koeller@baslerweb.com> @@ -10,266 +22,266 @@ 2003-09-23 Thomas Koeller <thomas.koel 2003-09-23 Andrew Lunn <andrew.lunn@ascom.ch> - * src/fs-ecos.c: Added test to detect known broken ARM compiler + * src/fs-ecos.c: Added test to detect known broken ARM compiler 2003-09-23 Thomas Koeller <thomas.koeller@baslerweb.com> * src/fs-ecos.c: Fixed broken hard link creation. - + 2003-09-23 Andrew Lunn <andrew.lunn@ascom.ch> - * src/fs-ecos.c: Minor optimization of previous patch. - + * src/fs-ecos.c: Minor optimization of previous patch. + 2003-09-21 Thomas Koeller <thomas.koeller@baslerweb.com> - Andrew Lunn <andrew.lunn@ascom.ch> + Andrew Lunn <andrew.lunn@ascom.ch> - * src/fs-ecos.c: Do not decrement mount count for unsuccessful - umount attempts. + * src/fs-ecos.c: Do not decrement mount count for unsuccessful + umount attempts. 2003-09-19 Thomas Koeller <thomas.koeller@baslerweb.com> - * src/fs-ecos.c: Another inode number fix. + * src/fs-ecos.c: Another inode number fix. 2003-09-18 Thomas Koeller <thomas.koeller@baslerweb.com> - * src/fs-ecos.c: Inode number returned by stat() was bogus. + * src/fs-ecos.c: Inode number returned by stat() was bogus. 2003-07-27 Andrew Lunn <andrew.lunn@ascom.ch> - * src/os-ecos.h: Added new #defines require for the recent jffs2 - import. + * src/os-ecos.h: Added new #defines require for the recent jffs2 + import. 2003-07-27 Michael Checky <Michael_Checky@ThermoKing.com> - * src/fs-ecos.c: Changed the return error code to be negative as - expected by jffs2_flash_read() and jffs2_flash_write(). - + * src/fs-ecos.c: Changed the return error code to be negative as + expected by jffs2_flash_read() and jffs2_flash_write(). + 2003-04-23 Bob Koninckx <bob.koninckx@mech.kuleuven.ac.be> - * src/fs-ecos.c: mtab -> cyg_mtab. Ditto cyg_mtab_end. + * src/fs-ecos.c: mtab -> cyg_mtab. Ditto cyg_mtab_end. 2003-03-25 Thomas Koeller <thomas.koeller@baslerweb.com> - * src/fs-ecos.c + * src/fs-ecos.c Fixed segmentation fault when unmounting file system. 2003-02-24 Jonathan Larmour <jifl@eCosCentric.com> - * cdl/jffs2.cdl: Fix doc link. + * cdl/jffs2.cdl: Fix doc link. 2003-02-05 Jonathan Larmour <jifl@eCosCentric.com> - * cdl/jffs2.cdl: Remove unused flash geometry CDL options. + * cdl/jffs2.cdl: Remove unused flash geometry CDL options. 2003-02-04 Gary Thomas <gary@mlbassoc.com> on behalf of -2003-02-04 David Woodhouse <dwmw2@cambridge.redhat.com> +2003-02-04 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/write.c: - * src/scan.c: - * src/readinode.c: - * src/read.c: - * src/pushpull.h: - * src/os-ecos.h: - * src/nodemgmt.c: - * src/nodelist.h: - * src/nodelist.c: - * src/malloc-ecos.c: - * src/jffs2port.h: - * src/gc.c: - * src/file-ecos.c: - * src/erase.c: - * src/dir-ecos.c: - * src/compr_zlib.c: - * src/compr_rubin.c: - * src/compr_rtime.c: - * src/compr.c: - * src/build.c: - * cdl/jffs2.cdl: Update to latest public JFFS2 codebase. + * src/write.c: + * src/scan.c: + * src/readinode.c: + * src/read.c: + * src/pushpull.h: + * src/os-ecos.h: + * src/nodemgmt.c: + * src/nodelist.h: + * src/nodelist.c: + * src/malloc-ecos.c: + * src/jffs2port.h: + * src/gc.c: + * src/file-ecos.c: + * src/erase.c: + * src/dir-ecos.c: + * src/compr_zlib.c: + * src/compr_rubin.c: + * src/compr_rtime.c: + * src/compr.c: + * src/build.c: + * cdl/jffs2.cdl: Update to latest public JFFS2 codebase. - * src/list.h: - * src/jffs2_fs_sb.h: - * src/jffs2_fs_i.h: - * src/jffs2.h: - * src/jffs2.c: - * src/crc32.h: - * src/background.c: Removed/renamed file(s). + * src/list.h: + * src/jffs2_fs_sb.h: + * src/jffs2_fs_i.h: + * src/jffs2.h: + * src/jffs2.c: + * src/crc32.h: + * src/background.c: Removed/renamed file(s). - * src/fs-ecos.c: - * include/linux/jffs2_fs_sb.h: - * include/linux/jffs2_fs_i.h: - * include/linux/jffs2.h: New file(s). + * src/fs-ecos.c: + * include/linux/jffs2_fs_sb.h: + * include/linux/jffs2_fs_i.h: + * include/linux/jffs2.h: New file(s). 2002-12-06 Andrew Lunn <andrew.lunn@ascom.ch> - * cdl/jffs2.cdl: Implements the CYGINT_IO_FILEIO_FS interface. + * cdl/jffs2.cdl: Implements the CYGINT_IO_FILEIO_FS interface. 2002-10-11 Andrew Lunn <andrew.lunn@ascom.ch> - * src/crc32.h (crc32): Use the CRC package for crc calculation - * tests/romfileio1.c (main): Pass the name of the device to mount + * src/crc32.h (crc32): Use the CRC package for crc calculation + * tests/romfileio1.c (main): Pass the name of the device to mount 2002-05-20 Jonathan Larmour <jlarmour@redhat.com> - * src/LICENCE: New file. Contains license for JFFS2, now GPL+exception. - * src/background.c: Point at LICENSE file instead of existing text. - * src/build.c: Ditto. - * src/compr.c: Ditto. - * src/compr_rtime.c: Ditto. - * src/compr_rubin.c: Ditto. - * src/compr_zlib.c: Ditto. - * src/dir-ecos.c: Ditto. - * src/erase.c: Ditto. - * src/file-ecos.c: Ditto. - * src/gc.c: Ditto. - * src/jffs2.h: Ditto. - * src/list.h: Ditto. - * src/malloc-ecos.c: Ditto. - * src/nodelist.c: Ditto. - * src/nodelist.h: Ditto. - * src/nodemgmt.c: Ditto. - * src/os-ecos.h: Ditto. - * src/pushpull.h: Ditto. - * src/read.c: Ditto. - * src/readinode.c: Ditto. - * src/scan.c: Ditto. - * src/write.c: Ditto. + * src/LICENCE: New file. Contains license for JFFS2, now GPL+exception. + * src/background.c: Point at LICENSE file instead of existing text. + * src/build.c: Ditto. + * src/compr.c: Ditto. + * src/compr_rtime.c: Ditto. + * src/compr_rubin.c: Ditto. + * src/compr_zlib.c: Ditto. + * src/dir-ecos.c: Ditto. + * src/erase.c: Ditto. + * src/file-ecos.c: Ditto. + * src/gc.c: Ditto. + * src/jffs2.h: Ditto. + * src/list.h: Ditto. + * src/malloc-ecos.c: Ditto. + * src/nodelist.c: Ditto. + * src/nodelist.h: Ditto. + * src/nodemgmt.c: Ditto. + * src/os-ecos.h: Ditto. + * src/pushpull.h: Ditto. + * src/read.c: Ditto. + * src/readinode.c: Ditto. + * src/scan.c: Ditto. + * src/write.c: Ditto. 2002-01-28 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/super-ecos.c: Removed. - * src/jffs2.c: Merge jffs2_write_super() and jffs2_put_super() into - the routines from which they were called, put jffs2_read_super() - in as a static function with a view to doing same RSN. - * src/jffs2port.h: Remove prototypes of functions that died. - * cdl/jffs2.cdl: Remove super-ecos.c - * src/dir-ecos.c src/write.c: Increase highest_version _before_ - assigning to new node, not after. - + * src/super-ecos.c: Removed. + * src/jffs2.c: Merge jffs2_write_super() and jffs2_put_super() into + the routines from which they were called, put jffs2_read_super() + in as a static function with a view to doing same RSN. + * src/jffs2port.h: Remove prototypes of functions that died. + * cdl/jffs2.cdl: Remove super-ecos.c + * src/dir-ecos.c src/write.c: Increase highest_version _before_ + assigning to new node, not after. + 2002-01-27 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/read.c (jffs2_read_inode_range): Deal correctly with - non-page-aligned read requests. We have to deal with the - case where we want to read from somewhere other than the - beginning of a frag. - * src/jffs2.c (jffs2_fo_read): Use jffs2_read_inode_range - instead of jffs2_readpage. + * src/read.c (jffs2_read_inode_range): Deal correctly with + non-page-aligned read requests. We have to deal with the + case where we want to read from somewhere other than the + beginning of a frag. + * src/jffs2.c (jffs2_fo_read): Use jffs2_read_inode_range + instead of jffs2_readpage. 2002-01-25 Jonathan Larmour <jlarmour@redhat.com> - * cdl/jffs2.cdl: We want CYGPKG_IO_FILEIO_INODE. - * src/dir-ecos.c (jffs2_symlink): Remove. eCos doesn't support symlinks. - (jffs2_mknod): Similar. - (jffs2_mkdir): Don't call d_instantiate - its a nop. - (jffs2_rename): Ditto. - * src/file-ecos.c (jffs2_commit_write): Don't set blocks. - * src/jffs2.c (jffs2_flash_writev): Rewrite to only write aligned - quantities to flash. - * src/jffs2port.h: Lots of decrufting. - * src/os-ecos.h: Ditto (a bit). - * src/readinode.c (jffs2_read_inode): Don't set blocks/blksize in inode. - * src/write.c (jffs2_new_inode): Ditto when __ECOS. - (jffs2_write_dnode): don't call writev with extra vectors - unnecessarily. - * src/super-ecos.c (jffs2_statfs): Remove - unused. + * cdl/jffs2.cdl: We want CYGPKG_IO_FILEIO_INODE. + * src/dir-ecos.c (jffs2_symlink): Remove. eCos doesn't support symlinks. + (jffs2_mknod): Similar. + (jffs2_mkdir): Don't call d_instantiate - its a nop. + (jffs2_rename): Ditto. + * src/file-ecos.c (jffs2_commit_write): Don't set blocks. + * src/jffs2.c (jffs2_flash_writev): Rewrite to only write aligned + quantities to flash. + * src/jffs2port.h: Lots of decrufting. + * src/os-ecos.h: Ditto (a bit). + * src/readinode.c (jffs2_read_inode): Don't set blocks/blksize in inode. + * src/write.c (jffs2_new_inode): Ditto when __ECOS. + (jffs2_write_dnode): don't call writev with extra vectors + unnecessarily. + * src/super-ecos.c (jffs2_statfs): Remove - unused. 2002-01-25 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/super-ecos.c: Cruftectomy. - * src/compr*.[ch] src/pushpull.h: Namespace cleanups merged from - mainstream sources. Bit push functions made inline. - * src/pushpull.c: Removed. - * cdl/jffs2.c: Remove reference to pushpull.c - * src/file-ecos.c: Cruftectomy. Lots of unused stuff here. - * src/jffs2.c src/jffs2port.h: Remove some functions from jffs2.c - which are no longer used, move some others to jffs2port.h as - inline functions so they don't pollute the namespace. + * src/super-ecos.c: Cruftectomy. + * src/compr*.[ch] src/pushpull.h: Namespace cleanups merged from + mainstream sources. Bit push functions made inline. + * src/pushpull.c: Removed. + * cdl/jffs2.c: Remove reference to pushpull.c + * src/file-ecos.c: Cruftectomy. Lots of unused stuff here. + * src/jffs2.c src/jffs2port.h: Remove some functions from jffs2.c + which are no longer used, move some others to jffs2port.h as + inline functions so they don't pollute the namespace. 2002-01-24 Jonathan Larmour <jlarmour@redhat.com> - * tests/fileio1.c: Check in listdir that the number of dirents is - correct, taking into account existing files in case it's live. + * tests/fileio1.c: Check in listdir that the number of dirents is + correct, taking into account existing files in case it's live. - * src/dir-ecos.c (jffs2_readdir): move to.... + * src/dir-ecos.c (jffs2_readdir): move to.... - * src/jffs2.c (jffs2_fo_dirread): here. And fix the return code - in the process so it now works. - (filldir): Make inline and simpler. - * src/jffs2port.h: remove filldir related stuff. + * src/jffs2.c (jffs2_fo_dirread): here. And fix the return code + in the process so it now works. + (filldir): Make inline and simpler. + * src/jffs2port.h: remove filldir related stuff. 2002-01-24 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/dir-ecos.c: Cruftectomy. Remove most of the stuff that was - commented out. Remove jffs2_do_{create,link,unlink} to write.c - * src/write.c: Add jffs2_do_{create,link,unlink} as in mainline. - * src/nodelist.h: Prototypes for the above. - * src/jffs2port.h: Don't include <pkgconf/kernel.h>. - + * src/dir-ecos.c: Cruftectomy. Remove most of the stuff that was + commented out. Remove jffs2_do_{create,link,unlink} to write.c + * src/write.c: Add jffs2_do_{create,link,unlink} as in mainline. + * src/nodelist.h: Prototypes for the above. + * src/jffs2port.h: Don't include <pkgconf/kernel.h>. + 2002-01-23 Jonathan Larmour <jlarmour@redhat.com> - * src/jffs2.c (jffs2_mount): Allow multiple FSs, and integration - with flash block device. - (jffs2_flash_read): Use flash block device. - (jffs2_flash_erase): Ditto. - (jffs2_flash_write): Ditto. - (do_flash_init): Remove - now done by block device layer - * src/list.h: Remove and reimplement from scratch to avoid GPL. - * src/os-ecos.h: Keep flash block dev handle in superblock. - eCos does support obsoleting as it isn't NAND only. - * src/dir-ecos.c (jffs2_readdir): Return correct value on success. - Merge in changes mutatis mutandis from between v1.49 and v1.51 of - dir.c in main repository. - * cdl/jffs2.cdl: CYGPKG_MEMALLOC more accurately CYGINT_ISO_MALLOC. - Only jffs2.c needs to be in libextras.a - Requires Flash block devices as an alternative for hardcoding - the sector size, flash size and base address. - * src/super-ecos.c (jffs2_read_super): Use flash block device for - sector and flash sizes. - * tests/fileio1.c: mount using block device (defined by CDL). - No need to init here - done by flash block device layer. - + * src/jffs2.c (jffs2_mount): Allow multiple FSs, and integration + with flash block device. + (jffs2_flash_read): Use flash block device. + (jffs2_flash_erase): Ditto. + (jffs2_flash_write): Ditto. + (do_flash_init): Remove - now done by block device layer + * src/list.h: Remove and reimplement from scratch to avoid GPL. + * src/os-ecos.h: Keep flash block dev handle in superblock. + eCos does support obsoleting as it isn't NAND only. + * src/dir-ecos.c (jffs2_readdir): Return correct value on success. + Merge in changes mutatis mutandis from between v1.49 and v1.51 of + dir.c in main repository. + * cdl/jffs2.cdl: CYGPKG_MEMALLOC more accurately CYGINT_ISO_MALLOC. + Only jffs2.c needs to be in libextras.a + Requires Flash block devices as an alternative for hardcoding + the sector size, flash size and base address. + * src/super-ecos.c (jffs2_read_super): Use flash block device for + sector and flash sizes. + * tests/fileio1.c: mount using block device (defined by CDL). + No need to init here - done by flash block device layer. + 2002-01-21 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/read.c: Obtain inode lock around reading symlink target. - * src/dir-ecos.c: Fix the long-standing up() without down() in - jffs2_readdir() when only the '.' and '..' entries are read, from - v1.52 of master dir.c. Merge copyright date change from v1.50 - note - that the portability cleanups from v1.51 aren't yet merged. - * src/os-ecos.h: Add jffs2_can_mark_obsolete() and the macros - for getting at generic inode fields from a struct jffs2_inode_info* - * src/nodemgmt.c: Remove an #ifndef __ECOS and use - jffs2_can_mark_obsolete() instead. - * src/jffs2port.h: up() is cyg_drv_mutex_unlock() not _release() - * src/gc.c: Merge portability cleanups - get rid of a lot of - references to 'struct inode'. Also include the attempt at NAND - support in jffs2_garbage_collect_deletion_dirent(). - + * src/read.c: Obtain inode lock around reading symlink target. + * src/dir-ecos.c: Fix the long-standing up() without down() in + jffs2_readdir() when only the '.' and '..' entries are read, from + v1.52 of master dir.c. Merge copyright date change from v1.50 - note + that the portability cleanups from v1.51 aren't yet merged. + * src/os-ecos.h: Add jffs2_can_mark_obsolete() and the macros + for getting at generic inode fields from a struct jffs2_inode_info* + * src/nodemgmt.c: Remove an #ifndef __ECOS and use + jffs2_can_mark_obsolete() instead. + * src/jffs2port.h: up() is cyg_drv_mutex_unlock() not _release() + * src/gc.c: Merge portability cleanups - get rid of a lot of + references to 'struct inode'. Also include the attempt at NAND + support in jffs2_garbage_collect_deletion_dirent(). + 2002-01-11 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/jffs2port.h: Switch semaphore emulation to cyg_drv_mutex_t, - remove some kernel includes which seem gratuitous. - * cdl/jffs2.cdl: Require CYGPKG_MEMALLOC - * src/compr_zlib.c src/compr.c: Merge changes from mainline code - to make mkfs.jffs2 compile again. - + * src/jffs2port.h: Switch semaphore emulation to cyg_drv_mutex_t, + remove some kernel includes which seem gratuitous. + * cdl/jffs2.cdl: Require CYGPKG_MEMALLOC + * src/compr_zlib.c src/compr.c: Merge changes from mainline code + to make mkfs.jffs2 compile again. + 2002-01-10 David Woodhouse <dwmw2@cambridge.redhat.com> - * src/jffs2.c: The 'mode' arg passed to jffs2_open() shouldn't - have been called 'mode'. It's 'oflags'. You have to make up a - mode for the newly-created file yourself. - * src/nodelist.h src/read.c: Fix jffs2_getlink() so it takes - portable arguments, not a dentry. Move it to read.c and symlink.c - becomes obsolete. - * src/symlink-ecos.c: Removed. - * cdl/jffs2.cdl: Remove symlink-ecos.c + * src/jffs2.c: The 'mode' arg passed to jffs2_open() shouldn't + have been called 'mode'. It's 'oflags'. You have to make up a + mode for the newly-created file yourself. + * src/nodelist.h src/read.c: Fix jffs2_getlink() so it takes + portable arguments, not a dentry. Move it to read.c and symlink.c + becomes obsolete. + * src/symlink-ecos.c: Removed. + * cdl/jffs2.cdl: Remove symlink-ecos.c 2002-01-09 David Woodhouse <dwmw2@cambridge.redhat.com> - * Import updated JFFS2 sources into eCos tree. + * Import updated JFFS2 sources into eCos tree. 2000-08-28 Dominic Ostrowski (dominic.ostrowski@3glab.com) * started on port of JFFS2 using ramfs as a template - + //=========================================================================== //####ECOSGPLCOPYRIGHTBEGIN#### // ------------------------------------------- @@ -304,5 +316,5 @@ 2000-08-28 Dominic Ostrowski (dominic.o // ------------------------------------------- //####ECOSGPLCOPYRIGHTEND#### //=========================================================================== - +
--- a/packages/fs/jffs2/current/cdl/jffs2.cdl +++ b/packages/fs/jffs2/current/cdl/jffs2.cdl @@ -148,6 +148,21 @@ cdl_package CYGPKG_FS_JFFS2 { effective." } + cdl_option CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE { + display "Memory pool size" + flavor data + default_value 0 + description " + In order to manage data stored in flash, the file system + needs to use data structures (jffs2_raw_node_ref) allocated + in RAM. You can specify here the maximum number of such + structures you expect to be used, which will then be allocated + statically. If this option is set to 0, the structures will + be allocated dynamically via malloc(), which may incur some + memory overhead depending on the particular malloc() + implementation used." + } + cdl_option CYGPKG_FS_JFFS2_CFLAGS_ADD { display "Additional compiler flags" flavor data
--- a/packages/fs/jffs2/current/src/fs-ecos.c +++ b/packages/fs/jffs2/current/src/fs-ecos.c @@ -41,39 +41,39 @@ static int jffs2_mount(cyg_fstab_entry * fste, cyg_mtab_entry * mte); static int jffs2_umount(cyg_mtab_entry * mte); static int jffs2_open(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - int mode, cyg_file * fte); + int mode, cyg_file * fte); static int jffs2_ops_unlink(cyg_mtab_entry * mte, cyg_dir dir, - const char *name); + const char *name); static int jffs2_ops_mkdir(cyg_mtab_entry * mte, cyg_dir dir, const char *name); static int jffs2_ops_rmdir(cyg_mtab_entry * mte, cyg_dir dir, const char *name); static int jffs2_ops_rename(cyg_mtab_entry * mte, cyg_dir dir1, - const char *name1, cyg_dir dir2, const char *name2); + const char *name1, cyg_dir dir2, const char *name2); static int jffs2_ops_link(cyg_mtab_entry * mte, cyg_dir dir1, const char *name1, - cyg_dir dir2, const char *name2, int type); + cyg_dir dir2, const char *name2, int type); static int jffs2_opendir(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - cyg_file * fte); + cyg_file * fte); static int jffs2_chdir(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - cyg_dir * dir_out); + cyg_dir * dir_out); static int jffs2_stat(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - struct stat *buf); + struct stat *buf); static int jffs2_getinfo(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - int key, void *buf, int len); + int key, void *buf, int len); static int jffs2_setinfo(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - int key, void *buf, int len); + int key, void *buf, int len); // File operations static int jffs2_fo_read(struct CYG_FILE_TAG *fp, struct CYG_UIO_TAG *uio); static int jffs2_fo_write(struct CYG_FILE_TAG *fp, struct CYG_UIO_TAG *uio); static int jffs2_fo_lseek(struct CYG_FILE_TAG *fp, off_t * pos, int whence); static int jffs2_fo_ioctl(struct CYG_FILE_TAG *fp, CYG_ADDRWORD com, - CYG_ADDRWORD data); + CYG_ADDRWORD data); static int jffs2_fo_fsync(struct CYG_FILE_TAG *fp, int mode); static int jffs2_fo_close(struct CYG_FILE_TAG *fp); static int jffs2_fo_fstat(struct CYG_FILE_TAG *fp, struct stat *buf); static int jffs2_fo_getinfo(struct CYG_FILE_TAG *fp, int key, void *buf, - int len); + int len); static int jffs2_fo_setinfo(struct CYG_FILE_TAG *fp, int key, void *buf, - int len); + int len); // Directory operations static int jffs2_fo_dirread(struct CYG_FILE_TAG *fp, struct CYG_UIO_TAG *uio); @@ -89,33 +89,33 @@ static int jffs2_fo_dirlseek(struct CYG_ // we should never block in any filesystem operations. FSTAB_ENTRY(jffs2_fste, "jffs2", 0, - CYG_SYNCMODE_FILE_FILESYSTEM | CYG_SYNCMODE_IO_FILESYSTEM, - jffs2_mount, - jffs2_umount, - jffs2_open, - jffs2_ops_unlink, - jffs2_ops_mkdir, - jffs2_ops_rmdir, - jffs2_ops_rename, - jffs2_ops_link, - jffs2_opendir, - jffs2_chdir, jffs2_stat, jffs2_getinfo, jffs2_setinfo); + CYG_SYNCMODE_FILE_FILESYSTEM | CYG_SYNCMODE_IO_FILESYSTEM, + jffs2_mount, + jffs2_umount, + jffs2_open, + jffs2_ops_unlink, + jffs2_ops_mkdir, + jffs2_ops_rmdir, + jffs2_ops_rename, + jffs2_ops_link, + jffs2_opendir, + jffs2_chdir, jffs2_stat, jffs2_getinfo, jffs2_setinfo); // ------------------------------------------------------------------------- // File operations. // This set of file operations are used for normal open files. static cyg_fileops jffs2_fileops = { - jffs2_fo_read, - jffs2_fo_write, - jffs2_fo_lseek, - jffs2_fo_ioctl, - cyg_fileio_seltrue, - jffs2_fo_fsync, - jffs2_fo_close, - jffs2_fo_fstat, - jffs2_fo_getinfo, - jffs2_fo_setinfo + jffs2_fo_read, + jffs2_fo_write, + jffs2_fo_lseek, + jffs2_fo_ioctl, + cyg_fileio_seltrue, + jffs2_fo_fsync, + jffs2_fo_close, + jffs2_fo_fstat, + jffs2_fo_getinfo, + jffs2_fo_setinfo }; // ------------------------------------------------------------------------- @@ -125,34 +125,35 @@ static cyg_fileops jffs2_fileops = { // close entries are functional. static cyg_fileops jffs2_dirops = { - jffs2_fo_dirread, - (cyg_fileop_write *) cyg_fileio_enosys, - jffs2_fo_dirlseek, - (cyg_fileop_ioctl *) cyg_fileio_enosys, - cyg_fileio_seltrue, - (cyg_fileop_fsync *) cyg_fileio_enosys, - jffs2_fo_close, - (cyg_fileop_fstat *) cyg_fileio_enosys, - (cyg_fileop_getinfo *) cyg_fileio_enosys, - (cyg_fileop_setinfo *) cyg_fileio_enosys + jffs2_fo_dirread, + (cyg_fileop_write *) cyg_fileio_enosys, + jffs2_fo_dirlseek, + (cyg_fileop_ioctl *) cyg_fileio_enosys, + cyg_fileio_seltrue, + (cyg_fileop_fsync *) cyg_fileio_enosys, + jffs2_fo_close, + (cyg_fileop_fstat *) cyg_fileio_enosys, + (cyg_fileop_getinfo *) cyg_fileio_enosys, + (cyg_fileop_setinfo *) cyg_fileio_enosys }; //========================================================================== // STATIC VARIABLES !!! -static char read_write_buffer[PAGE_CACHE_SIZE]; //avoids malloc when user may be under memory pressure -static char gc_buffer[PAGE_CACHE_SIZE]; //avoids malloc when user may be under memory pressure +static char read_write_buffer[PAGE_CACHE_SIZE]; //avoids malloc when user may be under memory pressure +static char gc_buffer[PAGE_CACHE_SIZE]; //avoids malloc when user may be under memory pressure +static unsigned char n_fs_mounted = 0; // a counter to track the number of jffs2 instances mounted //========================================================================== // Directory operations struct jffs2_dirsearch { - struct inode *dir; // directory to search - const char *path; // path to follow - struct inode *node; // Node found - const char *name; // last name fragment used - int namelen; // name fragment length - cyg_bool last; // last name in path? + struct inode *dir; // directory to search + const char *path; // path to follow + struct inode *node; // Node found + const char *name; // last name fragment used + int namelen; // name fragment length + cyg_bool last; // last name in path? }; typedef struct jffs2_dirsearch jffs2_dirsearch; @@ -167,46 +168,46 @@ typedef struct jffs2_dirsearch jffs2_dir static int dec_refcnt(struct inode *node) { - int err = ENOERR; - node->i_count--; + int err = ENOERR; + node->i_count--; - // In JFFS2 inode's are temporary in ram structures that are free'd when the usage i_count drops to 0 - // The i_nlink however is managed by JFFS2 and is unrelated to usage - if (node->i_count == 0) { - // This inode is not in use, so delete it. - iput(node); - } + // In JFFS2 inode's are temporary in ram structures that are free'd when the usage i_count drops to 0 + // The i_nlink however is managed by JFFS2 and is unrelated to usage + if (node->i_count == 0) { + // This inode is not in use, so delete it. + iput(node); + } - return err; + return err; } // FIXME: This seems like real cruft. Wouldn't it be better just to do the // right thing? static void icache_evict(struct inode *root_i, struct inode *i) { - struct inode *cached_inode; - struct inode *next_inode; + struct inode *cached_inode; + struct inode *next_inode; - D2(printf("icache_evict\n")); - // If this is an absolute search path from the root, - // remove all cached inodes with i_count of zero (these are only - // held where needed for dotdot filepaths) - if (i == root_i) { - for (cached_inode = root_i; cached_inode != NULL; - cached_inode = next_inode) { - next_inode = cached_inode->i_cache_next; - if (cached_inode->i_count == 0) { - cached_inode->i_cache_prev->i_cache_next = cached_inode->i_cache_next; // Previous entry points ahead of us - if (cached_inode->i_cache_next != NULL) - cached_inode->i_cache_next->i_cache_prev = cached_inode->i_cache_prev; // Next entry points behind us - jffs2_clear_inode(cached_inode); - D2(printf - ("free icache_evict inode %x $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$\n", - cached_inode)); - free(cached_inode); - } - } - } + D2(printf("icache_evict\n")); + // If this is an absolute search path from the root, + // remove all cached inodes with i_count of zero (these are only + // held where needed for dotdot filepaths) + if (i == root_i) { + for (cached_inode = root_i; cached_inode != NULL; + cached_inode = next_inode) { + next_inode = cached_inode->i_cache_next; + if (cached_inode->i_count == 0) { + cached_inode->i_cache_prev->i_cache_next = cached_inode->i_cache_next; // Previous entry points ahead of us + if (cached_inode->i_cache_next != NULL) + cached_inode->i_cache_next->i_cache_prev = cached_inode->i_cache_prev; // Next entry points behind us + jffs2_clear_inode(cached_inode); + D2(printf + ("free icache_evict inode %x $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$\n", + cached_inode)); + free(cached_inode); + } + } + } } //========================================================================== @@ -217,16 +218,16 @@ static void icache_evict(struct inode *r // Initialize a dirsearch object to start a search static void init_dirsearch(jffs2_dirsearch * ds, - struct inode *dir, const char *name) + struct inode *dir, const char *name) { - D2(printf("init_dirsearch name = %s\n", name)); - D2(printf("init_dirsearch dir = %x\n", dir)); - ds->dir = dir; - ds->path = name; - ds->node = dir; - ds->name = name; - ds->namelen = 0; - ds->last = false; + D2(printf("init_dirsearch name = %s\n", name)); + D2(printf("init_dirsearch dir = %x\n", dir)); + ds->dir = dir; + ds->path = name; + ds->node = dir; + ds->name = name; + ds->namelen = 0; + ds->last = false; } // ------------------------------------------------------------------------- @@ -236,87 +237,87 @@ static void init_dirsearch(jffs2_dirsear static int find_entry(jffs2_dirsearch * ds) { - unsigned long hash; - struct qstr this; - unsigned int c; - const char *hashname; + unsigned long hash; + struct qstr this; + unsigned int c; + const char *hashname; - struct inode *dir = ds->dir; - const char *name = ds->path; - const char *n = name; - char namelen = 0; - struct inode *d; + struct inode *dir = ds->dir; + const char *name = ds->path; + const char *n = name; + char namelen = 0; + struct inode *d; - D2(printf("find_entry\n")); + D2(printf("find_entry\n")); - // check that we really have a directory - if (!S_ISDIR(dir->i_mode)) - return ENOTDIR; + // check that we really have a directory + if (!S_ISDIR(dir->i_mode)) + return ENOTDIR; - // Isolate the next element of the path name. - while (*n != '\0' && *n != '/') - n++, namelen++; + // Isolate the next element of the path name. + while (*n != '\0' && *n != '/') + n++, namelen++; - // If we terminated on a NUL, set last flag. - if (*n == '\0') - ds->last = true; + // If we terminated on a NUL, set last flag. + if (*n == '\0') + ds->last = true; - // update name in dirsearch object - ds->name = name; - ds->namelen = namelen; + // update name in dirsearch object + ds->name = name; + ds->namelen = namelen; - if (name[0] == '.') - switch (namelen) { - default: - break; - case 2: - // Dot followed by not Dot, treat as any other name - if (name[1] != '.') - break; - // Dot Dot - // Move back up the search path - D2(printf("find_entry found ..\n")); - ds->node = ds->dir->i_parent; - if (ds->dir->i_count == 0) { - iput(ds->dir); // This inode may be evicted - ds->dir = NULL; - } - return ENOERR; - case 1: - // Dot is consumed - D2(printf("find_entry found .\n")); - ds->node = ds->dir; - return ENOERR; - } - // Here we have the name and its length set up. - // Search the directory for a matching entry + if (name[0] == '.') + switch (namelen) { + default: + break; + case 2: + // Dot followed by not Dot, treat as any other name + if (name[1] != '.') + break; + // Dot Dot + // Move back up the search path + D2(printf("find_entry found ..\n")); + ds->node = ds->dir->i_parent; + if (ds->dir->i_count == 0) { + iput(ds->dir); // This inode may be evicted + ds->dir = NULL; + } + return ENOERR; + case 1: + // Dot is consumed + D2(printf("find_entry found .\n")); + ds->node = ds->dir; + return ENOERR; + } + // Here we have the name and its length set up. + // Search the directory for a matching entry - hashname = name; - this.name = hashname; - c = *(const unsigned char *) hashname; + hashname = name; + this.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this.len = hashname - (const char *) this.name; - this.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this.len = hashname - (const char *) this.name; + this.hash = end_name_hash(hash); - D2(printf("find_entry for name = %s\n", ds->path)); - d = jffs2_lookup(dir, &this); - D2(printf("find_entry got dir = %x\n", d)); + D2(printf("find_entry for name = %s\n", ds->path)); + d = jffs2_lookup(dir, &this); + D2(printf("find_entry got dir = %x\n", d)); - if (d == NULL) - return ENOENT; + if (d == NULL) + return ENOENT; - // The back path for dotdot to follow - d->i_parent = dir; - // pass back the node we have found - ds->node = d; + // The back path for dotdot to follow + d->i_parent = dir; + // pass back the node we have found + ds->node = d; - return ENOERR; + return ENOERR; } @@ -327,33 +328,33 @@ static int find_entry(jffs2_dirsearch * static int jffs2_find(jffs2_dirsearch * d) { - int err; + int err; - D2(printf("jffs2_find for path =%s\n", d->path)); - // Short circuit empty paths - if (*(d->path) == '\0') - return ENOERR; + D2(printf("jffs2_find for path =%s\n", d->path)); + // Short circuit empty paths + if (*(d->path) == '\0') + return ENOERR; - // iterate down directory tree until we find the object - // we want. - for (;;) { - err = find_entry(d); + // iterate down directory tree until we find the object + // we want. + for (;;) { + err = find_entry(d); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - if (d->last) - return ENOERR; + if (d->last) + return ENOERR; - // every inode traversed in the find is temporary and should be free'd - //iput(d->dir); + // every inode traversed in the find is temporary and should be free'd + //iput(d->dir); - // Update dirsearch object to search next directory. - d->dir = d->node; - d->path += d->namelen; - if (*(d->path) == '/') - d->path++; // skip dirname separators - } + // Update dirsearch object to search next directory. + d->dir = d->node; + d->path += d->namelen; + if (*(d->path) == '/') + d->path++; // skip dirname separators + } } //========================================================================== @@ -362,70 +363,70 @@ static int jffs2_find(jffs2_dirsearch * static int jffs2_pathconf(struct inode *node, struct cyg_pathconf_info *info) { - int err = ENOERR; - D2(printf("jffs2_pathconf\n")); + int err = ENOERR; + D2(printf("jffs2_pathconf\n")); - switch (info->name) { - case _PC_LINK_MAX: - info->value = LINK_MAX; - break; + switch (info->name) { + case _PC_LINK_MAX: + info->value = LINK_MAX; + break; - case _PC_MAX_CANON: - info->value = -1; // not supported - err = EINVAL; - break; + case _PC_MAX_CANON: + info->value = -1; // not supported + err = EINVAL; + break; - case _PC_MAX_INPUT: - info->value = -1; // not supported - err = EINVAL; - break; + case _PC_MAX_INPUT: + info->value = -1; // not supported + err = EINVAL; + break; - case _PC_NAME_MAX: - info->value = NAME_MAX; - break; + case _PC_NAME_MAX: + info->value = NAME_MAX; + break; - case _PC_PATH_MAX: - info->value = PATH_MAX; - break; + case _PC_PATH_MAX: + info->value = PATH_MAX; + break; - case _PC_PIPE_BUF: - info->value = -1; // not supported - err = EINVAL; - break; + case _PC_PIPE_BUF: + info->value = -1; // not supported + err = EINVAL; + break; - case _PC_ASYNC_IO: - info->value = -1; // not supported - err = EINVAL; - break; + case _PC_ASYNC_IO: + info->value = -1; // not supported + err = EINVAL; + break; - case _PC_CHOWN_RESTRICTED: - info->value = -1; // not supported - err = EINVAL; - break; + case _PC_CHOWN_RESTRICTED: + info->value = -1; // not supported + err = EINVAL; + break; - case _PC_NO_TRUNC: - info->value = 0; - break; + case _PC_NO_TRUNC: + info->value = 0; + break; - case _PC_PRIO_IO: - info->value = 0; - break; + case _PC_PRIO_IO: + info->value = 0; + break; - case _PC_SYNC_IO: - info->value = 0; - break; + case _PC_SYNC_IO: + info->value = 0; + break; - case _PC_VDISABLE: - info->value = -1; // not supported - err = EINVAL; - break; + case _PC_VDISABLE: + info->value = -1; // not supported + err = EINVAL; + break; - default: - err = EINVAL; - break; - } + default: + err = EINVAL; + break; + } - return err; + return err; } //========================================================================== @@ -437,145 +438,150 @@ static int jffs2_pathconf(struct inode * // filesystem. static int jffs2_read_super(struct super_block *sb) { - struct jffs2_sb_info *c; - struct inode *root_i; - Cyg_ErrNo err; - cyg_uint32 len; - cyg_io_flash_getconfig_devsize_t ds; - cyg_io_flash_getconfig_blocksize_t bs; + struct jffs2_sb_info *c; + struct inode *root_i; + Cyg_ErrNo err; + cyg_uint32 len; + cyg_io_flash_getconfig_devsize_t ds; + cyg_io_flash_getconfig_blocksize_t bs; - D1(printk(KERN_DEBUG "jffs2: read_super\n")); + D1(printk(KERN_DEBUG "jffs2: read_super\n")); - c = JFFS2_SB_INFO(sb); + c = JFFS2_SB_INFO(sb); - len = sizeof (ds); - err = cyg_io_get_config(sb->s_dev, - CYG_IO_GET_CONFIG_FLASH_DEVSIZE, &ds, &len); - if (err != ENOERR) { - D1(printf - ("jffs2: cyg_io_get_config failed to get dev size: %d\n", - err)); - return err; - } - len = sizeof (bs); - bs.offset = 0; - err = cyg_io_get_config(sb->s_dev, - CYG_IO_GET_CONFIG_FLASH_BLOCKSIZE, &bs, &len); - if (err != ENOERR) { - D1(printf - ("jffs2: cyg_io_get_config failed to get block size: %d\n", - err)); - return err; - } + len = sizeof (ds); + err = cyg_io_get_config(sb->s_dev, + CYG_IO_GET_CONFIG_FLASH_DEVSIZE, &ds, &len); + if (err != ENOERR) { + D1(printf + ("jffs2: cyg_io_get_config failed to get dev size: %d\n", + err)); + return err; + } + len = sizeof (bs); + bs.offset = 0; + err = cyg_io_get_config(sb->s_dev, + CYG_IO_GET_CONFIG_FLASH_BLOCKSIZE, &bs, &len); + if (err != ENOERR) { + D1(printf + ("jffs2: cyg_io_get_config failed to get block size: %d\n", + err)); + return err; + } - c->sector_size = bs.block_size; - c->flash_size = ds.dev_size; - c->cleanmarker_size = sizeof(struct jffs2_unknown_node); + c->sector_size = bs.block_size; + c->flash_size = ds.dev_size; + c->cleanmarker_size = sizeof(struct jffs2_unknown_node); - err = jffs2_do_mount_fs(c); - if (err) - return -err; + err = jffs2_do_mount_fs(c); + if (err) + return -err; - D1(printk(KERN_DEBUG "jffs2_read_super(): Getting root inode\n")); - root_i = iget(sb, 1); - if (is_bad_inode(root_i)) { - D1(printk(KERN_WARNING "get root inode failed\n")); - err = EIO; - goto out_nodes; - } + D1(printk(KERN_DEBUG "jffs2_read_super(): Getting root inode\n")); + root_i = iget(sb, 1); + if (is_bad_inode(root_i)) { + D1(printk(KERN_WARNING "get root inode failed\n")); + err = EIO; + goto out_nodes; + } - D1(printk(KERN_DEBUG "jffs2_read_super(): d_alloc_root()\n")); - sb->s_root = d_alloc_root(root_i); - if (!sb->s_root) { - err = ENOMEM; - goto out_root_i; - } - sb->s_blocksize = PAGE_CACHE_SIZE; - sb->s_blocksize_bits = PAGE_CACHE_SHIFT; - sb->s_magic = JFFS2_SUPER_MAGIC; + D1(printk(KERN_DEBUG "jffs2_read_super(): d_alloc_root()\n")); + sb->s_root = d_alloc_root(root_i); + if (!sb->s_root) { + err = ENOMEM; + goto out_root_i; + } + sb->s_blocksize = PAGE_CACHE_SIZE; + sb->s_blocksize_bits = PAGE_CACHE_SHIFT; + sb->s_magic = JFFS2_SUPER_MAGIC; - return 0; + return 0; out_root_i: - iput(root_i); + iput(root_i); out_nodes: - jffs2_free_ino_caches(c); - jffs2_free_raw_node_refs(c); - free(c->blocks); + jffs2_free_ino_caches(c); + jffs2_free_raw_node_refs(c); + free(c->blocks); - return err; + return err; } static int jffs2_mount(cyg_fstab_entry * fste, cyg_mtab_entry * mte) { - extern cyg_mtab_entry cyg_mtab[], cyg_mtab_end; - struct super_block *jffs2_sb = NULL; - struct jffs2_sb_info *c; - cyg_mtab_entry *m; - cyg_io_handle_t t; - Cyg_ErrNo err; + extern cyg_mtab_entry cyg_mtab[], cyg_mtab_end; + struct super_block *jffs2_sb = NULL; + struct jffs2_sb_info *c; + cyg_mtab_entry *m; + cyg_io_handle_t t; + Cyg_ErrNo err; - D2(printf("jffs2_mount\n")); + D2(printf("jffs2_mount\n")); - err = cyg_io_lookup(mte->devname, &t); - if (err != ENOERR) - return -err; + err = cyg_io_lookup(mte->devname, &t); + if (err != ENOERR) + return -err; - // Iterate through the mount table to see if we're mounted - // FIXME: this should be done better - perhaps if the superblock - // can be stored as an inode in the icache. - for (m = &cyg_mtab[0]; m != &cyg_mtab_end; m++) { - // stop if there are more than the configured maximum - if (m - &cyg_mtab[0] >= CYGNUM_FILEIO_MTAB_MAX) { - m = &cyg_mtab_end; - break; - } - if (m->valid && strcmp(m->fsname, "jffs2") == 0 && - strcmp(m->devname, mte->devname) == 0) { - jffs2_sb = (struct super_block *) m->data; - } - } + // Iterate through the mount table to see if we're mounted + // FIXME: this should be done better - perhaps if the superblock + // can be stored as an inode in the icache. + for (m = &cyg_mtab[0]; m != &cyg_mtab_end; m++) { + // stop if there are more than the configured maximum + if (m - &cyg_mtab[0] >= CYGNUM_FILEIO_MTAB_MAX) { + m = &cyg_mtab_end; + break; + } + if (m->valid && strcmp(m->fsname, "jffs2") == 0 && + strcmp(m->devname, mte->devname) == 0) { + jffs2_sb = (struct super_block *) m->data; + } + } - if (jffs2_sb == NULL) { - jffs2_sb = malloc(sizeof (struct super_block)); + if (jffs2_sb == NULL) { + jffs2_sb = malloc(sizeof (struct super_block)); - if (jffs2_sb == NULL) - return ENOMEM; + if (jffs2_sb == NULL) + return ENOMEM; - c = JFFS2_SB_INFO(jffs2_sb); - memset(jffs2_sb, 0, sizeof (struct super_block)); - jffs2_sb->s_dev = t; + c = JFFS2_SB_INFO(jffs2_sb); + memset(jffs2_sb, 0, sizeof (struct super_block)); + jffs2_sb->s_dev = t; - c->inocache_list = malloc(sizeof(struct jffs2_inode_cache *) * INOCACHE_HASHSIZE); - if (!c->inocache_list) { - free(jffs2_sb); - return ENOMEM; - } - memset(c->inocache_list, 0, sizeof(struct jffs2_inode_cache *) * INOCACHE_HASHSIZE); + c->inocache_list = malloc(sizeof(struct jffs2_inode_cache *) * INOCACHE_HASHSIZE); + if (!c->inocache_list) { + free(jffs2_sb); + return ENOMEM; + } + memset(c->inocache_list, 0, sizeof(struct jffs2_inode_cache *) * INOCACHE_HASHSIZE); + if (n_fs_mounted++ == 0) + jffs2_create_slab_caches(); // No error check, cannot fail - err = jffs2_read_super(jffs2_sb); + err = jffs2_read_super(jffs2_sb); - if (err) { - free(jffs2_sb); - free(c->inocache_list); - return err; - } - - jffs2_sb->s_root->i_parent = jffs2_sb->s_root; // points to itself, no dotdot paths above mountpoint - jffs2_sb->s_root->i_cache_prev = NULL; // root inode, so always null - jffs2_sb->s_root->i_cache_next = NULL; - jffs2_sb->s_root->i_count = 1; // Ensures the root inode is always in ram until umount + if (err) { + if (--n_fs_mounted == 0) + jffs2_destroy_slab_caches(); + + free(jffs2_sb); + free(c->inocache_list); + return err; + } - D2(printf("jffs2_mount erasing pending blocks\n")); - jffs2_erase_pending_blocks(c,0); - } - mte->data = (CYG_ADDRWORD) jffs2_sb; + jffs2_sb->s_root->i_parent = jffs2_sb->s_root; // points to itself, no dotdot paths above mountpoint + jffs2_sb->s_root->i_cache_prev = NULL; // root inode, so always null + jffs2_sb->s_root->i_cache_next = NULL; + jffs2_sb->s_root->i_count = 1; // Ensures the root inode is always in ram until umount - jffs2_sb->s_mount_count++; - mte->root = (cyg_dir) jffs2_sb->s_root; - D2(printf("jffs2_mounted superblock at %x\n", mte->root)); + D2(printf("jffs2_mount erasing pending blocks\n")); + jffs2_erase_pending_blocks(c,0); + } + mte->data = (CYG_ADDRWORD) jffs2_sb; - return ENOERR; + jffs2_sb->s_mount_count++; + mte->root = (cyg_dir) jffs2_sb->s_root; + D2(printf("jffs2_mounted superblock at %x\n", mte->root)); + + return ENOERR; } // ------------------------------------------------------------------------- @@ -584,40 +590,41 @@ static int jffs2_mount(cyg_fstab_entry * static int jffs2_umount(cyg_mtab_entry * mte) { - struct inode *root = (struct inode *) mte->root; - struct super_block *jffs2_sb = root->i_sb; - struct jffs2_sb_info *c = JFFS2_SB_INFO(jffs2_sb); + struct inode *root = (struct inode *) mte->root; + struct super_block *jffs2_sb = root->i_sb; + struct jffs2_sb_info *c = JFFS2_SB_INFO(jffs2_sb); - D2(printf("jffs2_umount\n")); + D2(printf("jffs2_umount\n")); - // Only really umount if this is the only mount - if (jffs2_sb->s_mount_count == 1) { + // Only really umount if this is the only mount + if (jffs2_sb->s_mount_count == 1) { - if (root->i_cache_next != NULL) // root icount was set to 1 on mount - return EBUSY; + if (root->i_cache_next != NULL) // root icount was set to 1 on mount + return EBUSY; - dec_refcnt(root); // Time to free the root inode + dec_refcnt(root); // Time to free the root inode - //Clear root inode - //root_i = NULL; + //Clear root inode + //root_i = NULL; - // Clean up the super block and root inode - jffs2_free_ino_caches(c); - jffs2_free_raw_node_refs(c); - free(c->blocks); - free(c->inocache_list); - free(jffs2_sb); - // Clear superblock & root pointer - mte->root = CYG_DIR_NULL; + // Clean up the super block and root inode + jffs2_free_ino_caches(c); + jffs2_free_raw_node_refs(c); + free(c->blocks); + free(c->inocache_list); + free(jffs2_sb); + // Clear superblock & root pointer + mte->root = CYG_DIR_NULL; mte->data = 0; - mte->fs->data = 0; // fstab entry, visible to all mounts. No current mount - // That's all folks. - D2(printf("jffs2_umount No current mounts\n")); - } else { - jffs2_sb->s_mount_count--; + mte->fs->data = 0; // fstab entry, visible to all mounts. No current mount + // That's all folks. + D2(printf("jffs2_umount No current mounts\n")); + } else { + jffs2_sb->s_mount_count--; } - - return ENOERR; + if (--n_fs_mounted == 0) + jffs2_destroy_slab_caches(); + return ENOERR; } // ------------------------------------------------------------------------- @@ -625,94 +632,94 @@ static int jffs2_umount(cyg_mtab_entry * // Open a file for reading or writing. static int jffs2_open(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - int mode, cyg_file * file) + int mode, cyg_file * file) { - jffs2_dirsearch ds; - struct inode *node = NULL; - int err; + jffs2_dirsearch ds; + struct inode *node = NULL; + int err; - D2(printf("jffs2_open\n")); + D2(printf("jffs2_open\n")); - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err == ENOENT) { - if (ds.last && (mode & O_CREAT)) { - unsigned long hash; - struct qstr this; - unsigned int c; - const char *hashname; + if (err == ENOENT) { + if (ds.last && (mode & O_CREAT)) { + unsigned long hash; + struct qstr this; + unsigned int c; + const char *hashname; - // No node there, if the O_CREAT bit is set then we must - // create a new one. The dir and name fields of the dirsearch - // object will have been updated so we know where to put it. + // No node there, if the O_CREAT bit is set then we must + // create a new one. The dir and name fields of the dirsearch + // object will have been updated so we know where to put it. - hashname = ds.name; - this.name = hashname; - c = *(const unsigned char *) hashname; + hashname = ds.name; + this.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this.len = hashname - (const char *) this.name; - this.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this.len = hashname - (const char *) this.name; + this.hash = end_name_hash(hash); - err = jffs2_create(ds.dir, &this, S_IRUGO|S_IXUGO|S_IWUSR|S_IFREG, &node); + err = jffs2_create(ds.dir, &this, S_IRUGO|S_IXUGO|S_IWUSR|S_IFREG, &node); - if (err != 0) { - //Possible orphaned inode on the flash - but will be gc'd - return err; - } + if (err != 0) { + //Possible orphaned inode on the flash - but will be gc'd + return err; + } - err = ENOERR; - } - } else if (err == ENOERR) { - // The node exists. If the O_CREAT and O_EXCL bits are set, we - // must fail the open. + err = ENOERR; + } + } else if (err == ENOERR) { + // The node exists. If the O_CREAT and O_EXCL bits are set, we + // must fail the open. - if ((mode & (O_CREAT | O_EXCL)) == (O_CREAT | O_EXCL)) - err = EEXIST; - else - node = ds.node; - } + if ((mode & (O_CREAT | O_EXCL)) == (O_CREAT | O_EXCL)) + err = EEXIST; + else + node = ds.node; + } - if (err == ENOERR && (mode & O_TRUNC)) { - struct jffs2_inode_info *f = JFFS2_INODE_INFO(node); - struct jffs2_sb_info *c = JFFS2_SB_INFO(node->i_sb); - // If the O_TRUNC bit is set we must clean out the file data. + if (err == ENOERR && (mode & O_TRUNC)) { + struct jffs2_inode_info *f = JFFS2_INODE_INFO(node); + struct jffs2_sb_info *c = JFFS2_SB_INFO(node->i_sb); + // If the O_TRUNC bit is set we must clean out the file data. - node->i_size = 0; - jffs2_truncate_fraglist(c, &f->fragtree, 0); - // Update file times - node->i_ctime = node->i_mtime = cyg_timestamp(); - } + node->i_size = 0; + jffs2_truncate_fraglist(c, &f->fragtree, 0); + // Update file times + node->i_ctime = node->i_mtime = cyg_timestamp(); + } - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // Check that we actually have a file here - if (S_ISDIR(node->i_mode)) - return EISDIR; + // Check that we actually have a file here + if (S_ISDIR(node->i_mode)) + return EISDIR; - node->i_count++; // Count successful open + node->i_count++; // Count successful open - // Initialize the file object + // Initialize the file object - file->f_flag |= mode & CYG_FILE_MODE_MASK; - file->f_type = CYG_FILE_TYPE_FILE; - file->f_ops = &jffs2_fileops; - file->f_offset = (mode & O_APPEND) ? node->i_size : 0; - file->f_data = (CYG_ADDRWORD) node; - file->f_xops = 0; + file->f_flag |= mode & CYG_FILE_MODE_MASK; + file->f_type = CYG_FILE_TYPE_FILE; + file->f_ops = &jffs2_fileops; + file->f_offset = (mode & O_APPEND) ? node->i_size : 0; + file->f_data = (CYG_ADDRWORD) node; + file->f_xops = 0; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -721,46 +728,46 @@ static int jffs2_open(cyg_mtab_entry * m static int jffs2_ops_unlink(cyg_mtab_entry * mte, cyg_dir dir, const char *name) { - unsigned long hash; - struct qstr this; - unsigned int c; - const char *hashname; - jffs2_dirsearch ds; - int err; + unsigned long hash; + struct qstr this; + unsigned int c; + const char *hashname; + jffs2_dirsearch ds; + int err; - D2(printf("jffs2_ops_unlink\n")); + D2(printf("jffs2_ops_unlink\n")); - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // Cannot unlink directories, use rmdir() instead - if (S_ISDIR(ds.node->i_mode)) - return EPERM; + // Cannot unlink directories, use rmdir() instead + if (S_ISDIR(ds.node->i_mode)) + return EPERM; - // Delete it from its directory + // Delete it from its directory - hashname = ds.name; - this.name = hashname; - c = *(const unsigned char *) hashname; + hashname = ds.name; + this.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this.len = hashname - (const char *) this.name; - this.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this.len = hashname - (const char *) this.name; + this.hash = end_name_hash(hash); - err = jffs2_unlink(ds.dir, ds.node, &this); + err = jffs2_unlink(ds.dir, ds.node, &this); - return err; + return err; } // ------------------------------------------------------------------------- @@ -769,57 +776,57 @@ static int jffs2_ops_unlink(cyg_mtab_ent static int jffs2_ops_mkdir(cyg_mtab_entry * mte, cyg_dir dir, const char *name) { - jffs2_dirsearch ds; - struct inode *node = NULL; - int err; + jffs2_dirsearch ds; + struct inode *node = NULL; + int err; - D2(printf("jffs2_ops_mkdir\n")); + D2(printf("jffs2_ops_mkdir\n")); - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err == ENOENT) { - if (ds.last) { - unsigned long hash; - struct qstr this; - unsigned int c; - const char *hashname; - // The entry does not exist, and it is the last element in - // the pathname, so we can create it here. + if (err == ENOENT) { + if (ds.last) { + unsigned long hash; + struct qstr this; + unsigned int c; + const char *hashname; + // The entry does not exist, and it is the last element in + // the pathname, so we can create it here. - hashname = ds.name; - this.name = hashname; - c = *(const unsigned char *) hashname; + hashname = ds.name; + this.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this.len = hashname - (const char *) this.name; - this.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this.len = hashname - (const char *) this.name; + this.hash = end_name_hash(hash); - err = jffs2_mkdir(ds.dir, &this, 0, &node); + err = jffs2_mkdir(ds.dir, &this, 0, &node); - if (err != 0) - return ENOSPC; + if (err != 0) + return ENOSPC; - } - // If this was not the last element, then and intermediate - // directory does not exist. - } else { - // If there we no error, something already exists with that - // name, so we cannot create another one. + } + // If this was not the last element, then and intermediate + // directory does not exist. + } else { + // If there we no error, something already exists with that + // name, so we cannot create another one. - if (err == ENOERR) - err = EEXIST; - } + if (err == ENOERR) + err = EEXIST; + } - return err; + return err; } // ------------------------------------------------------------------------- @@ -828,47 +835,47 @@ static int jffs2_ops_mkdir(cyg_mtab_entr static int jffs2_ops_rmdir(cyg_mtab_entry * mte, cyg_dir dir, const char *name) { - unsigned long hash; - struct qstr this; - unsigned int c; - const char *hashname; - jffs2_dirsearch ds; - int err; + unsigned long hash; + struct qstr this; + unsigned int c; + const char *hashname; + jffs2_dirsearch ds; + int err; - D2(printf("jffs2_ops_rmdir\n")); + D2(printf("jffs2_ops_rmdir\n")); - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // Check that this is actually a directory. - if (!S_ISDIR(ds.node->i_mode)) - return EPERM; + // Check that this is actually a directory. + if (!S_ISDIR(ds.node->i_mode)) + return EPERM; - // Delete the entry. - hashname = ds.name; - this.name = hashname; - c = *(const unsigned char *) hashname; + // Delete the entry. + hashname = ds.name; + this.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this.len = hashname - (const char *) this.name; - this.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this.len = hashname - (const char *) this.name; + this.hash = end_name_hash(hash); - err = jffs2_rmdir(ds.dir, ds.node, &this); + err = jffs2_rmdir(ds.dir, ds.node, &this); - return err; + return err; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -876,96 +883,96 @@ static int jffs2_ops_rmdir(cyg_mtab_entr // Rename a file/dir. static int jffs2_ops_rename(cyg_mtab_entry * mte, cyg_dir dir1, - const char *name1, cyg_dir dir2, const char *name2) + const char *name1, cyg_dir dir2, const char *name2) { - unsigned long hash; - struct qstr this1, this2; - unsigned int c; - const char *hashname; - jffs2_dirsearch ds1, ds2; - int err; + unsigned long hash; + struct qstr this1, this2; + unsigned int c; + const char *hashname; + jffs2_dirsearch ds1, ds2; + int err; - D2(printf("jffs2_ops_rename\n")); + D2(printf("jffs2_ops_rename\n")); - init_dirsearch(&ds1, (struct inode *) dir1, name1); + init_dirsearch(&ds1, (struct inode *) dir1, name1); - err = jffs2_find(&ds1); + err = jffs2_find(&ds1); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - init_dirsearch(&ds2, (struct inode *) dir2, name2); + init_dirsearch(&ds2, (struct inode *) dir2, name2); - err = jffs2_find(&ds2); + err = jffs2_find(&ds2); - // Allow through renames to non-existent objects. - if (ds2.last && err == ENOENT) - ds2.node = NULL, err = ENOERR; + // Allow through renames to non-existent objects. + if (ds2.last && err == ENOENT) + ds2.node = NULL, err = ENOERR; - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // Null rename, just return - if (ds1.node == ds2.node) - return ENOERR; + // Null rename, just return + if (ds1.node == ds2.node) + return ENOERR; - hashname = ds1.name; - this1.name = hashname; - c = *(const unsigned char *) hashname; + hashname = ds1.name; + this1.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this1.len = hashname - (const char *) this1.name; - this1.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this1.len = hashname - (const char *) this1.name; + this1.hash = end_name_hash(hash); - hashname = ds2.name; - this2.name = hashname; - c = *(const unsigned char *) hashname; + hashname = ds2.name; + this2.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this2.len = hashname - (const char *) this2.name; - this2.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this2.len = hashname - (const char *) this2.name; + this2.hash = end_name_hash(hash); - // First deal with any entry that is at the destination - if (ds2.node) { - // Check that we are renaming like-for-like + // First deal with any entry that is at the destination + if (ds2.node) { + // Check that we are renaming like-for-like - if (!S_ISDIR(ds1.node->i_mode) && S_ISDIR(ds2.node->i_mode)) - return EISDIR; + if (!S_ISDIR(ds1.node->i_mode) && S_ISDIR(ds2.node->i_mode)) + return EISDIR; - if (S_ISDIR(ds1.node->i_mode) && !S_ISDIR(ds2.node->i_mode)) - return ENOTDIR; + if (S_ISDIR(ds1.node->i_mode) && !S_ISDIR(ds2.node->i_mode)) + return ENOTDIR; - // Now delete the destination directory entry + // Now delete the destination directory entry - err = jffs2_unlink(ds2.dir, ds2.node, &this2); + err = jffs2_unlink(ds2.dir, ds2.node, &this2); - if (err != 0) - return err; + if (err != 0) + return err; - } - // Now we know that there is no clashing node at the destination, - // make a new direntry at the destination and delete the old entry - // at the source. + } + // Now we know that there is no clashing node at the destination, + // make a new direntry at the destination and delete the old entry + // at the source. - err = jffs2_rename(ds1.dir, ds1.node, &this1, ds2.dir, &this2); + err = jffs2_rename(ds1.dir, ds1.node, &this1, ds2.dir, &this2); - // Update directory times - if (err == 0) - ds1.dir->i_ctime = - ds1.dir->i_mtime = - ds2.dir->i_ctime = ds2.dir->i_mtime = cyg_timestamp(); + // Update directory times + if (err == 0) + ds1.dir->i_ctime = + ds1.dir->i_mtime = + ds2.dir->i_ctime = ds2.dir->i_mtime = cyg_timestamp(); - return err; + return err; } // ------------------------------------------------------------------------- @@ -973,66 +980,66 @@ static int jffs2_ops_rename(cyg_mtab_ent // Make a new directory entry for a file. static int jffs2_ops_link(cyg_mtab_entry * mte, cyg_dir dir1, const char *name1, - cyg_dir dir2, const char *name2, int type) + cyg_dir dir2, const char *name2, int type) { - unsigned long hash; - struct qstr this; - unsigned int c; - const char *hashname; - jffs2_dirsearch ds1, ds2; - int err; + unsigned long hash; + struct qstr this; + unsigned int c; + const char *hashname; + jffs2_dirsearch ds1, ds2; + int err; - D2(printf("jffs2_ops_link\n")); + D2(printf("jffs2_ops_link\n")); - // Only do hard links for now in this filesystem - if (type != CYG_FSLINK_HARD) - return EINVAL; + // Only do hard links for now in this filesystem + if (type != CYG_FSLINK_HARD) + return EINVAL; - init_dirsearch(&ds1, (struct inode *) dir1, name1); + init_dirsearch(&ds1, (struct inode *) dir1, name1); - err = jffs2_find(&ds1); + err = jffs2_find(&ds1); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - init_dirsearch(&ds2, (struct inode *) dir2, name2); + init_dirsearch(&ds2, (struct inode *) dir2, name2); - err = jffs2_find(&ds2); + err = jffs2_find(&ds2); - // Don't allow links to existing objects - if (err == ENOERR) - return EEXIST; + // Don't allow links to existing objects + if (err == ENOERR) + return EEXIST; - // Allow through links to non-existing terminal objects - if (ds2.last && err == ENOENT) - ds2.node = NULL, err = ENOERR; + // Allow through links to non-existing terminal objects + if (ds2.last && err == ENOENT) + ds2.node = NULL, err = ENOERR; - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // Now we know that there is no existing node at the destination, - // make a new direntry at the destination. + // Now we know that there is no existing node at the destination, + // make a new direntry at the destination. - hashname = ds2.name; - this.name = hashname; - c = *(const unsigned char *) hashname; + hashname = ds2.name; + this.name = hashname; + c = *(const unsigned char *) hashname; - hash = init_name_hash(); - do { - hashname++; - hash = partial_name_hash(c, hash); - c = *(const unsigned char *) hashname; - } while (c && (c != '/')); - this.len = hashname - (const char *) this.name; - this.hash = end_name_hash(hash); + hash = init_name_hash(); + do { + hashname++; + hash = partial_name_hash(c, hash); + c = *(const unsigned char *) hashname; + } while (c && (c != '/')); + this.len = hashname - (const char *) this.name; + this.hash = end_name_hash(hash); - err = jffs2_link(ds1.node, ds2.dir, &this); + err = jffs2_link(ds1.node, ds2.dir, &this); - if (err == 0) - ds1.node->i_ctime = - ds2.dir->i_ctime = ds2.dir->i_mtime = cyg_timestamp(); + if (err == 0) + ds1.node->i_ctime = + ds2.dir->i_ctime = ds2.dir->i_mtime = cyg_timestamp(); - return err; + return err; } // ------------------------------------------------------------------------- @@ -1040,38 +1047,38 @@ static int jffs2_ops_link(cyg_mtab_entry // Open a directory for reading. static int jffs2_opendir(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - cyg_file * file) + cyg_file * file) { - jffs2_dirsearch ds; - int err; + jffs2_dirsearch ds; + int err; - D2(printf("jffs2_opendir\n")); + D2(printf("jffs2_opendir\n")); - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // check it is really a directory. - if (!S_ISDIR(ds.node->i_mode)) - return ENOTDIR; + // check it is really a directory. + if (!S_ISDIR(ds.node->i_mode)) + return ENOTDIR; - ds.node->i_count++; // Count successful open + ds.node->i_count++; // Count successful open - // Initialize the file object, setting the f_ops field to a - // special set of file ops. + // Initialize the file object, setting the f_ops field to a + // special set of file ops. - file->f_type = CYG_FILE_TYPE_FILE; - file->f_ops = &jffs2_dirops; - file->f_offset = 0; - file->f_data = (CYG_ADDRWORD) ds.node; - file->f_xops = 0; + file->f_type = CYG_FILE_TYPE_FILE; + file->f_ops = &jffs2_dirops; + file->f_offset = 0; + file->f_data = (CYG_ADDRWORD) ds.node; + file->f_xops = 0; - return ENOERR; + return ENOERR; } @@ -1080,48 +1087,48 @@ static int jffs2_opendir(cyg_mtab_entry // Change directory support. static int jffs2_chdir(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - cyg_dir * dir_out) + cyg_dir * dir_out) { - D2(printf("jffs2_chdir\n")); + D2(printf("jffs2_chdir\n")); - if (dir_out != NULL) { - // This is a request to get a new directory pointer in - // *dir_out. + if (dir_out != NULL) { + // This is a request to get a new directory pointer in + // *dir_out. - jffs2_dirsearch ds; - int err; + jffs2_dirsearch ds; + int err; - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // check it is a directory - if (!S_ISDIR(ds.node->i_mode)) - return ENOTDIR; + // check it is a directory + if (!S_ISDIR(ds.node->i_mode)) + return ENOTDIR; - // Increment ref count to keep this directory in existance - // while it is the current cdir. - ds.node->i_count++; + // Increment ref count to keep this directory in existance + // while it is the current cdir. + ds.node->i_count++; - // Pass it out - *dir_out = (cyg_dir) ds.node; - } else { - // If no output dir is required, this means that the mte and - // dir arguments are the current cdir setting and we should - // forget this fact. + // Pass it out + *dir_out = (cyg_dir) ds.node; + } else { + // If no output dir is required, this means that the mte and + // dir arguments are the current cdir setting and we should + // forget this fact. - struct inode *node = (struct inode *) dir; + struct inode *node = (struct inode *) dir; - // Just decrement directory reference count. - dec_refcnt(node); - } + // Just decrement directory reference count. + dec_refcnt(node); + } - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1129,37 +1136,37 @@ static int jffs2_chdir(cyg_mtab_entry * // Get struct stat info for named object. static int jffs2_stat(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - struct stat *buf) + struct stat *buf) { - jffs2_dirsearch ds; - int err; + jffs2_dirsearch ds; + int err; - D2(printf("jffs2_stat\n")); + D2(printf("jffs2_stat\n")); - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - // Fill in the status - buf->st_mode = ds.node->i_mode; - buf->st_ino = ds.node->i_ino; - buf->st_dev = 0; - buf->st_nlink = ds.node->i_nlink; - buf->st_uid = 0; - buf->st_gid = 0; - buf->st_size = ds.node->i_size; - buf->st_atime = ds.node->i_atime; - buf->st_mtime = ds.node->i_mtime; - buf->st_ctime = ds.node->i_ctime; + // Fill in the status + buf->st_mode = ds.node->i_mode; + buf->st_ino = ds.node->i_ino; + buf->st_dev = 0; + buf->st_nlink = ds.node->i_nlink; + buf->st_uid = 0; + buf->st_gid = 0; + buf->st_size = ds.node->i_size; + buf->st_atime = ds.node->i_atime; + buf->st_mtime = ds.node->i_mtime; + buf->st_ctime = ds.node->i_ctime; - return err; + return err; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1167,33 +1174,33 @@ static int jffs2_stat(cyg_mtab_entry * m // Getinfo. Currently only support pathconf(). static int jffs2_getinfo(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - int key, void *buf, int len) + int key, void *buf, int len) { - jffs2_dirsearch ds; - int err; + jffs2_dirsearch ds; + int err; - D2(printf("jffs2_getinfo\n")); + D2(printf("jffs2_getinfo\n")); - icache_evict((struct inode *) mte->root, (struct inode *) dir); + icache_evict((struct inode *) mte->root, (struct inode *) dir); - init_dirsearch(&ds, (struct inode *) dir, name); + init_dirsearch(&ds, (struct inode *) dir, name); - err = jffs2_find(&ds); + err = jffs2_find(&ds); - if (err != ENOERR) - return err; + if (err != ENOERR) + return err; - switch (key) { - case FS_INFO_CONF: - err = jffs2_pathconf(ds.node, (struct cyg_pathconf_info *) buf); - break; + switch (key) { + case FS_INFO_CONF: + err = jffs2_pathconf(ds.node, (struct cyg_pathconf_info *) buf); + break; - default: - err = EINVAL; - } - return err; + default: + err = EINVAL; + } + return err; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1201,13 +1208,13 @@ static int jffs2_getinfo(cyg_mtab_entry // Setinfo. Nothing to support here at present. static int jffs2_setinfo(cyg_mtab_entry * mte, cyg_dir dir, const char *name, - int key, void *buf, int len) + int key, void *buf, int len) { - // No setinfo keys supported at present + // No setinfo keys supported at present - D2(printf("jffs2_setinfo\n")); + D2(printf("jffs2_setinfo\n")); - return EINVAL; + return EINVAL; } //========================================================================== @@ -1219,50 +1226,50 @@ static int jffs2_setinfo(cyg_mtab_entry static int jffs2_fo_read(struct CYG_FILE_TAG *fp, struct CYG_UIO_TAG *uio) { - struct inode *inode = (struct inode *) fp->f_data; - struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); - struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb); - int i; - ssize_t resid = uio->uio_resid; - off_t pos = fp->f_offset; + struct inode *inode = (struct inode *) fp->f_data; + struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); + struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb); + int i; + ssize_t resid = uio->uio_resid; + off_t pos = fp->f_offset; - down(&f->sem); + down(&f->sem); - // Loop over the io vectors until there are none left - for (i = 0; i < uio->uio_iovcnt && pos < inode->i_size; i++) { - int ret; - cyg_iovec *iov = &uio->uio_iov[i]; - off_t len = min(iov->iov_len, inode->i_size - pos); + // Loop over the io vectors until there are none left + for (i = 0; i < uio->uio_iovcnt && pos < inode->i_size; i++) { + int ret; + cyg_iovec *iov = &uio->uio_iov[i]; + off_t len = min(iov->iov_len, inode->i_size - pos); - D2(printf("jffs2_fo_read inode size %d\n", inode->i_size)); + D2(printf("jffs2_fo_read inode size %d\n", inode->i_size)); - ret = - jffs2_read_inode_range(c, f, - (unsigned char *) iov->iov_base, pos, - len); - if (ret) { - D1(printf - ("jffs2_fo_read(): read_inode_range failed %d\n", - ret)); - uio->uio_resid = resid; - up(&f->sem); - return -ret; - } - resid -= len; - pos += len; - } + ret = + jffs2_read_inode_range(c, f, + (unsigned char *) iov->iov_base, pos, + len); + if (ret) { + D1(printf + ("jffs2_fo_read(): read_inode_range failed %d\n", + ret)); + uio->uio_resid = resid; + up(&f->sem); + return -ret; + } + resid -= len; + pos += len; + } - // We successfully read some data, update the node's access time - // and update the file offset and transfer residue. + // We successfully read some data, update the node's access time + // and update the file offset and transfer residue. - inode->i_atime = cyg_timestamp(); + inode->i_atime = cyg_timestamp(); - uio->uio_resid = resid; - fp->f_offset = pos; + uio->uio_resid = resid; + fp->f_offset = pos; - up(&f->sem); + up(&f->sem); - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1271,98 +1278,98 @@ static int jffs2_fo_read(struct CYG_FILE static int jffs2_fo_write(struct CYG_FILE_TAG *fp, struct CYG_UIO_TAG *uio) { - struct page write_page; - off_t page_start_pos; - struct inode *node = (struct inode *) fp->f_data; - off_t pos = fp->f_offset; - ssize_t resid = uio->uio_resid; - int i; + struct page write_page; + off_t page_start_pos; + struct inode *node = (struct inode *) fp->f_data; + off_t pos = fp->f_offset; + ssize_t resid = uio->uio_resid; + int i; - memset(&read_write_buffer, 0, PAGE_CACHE_SIZE); - write_page.virtual = &read_write_buffer; + memset(&read_write_buffer, 0, PAGE_CACHE_SIZE); + write_page.virtual = &read_write_buffer; - // If the APPEND mode bit was supplied, force all writes to - // the end of the file. - if (fp->f_flag & CYG_FAPPEND) - pos = fp->f_offset = node->i_size; + // If the APPEND mode bit was supplied, force all writes to + // the end of the file. + if (fp->f_flag & CYG_FAPPEND) + pos = fp->f_offset = node->i_size; - // Check that pos is within current file size, or at the very end. - if (pos < 0 || pos > node->i_size) - return EINVAL; + // Check that pos is within current file size, or at the very end. + if (pos < 0 || pos > node->i_size) + return EINVAL; - // Now loop over the iovecs until they are all done, or - // we get an error. - for (i = 0; i < uio->uio_iovcnt; i++) { - cyg_iovec *iov = &uio->uio_iov[i]; - char *buf = (char *) iov->iov_base; - off_t len = iov->iov_len; + // Now loop over the iovecs until they are all done, or + // we get an error. + for (i = 0; i < uio->uio_iovcnt; i++) { + cyg_iovec *iov = &uio->uio_iov[i]; + char *buf = (char *) iov->iov_base; + off_t len = iov->iov_len; - // loop over the vector writing it to the file until it has - // all been done. - while (len > 0) { - //cyg_uint8 *fbuf; - //size_t bsize; - size_t writtenlen; - off_t l = len; - int err; + // loop over the vector writing it to the file until it has + // all been done. + while (len > 0) { + //cyg_uint8 *fbuf; + //size_t bsize; + size_t writtenlen; + off_t l = len; + int err; - write_page.index = 0; + write_page.index = 0; - page_start_pos = pos; - while (page_start_pos >= (PAGE_CACHE_SIZE)) { - write_page.index++; - page_start_pos -= PAGE_CACHE_SIZE; - } + page_start_pos = pos; + while (page_start_pos >= (PAGE_CACHE_SIZE)) { + write_page.index++; + page_start_pos -= PAGE_CACHE_SIZE; + } - if (l > PAGE_CACHE_SIZE - page_start_pos) - l = PAGE_CACHE_SIZE - page_start_pos; + if (l > PAGE_CACHE_SIZE - page_start_pos) + l = PAGE_CACHE_SIZE - page_start_pos; - D2(printf - ("jffs2_fo_write write_page.index %d\n", - write_page.index)); - D2(printf - ("jffs2_fo_write page_start_pos %d\n", - page_start_pos)); - D2(printf("jffs2_fo_write transfer size %d\n", l)); + D2(printf + ("jffs2_fo_write write_page.index %d\n", + write_page.index)); + D2(printf + ("jffs2_fo_write page_start_pos %d\n", + page_start_pos)); + D2(printf("jffs2_fo_write transfer size %d\n", l)); - err = - jffs2_prepare_write(node, &write_page, - page_start_pos, - page_start_pos + l); + err = + jffs2_prepare_write(node, &write_page, + page_start_pos, + page_start_pos + l); - if (err != 0) - return err; + if (err != 0) + return err; - // copy data in - memcpy(&read_write_buffer[page_start_pos], buf, l); + // copy data in + memcpy(&read_write_buffer[page_start_pos], buf, l); - writtenlen = - jffs2_commit_write(node, &write_page, - page_start_pos, - page_start_pos + l); + writtenlen = + jffs2_commit_write(node, &write_page, + page_start_pos, + page_start_pos + l); - if (writtenlen != l) - return ENOSPC; + if (writtenlen != l) + return ENOSPC; - // Update working vars - len -= l; - buf += l; - pos += l; - resid -= l; - } - } + // Update working vars + len -= l; + buf += l; + pos += l; + resid -= l; + } + } - // We wrote some data successfully, update the modified and access - // times of the node, increase its size appropriately, and update - // the file offset and transfer residue. - node->i_mtime = node->i_ctime = cyg_timestamp(); - if (pos > node->i_size) - node->i_size = pos; + // We wrote some data successfully, update the modified and access + // times of the node, increase its size appropriately, and update + // the file offset and transfer residue. + node->i_mtime = node->i_ctime = cyg_timestamp(); + if (pos > node->i_size) + node->i_size = pos; - uio->uio_resid = resid; - fp->f_offset = pos; + uio->uio_resid = resid; + fp->f_offset = pos; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1371,39 +1378,39 @@ static int jffs2_fo_write(struct CYG_FIL static int jffs2_fo_lseek(struct CYG_FILE_TAG *fp, off_t * apos, int whence) { - struct inode *node = (struct inode *) fp->f_data; - off_t pos = *apos; + struct inode *node = (struct inode *) fp->f_data; + off_t pos = *apos; - D2(printf("jffs2_fo_lseek\n")); + D2(printf("jffs2_fo_lseek\n")); - switch (whence) { - case SEEK_SET: - // Pos is already where we want to be. - break; + switch (whence) { + case SEEK_SET: + // Pos is already where we want to be. + break; - case SEEK_CUR: - // Add pos to current offset. - pos += fp->f_offset; - break; + case SEEK_CUR: + // Add pos to current offset. + pos += fp->f_offset; + break; - case SEEK_END: - // Add pos to file size. - pos += node->i_size; - break; + case SEEK_END: + // Add pos to file size. + pos += node->i_size; + break; - default: - return EINVAL; - } + default: + return EINVAL; + } - // Check that pos is still within current file size, or at the - // very end. - if (pos < 0 || pos > node->i_size) - return EINVAL; + // Check that pos is still within current file size, or at the + // very end. + if (pos < 0 || pos > node->i_size) + return EINVAL; - // All OK, set fp offset and return new position. - *apos = fp->f_offset = pos; + // All OK, set fp offset and return new position. + *apos = fp->f_offset = pos; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1411,13 +1418,13 @@ static int jffs2_fo_lseek(struct CYG_FIL // Handle ioctls. Currently none are defined. static int jffs2_fo_ioctl(struct CYG_FILE_TAG *fp, CYG_ADDRWORD com, - CYG_ADDRWORD data) + CYG_ADDRWORD data) { - // No Ioctls currenly defined. + // No Ioctls currenly defined. - D2(printf("jffs2_fo_ioctl\n")); + D2(printf("jffs2_fo_ioctl\n")); - return EINVAL; + return EINVAL; } // ------------------------------------------------------------------------- @@ -1426,12 +1433,12 @@ static int jffs2_fo_ioctl(struct CYG_FIL static int jffs2_fo_fsync(struct CYG_FILE_TAG *fp, int mode) { - // Data is always permanently where it belongs, nothing to do - // here. + // Data is always permanently where it belongs, nothing to do + // here. - D2(printf("jffs2_fo_fsync\n")); + D2(printf("jffs2_fo_fsync\n")); - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1441,15 +1448,15 @@ static int jffs2_fo_fsync(struct CYG_FIL static int jffs2_fo_close(struct CYG_FILE_TAG *fp) { - struct inode *node = (struct inode *) fp->f_data; + struct inode *node = (struct inode *) fp->f_data; - D2(printf("jffs2_fo_close\n")); + D2(printf("jffs2_fo_close\n")); - dec_refcnt(node); + dec_refcnt(node); - fp->f_data = 0; // zero data pointer + fp->f_data = 0; // zero data pointer - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1458,23 +1465,23 @@ static int jffs2_fo_close(struct CYG_FIL static int jffs2_fo_fstat(struct CYG_FILE_TAG *fp, struct stat *buf) { - struct inode *node = (struct inode *) fp->f_data; + struct inode *node = (struct inode *) fp->f_data; - D2(printf("jffs2_fo_fstat\n")); + D2(printf("jffs2_fo_fstat\n")); - // Fill in the status - buf->st_mode = node->i_mode; - buf->st_ino = node->i_ino; - buf->st_dev = 0; - buf->st_nlink = node->i_nlink; - buf->st_uid = 0; - buf->st_gid = 0; - buf->st_size = node->i_size; - buf->st_atime = node->i_atime; - buf->st_mtime = node->i_mtime; - buf->st_ctime = node->i_ctime; + // Fill in the status + buf->st_mode = node->i_mode; + buf->st_ino = node->i_ino; + buf->st_dev = 0; + buf->st_nlink = node->i_nlink; + buf->st_uid = 0; + buf->st_gid = 0; + buf->st_size = node->i_size; + buf->st_atime = node->i_atime; + buf->st_mtime = node->i_mtime; + buf->st_ctime = node->i_ctime; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1482,24 +1489,24 @@ static int jffs2_fo_fstat(struct CYG_FIL // Get info. Currently only supports fpathconf(). static int jffs2_fo_getinfo(struct CYG_FILE_TAG *fp, int key, void *buf, - int len) + int len) { - struct inode *node = (struct inode *) fp->f_data; - int err; + struct inode *node = (struct inode *) fp->f_data; + int err; - D2(printf("jffs2_fo_getinfo\n")); + D2(printf("jffs2_fo_getinfo\n")); - switch (key) { - case FS_INFO_CONF: - err = jffs2_pathconf(node, (struct cyg_pathconf_info *) buf); - break; + switch (key) { + case FS_INFO_CONF: + err = jffs2_pathconf(node, (struct cyg_pathconf_info *) buf); + break; - default: - err = EINVAL; - } - return err; + default: + err = EINVAL; + } + return err; - return ENOERR; + return ENOERR; } // ------------------------------------------------------------------------- @@ -1507,13 +1514,13 @@ static int jffs2_fo_getinfo(struct CYG_F // Set info. Nothing supported here. static int jffs2_fo_setinfo(struct CYG_FILE_TAG *fp, int key, void *buf, - int len) + int len) { - // No setinfo key supported at present + // No setinfo key supported at present - D2(printf("jffs2_fo_setinfo\n")); + D2(printf("jffs2_fo_setinfo\n")); - return ENOERR; + return ENOERR; } //========================================================================== @@ -1525,83 +1532,83 @@ static int jffs2_fo_setinfo(struct CYG_F static __inline void filldir(char *nbuf, int nlen, const char *name, int namlen) { - int len = nlen < namlen ? nlen : namlen; - memcpy(nbuf, name, len); - nbuf[len] = '\0'; + int len = nlen < namlen ? nlen : namlen; + memcpy(nbuf, name, len); + nbuf[len] = '\0'; } static int jffs2_fo_dirread(struct CYG_FILE_TAG *fp, struct CYG_UIO_TAG *uio) { - struct inode *d_inode = (struct inode *) fp->f_data; - struct dirent *ent = (struct dirent *) uio->uio_iov[0].iov_base; - char *nbuf = ent->d_name; - int nlen = sizeof (ent->d_name) - 1; - off_t len = uio->uio_iov[0].iov_len; - struct jffs2_inode_info *f; - struct jffs2_sb_info *c; - struct inode *inode = d_inode; - struct jffs2_full_dirent *fd; - unsigned long offset, curofs; - int found = 1; + struct inode *d_inode = (struct inode *) fp->f_data; + struct dirent *ent = (struct dirent *) uio->uio_iov[0].iov_base; + char *nbuf = ent->d_name; + int nlen = sizeof (ent->d_name) - 1; + off_t len = uio->uio_iov[0].iov_len; + struct jffs2_inode_info *f; + struct jffs2_sb_info *c; + struct inode *inode = d_inode; + struct jffs2_full_dirent *fd; + unsigned long offset, curofs; + int found = 1; - if (len < sizeof (struct dirent)) - return EINVAL; + if (len < sizeof (struct dirent)) + return EINVAL; - D1(printk - (KERN_DEBUG "jffs2_readdir() for dir_i #%lu\n", d_inode->i_ino)); + D1(printk + (KERN_DEBUG "jffs2_readdir() for dir_i #%lu\n", d_inode->i_ino)); - f = JFFS2_INODE_INFO(inode); - c = JFFS2_SB_INFO(inode->i_sb); + f = JFFS2_INODE_INFO(inode); + c = JFFS2_SB_INFO(inode->i_sb); - offset = fp->f_offset; + offset = fp->f_offset; - if (offset == 0) { - D1(printk - (KERN_DEBUG "Dirent 0: \".\", ino #%lu\n", inode->i_ino)); - filldir(nbuf, nlen, ".", 1); - goto out; - } - if (offset == 1) { - filldir(nbuf, nlen, "..", 2); - goto out; - } + if (offset == 0) { + D1(printk + (KERN_DEBUG "Dirent 0: \".\", ino #%lu\n", inode->i_ino)); + filldir(nbuf, nlen, ".", 1); + goto out; + } + if (offset == 1) { + filldir(nbuf, nlen, "..", 2); + goto out; + } - curofs = 1; - down(&f->sem); - for (fd = f->dents; fd; fd = fd->next) { + curofs = 1; + down(&f->sem); + for (fd = f->dents; fd; fd = fd->next) { - curofs++; - /* First loop: curofs = 2; offset = 2 */ - if (curofs < offset) { - D2(printk - (KERN_DEBUG - "Skipping dirent: \"%s\", ino #%u, type %d, because curofs %ld < offset %ld\n", - fd->name, fd->ino, fd->type, curofs, offset)); - continue; - } - if (!fd->ino) { - D2(printk - (KERN_DEBUG "Skipping deletion dirent \"%s\"\n", - fd->name)); - offset++; - continue; - } - D2(printk - (KERN_DEBUG "Dirent %ld: \"%s\", ino #%u, type %d\n", offset, - fd->name, fd->ino, fd->type)); - filldir(nbuf, nlen, fd->name, strlen(fd->name)); - goto out_sem; - } - /* Reached the end of the directory */ - found = 0; + curofs++; + /* First loop: curofs = 2; offset = 2 */ + if (curofs < offset) { + D2(printk + (KERN_DEBUG + "Skipping dirent: \"%s\", ino #%u, type %d, because curofs %ld < offset %ld\n", + fd->name, fd->ino, fd->type, curofs, offset)); + continue; + } + if (!fd->ino) { + D2(printk + (KERN_DEBUG "Skipping deletion dirent \"%s\"\n", + fd->name)); + offset++; + continue; + } + D2(printk + (KERN_DEBUG "Dirent %ld: \"%s\", ino #%u, type %d\n", offset, + fd->name, fd->ino, fd->type)); + filldir(nbuf, nlen, fd->name, strlen(fd->name)); + goto out_sem; + } + /* Reached the end of the directory */ + found = 0; out_sem: - up(&f->sem); + up(&f->sem); out: - fp->f_offset = ++offset; - if (found) { - uio->uio_resid -= sizeof (struct dirent); - } - return ENOERR; + fp->f_offset = ++offset; + if (found) { + uio->uio_resid -= sizeof (struct dirent); + } + return ENOERR; } // ------------------------------------------------------------------------- @@ -1610,16 +1617,16 @@ static int jffs2_fo_dirread(struct CYG_F static int jffs2_fo_dirlseek(struct CYG_FILE_TAG *fp, off_t * pos, int whence) { - // Only allow SEEK_SET to zero + // Only allow SEEK_SET to zero - D2(printf("jffs2_fo_dirlseek\n")); + D2(printf("jffs2_fo_dirlseek\n")); - if (whence != SEEK_SET || *pos != 0) - return EINVAL; + if (whence != SEEK_SET || *pos != 0) + return EINVAL; - *pos = fp->f_offset = 0; + *pos = fp->f_offset = 0; - return ENOERR; + return ENOERR; } //========================================================================== @@ -1631,167 +1638,167 @@ static int jffs2_fo_dirlseek(struct CYG_ //========================================================================== struct page *read_cache_page(unsigned long index, - int (*filler) (void *, struct page *), void *data) + int (*filler) (void *, struct page *), void *data) { - // Only called in gc.c jffs2_garbage_collect_dnode - // but gets a real page for the specified inode + // Only called in gc.c jffs2_garbage_collect_dnode + // but gets a real page for the specified inode - int err; - struct page *gc_page = malloc(sizeof (struct page)); + int err; + struct page *gc_page = malloc(sizeof (struct page)); - printf("read_cache_page\n"); - memset(&gc_buffer, 0, PAGE_CACHE_SIZE); + printf("read_cache_page\n"); + memset(&gc_buffer, 0, PAGE_CACHE_SIZE); - if (gc_page != NULL) { - gc_page->virtual = &gc_buffer; - gc_page->index = index; + if (gc_page != NULL) { + gc_page->virtual = &gc_buffer; + gc_page->index = index; - err = filler(data, gc_page); - if (err < 0) { - free(gc_page); - gc_page = NULL; - } - } + err = filler(data, gc_page); + if (err < 0) { + free(gc_page); + gc_page = NULL; + } + } - return gc_page; + return gc_page; } void page_cache_release(struct page *pg) { - // Only called in gc.c jffs2_garbage_collect_dnode - // but should free the page malloc'd by read_cache_page + // Only called in gc.c jffs2_garbage_collect_dnode + // but should free the page malloc'd by read_cache_page - printf("page_cache_release\n"); - free(pg); + printf("page_cache_release\n"); + free(pg); } struct inode *new_inode(struct super_block *sb) { - // Only called in write.c jffs2_new_inode - // Always adds itself to inode cache + // Only called in write.c jffs2_new_inode + // Always adds itself to inode cache - struct inode *inode; - struct inode *cached_inode; + struct inode *inode; + struct inode *cached_inode; - inode = malloc(sizeof (struct inode)); - if (inode == NULL) - return 0; + inode = malloc(sizeof (struct inode)); + if (inode == NULL) + return 0; - D2(printf - ("malloc new_inode %x ####################################\n", - inode)); + D2(printf + ("malloc new_inode %x ####################################\n", + inode)); - memset(inode, 0, sizeof (struct inode)); - inode->i_sb = sb; - inode->i_ino = 1; - inode->i_count = 0; //1; // Let ecos manage the open count + memset(inode, 0, sizeof (struct inode)); + inode->i_sb = sb; + inode->i_ino = 1; + inode->i_count = 0; //1; // Let ecos manage the open count - inode->i_nlink = 1; // Let JFFS2 manage the link count - inode->i_size = 0; + inode->i_nlink = 1; // Let JFFS2 manage the link count + inode->i_size = 0; - inode->i_cache_next = NULL; // Newest inode, about to be cached + inode->i_cache_next = NULL; // Newest inode, about to be cached - // Add to the icache - for (cached_inode = sb->s_root; cached_inode != NULL; - cached_inode = cached_inode->i_cache_next) { - if (cached_inode->i_cache_next == NULL) { - cached_inode->i_cache_next = inode; // Current last in cache points to newcomer - inode->i_cache_prev = cached_inode; // Newcomer points back to last - break; - } - } + // Add to the icache + for (cached_inode = sb->s_root; cached_inode != NULL; + cached_inode = cached_inode->i_cache_next) { + if (cached_inode->i_cache_next == NULL) { + cached_inode->i_cache_next = inode; // Current last in cache points to newcomer + inode->i_cache_prev = cached_inode; // Newcomer points back to last + break; + } + } - return inode; + return inode; } struct inode *ilookup(struct super_block *sb, cyg_uint32 ino) { - struct inode *inode = NULL; + struct inode *inode = NULL; - D2(printf("ilookup\n")); - // Check for this inode in the cache - for (inode = sb->s_root; inode != NULL; inode = inode->i_cache_next) { - if (inode->i_ino == ino) { - inode->i_count++; - break; - } - } + D2(printf("ilookup\n")); + // Check for this inode in the cache + for (inode = sb->s_root; inode != NULL; inode = inode->i_cache_next) { + if (inode->i_ino == ino) { + inode->i_count++; + break; + } + } - return inode; + return inode; } struct inode *iget(struct super_block *sb, cyg_uint32 ino) { - // Substitute for iget drops straight through to reading the - // inode from disk if it is not in the inode cache + // Substitute for iget drops straight through to reading the + // inode from disk if it is not in the inode cache - // Called in super.c jffs2_read_super, dir.c jffs2_lookup, - // and gc.c jffs2_garbage_collect_pass + // Called in super.c jffs2_read_super, dir.c jffs2_lookup, + // and gc.c jffs2_garbage_collect_pass - // Must first check for cached inode - // If this fails let new_inode create one + // Must first check for cached inode + // If this fails let new_inode create one - struct inode *inode; + struct inode *inode; - D2(printf("iget\n")); + D2(printf("iget\n")); - inode = ilookup(sb, ino); - if (inode) - return inode; + inode = ilookup(sb, ino); + if (inode) + return inode; - // Not cached, so malloc it - inode = new_inode(sb); - if (inode == NULL) - return 0; + // Not cached, so malloc it + inode = new_inode(sb); + if (inode == NULL) + return 0; - inode->i_ino = ino; - jffs2_read_inode(inode); - inode->i_count = 1; - return inode; + inode->i_ino = ino; + jffs2_read_inode(inode); + inode->i_count = 1; + return inode; } void iput(struct inode *i) { - // Called in dec_refcnt, jffs2_find - // (and jffs2_open and jffs2_ops_mkdir?) - // super.c jffs2_read_super, - // and gc.c jffs2_garbage_collect_pass + // Called in dec_refcnt, jffs2_find + // (and jffs2_open and jffs2_ops_mkdir?) + // super.c jffs2_read_super, + // and gc.c jffs2_garbage_collect_pass - struct inode *cached_inode; + struct inode *cached_inode; - D2(printf - ("free iput inode %x $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$\n", i)); - if (i && i->i_count) { - /* Added by dwmw2. iget/iput in Linux track the use count, - don't just unconditionally free it */ - printf("iput called for used inode\n"); - return; - } - if (i != NULL) { - // Remove from the icache - for (cached_inode = i->i_sb->s_root; cached_inode != NULL; - cached_inode = cached_inode->i_cache_next) { - if (cached_inode == i) { + D2(printf + ("free iput inode %x $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$\n", i)); + if (i && i->i_count) { + /* Added by dwmw2. iget/iput in Linux track the use count, + don't just unconditionally free it */ + printf("iput called for used inode\n"); + return; + } + if (i != NULL) { + // Remove from the icache + for (cached_inode = i->i_sb->s_root; cached_inode != NULL; + cached_inode = cached_inode->i_cache_next) { + if (cached_inode == i) { if(cached_inode->i_cache_prev != NULL) { - cached_inode->i_cache_prev->i_cache_next = cached_inode->i_cache_next; // Previous entry points ahead of us - if (cached_inode->i_cache_next != NULL) - cached_inode->i_cache_next->i_cache_prev = cached_inode->i_cache_prev; // Next entry points behind us + cached_inode->i_cache_prev->i_cache_next = cached_inode->i_cache_next; // Previous entry points ahead of us + if (cached_inode->i_cache_next != NULL) + cached_inode->i_cache_next->i_cache_prev = cached_inode->i_cache_prev; // Next entry points behind us } - break; - } - } - // inode has been seperated from the cache - jffs2_clear_inode(i); - free(i); - } + break; + } + } + // inode has been seperated from the cache + jffs2_clear_inode(i); + free(i); + } } static int return_EIO(void) { - return -EIO; + return -EIO; } #define EIO_ERROR ((void *) (return_EIO)) @@ -1799,173 +1806,173 @@ static int return_EIO(void) void make_bad_inode(struct inode *inode) { - // In readinode.c JFFS2 checks whether the inode has appropriate - // content for its marked type + // In readinode.c JFFS2 checks whether the inode has appropriate + // content for its marked type - D2(printf("make_bad_inode\n")); + D2(printf("make_bad_inode\n")); - inode->i_mode = S_IFREG; - inode->i_atime = inode->i_mtime = inode->i_ctime = cyg_timestamp(); - inode->i_op = EIO_ERROR; - inode->i_fop = EIO_ERROR; + inode->i_mode = S_IFREG; + inode->i_atime = inode->i_mtime = inode->i_ctime = cyg_timestamp(); + inode->i_op = EIO_ERROR; + inode->i_fop = EIO_ERROR; } int is_bad_inode(struct inode *inode) { - // Called in super.c jffs2_read_super, - // and gc.c jffs2_garbage_collect_pass + // Called in super.c jffs2_read_super, + // and gc.c jffs2_garbage_collect_pass - D2(printf("is_bad_inode\n")); + D2(printf("is_bad_inode\n")); - return (inode->i_op == EIO_ERROR); - /*if(i == NULL) - return 1; - return 0; */ + return (inode->i_op == EIO_ERROR); + /*if(i == NULL) + return 1; + return 0; */ } cyg_bool jffs2_flash_read(struct jffs2_sb_info * c, - cyg_uint32 read_buffer_offset, const size_t size, - size_t * return_size, char *write_buffer) + cyg_uint32 read_buffer_offset, const size_t size, + size_t * return_size, char *write_buffer) { - Cyg_ErrNo err; - cyg_uint32 len = size; - struct super_block *sb = OFNI_BS_2SFFJ(c); + Cyg_ErrNo err; + cyg_uint32 len = size; + struct super_block *sb = OFNI_BS_2SFFJ(c); - //D2(printf("FLASH READ\n")); - //D2(printf("read address = %x\n", CYGNUM_FS_JFFS2_BASE_ADDRESS + read_buffer_offset)); - //D2(printf("write address = %x\n", write_buffer)); - //D2(printf("size = %x\n", size)); - err = cyg_io_bread(sb->s_dev, write_buffer, &len, read_buffer_offset); + //D2(printf("FLASH READ\n")); + //D2(printf("read address = %x\n", CYGNUM_FS_JFFS2_BASE_ADDRESS + read_buffer_offset)); + //D2(printf("write address = %x\n", write_buffer)); + //D2(printf("size = %x\n", size)); + err = cyg_io_bread(sb->s_dev, write_buffer, &len, read_buffer_offset); - *return_size = (size_t) len; - return ((err == ENOERR) ? ENOERR : -EIO); + *return_size = (size_t) len; + return ((err == ENOERR) ? ENOERR : -EIO); } cyg_bool jffs2_flash_write(struct jffs2_sb_info * c, - cyg_uint32 write_buffer_offset, const size_t size, - size_t * return_size, char *read_buffer) + cyg_uint32 write_buffer_offset, const size_t size, + size_t * return_size, char *read_buffer) { - Cyg_ErrNo err; - cyg_uint32 len = size; - struct super_block *sb = OFNI_BS_2SFFJ(c); + Cyg_ErrNo err; + cyg_uint32 len = size; + struct super_block *sb = OFNI_BS_2SFFJ(c); - // D2(printf("FLASH WRITE ENABLED!!!\n")); - // D2(printf("write address = %x\n", CYGNUM_FS_JFFS2_BASE_ADDRESS + write_buffer_offset)); - // D2(printf("read address = %x\n", read_buffer)); - // D2(printf("size = %x\n", size)); + // D2(printf("FLASH WRITE ENABLED!!!\n")); + // D2(printf("write address = %x\n", CYGNUM_FS_JFFS2_BASE_ADDRESS + write_buffer_offset)); + // D2(printf("read address = %x\n", read_buffer)); + // D2(printf("size = %x\n", size)); - err = cyg_io_bwrite(sb->s_dev, read_buffer, &len, write_buffer_offset); - *return_size = (size_t) len; + err = cyg_io_bwrite(sb->s_dev, read_buffer, &len, write_buffer_offset); + *return_size = (size_t) len; - return ((err == ENOERR) ? ENOERR : -EIO); + return ((err == ENOERR) ? ENOERR : -EIO); } int jffs2_flash_direct_writev(struct jffs2_sb_info *c, const struct iovec *vecs, - unsigned long count, loff_t to, size_t * retlen) + unsigned long count, loff_t to, size_t * retlen) { - unsigned long i; - size_t totlen = 0, thislen; - int ret = 0; + unsigned long i; + size_t totlen = 0, thislen; + int ret = 0; - for (i = 0; i < count; i++) { - // writes need to be aligned but the data we're passed may not be - // Observation suggests most unaligned writes are small, so we - // optimize for that case. + for (i = 0; i < count; i++) { + // writes need to be aligned but the data we're passed may not be + // Observation suggests most unaligned writes are small, so we + // optimize for that case. - if (((vecs[i].iov_len & (sizeof (int) - 1))) || - (((unsigned long) vecs[i]. - iov_base & (sizeof (unsigned long) - 1)))) { - // are there iov's after this one? Or is it so much we'd need - // to do multiple writes anyway? - if ((i + 1) < count || vecs[i].iov_len > 256) { - // cop out and malloc - unsigned long j; - ssize_t sizetomalloc = 0, totvecsize = 0; - char *cbuf, *cbufptr; + if (((vecs[i].iov_len & (sizeof (int) - 1))) || + (((unsigned long) vecs[i]. + iov_base & (sizeof (unsigned long) - 1)))) { + // are there iov's after this one? Or is it so much we'd need + // to do multiple writes anyway? + if ((i + 1) < count || vecs[i].iov_len > 256) { + // cop out and malloc + unsigned long j; + ssize_t sizetomalloc = 0, totvecsize = 0; + char *cbuf, *cbufptr; - for (j = i; j < count; j++) - totvecsize += vecs[j].iov_len; + for (j = i; j < count; j++) + totvecsize += vecs[j].iov_len; - // pad up in case unaligned - sizetomalloc = totvecsize + sizeof (int) - 1; - sizetomalloc &= ~(sizeof (int) - 1); - cbuf = (char *) malloc(sizetomalloc); - // malloc returns aligned memory - if (!cbuf) { - ret = -ENOMEM; - goto writev_out; - } - cbufptr = cbuf; - for (j = i; j < count; j++) { - memcpy(cbufptr, vecs[j].iov_base, - vecs[j].iov_len); - cbufptr += vecs[j].iov_len; - } - ret = - jffs2_flash_write(c, to, sizetomalloc, - &thislen, cbuf); - if (thislen > totvecsize) // in case it was aligned up - thislen = totvecsize; - totlen += thislen; - free(cbuf); - goto writev_out; - } else { - // otherwise optimize for the common case - int buf[256 / sizeof (int)]; // int, so int aligned - size_t lentowrite; + // pad up in case unaligned + sizetomalloc = totvecsize + sizeof (int) - 1; + sizetomalloc &= ~(sizeof (int) - 1); + cbuf = (char *) malloc(sizetomalloc); + // malloc returns aligned memory + if (!cbuf) { + ret = -ENOMEM; + goto writev_out; + } + cbufptr = cbuf; + for (j = i; j < count; j++) { + memcpy(cbufptr, vecs[j].iov_base, + vecs[j].iov_len); + cbufptr += vecs[j].iov_len; + } + ret = + jffs2_flash_write(c, to, sizetomalloc, + &thislen, cbuf); + if (thislen > totvecsize) // in case it was aligned up + thislen = totvecsize; + totlen += thislen; + free(cbuf); + goto writev_out; + } else { + // otherwise optimize for the common case + int buf[256 / sizeof (int)]; // int, so int aligned + size_t lentowrite; - lentowrite = vecs[i].iov_len; - // pad up in case its unaligned - lentowrite += sizeof (int) - 1; - lentowrite &= ~(sizeof (int) - 1); - memcpy(buf, vecs[i].iov_base, lentowrite); + lentowrite = vecs[i].iov_len; + // pad up in case its unaligned + lentowrite += sizeof (int) - 1; + lentowrite &= ~(sizeof (int) - 1); + memcpy(buf, vecs[i].iov_base, lentowrite); - ret = - jffs2_flash_write(c, to, lentowrite, - &thislen, (char *) &buf); - if (thislen > vecs[i].iov_len) - thislen = vecs[i].iov_len; - } // else - } else - ret = - jffs2_flash_write(c, to, vecs[i].iov_len, &thislen, - vecs[i].iov_base); - totlen += thislen; - if (ret || thislen != vecs[i].iov_len) - break; - to += vecs[i].iov_len; - } + ret = + jffs2_flash_write(c, to, lentowrite, + &thislen, (char *) &buf); + if (thislen > vecs[i].iov_len) + thislen = vecs[i].iov_len; + } // else + } else + ret = + jffs2_flash_write(c, to, vecs[i].iov_len, &thislen, + vecs[i].iov_base); + totlen += thislen; + if (ret || thislen != vecs[i].iov_len) + break; + to += vecs[i].iov_len; + } writev_out: - if (retlen) - *retlen = totlen; + if (retlen) + *retlen = totlen; - return ret; + return ret; } cyg_bool jffs2_flash_erase(struct jffs2_sb_info * c, - struct jffs2_eraseblock * jeb) + struct jffs2_eraseblock * jeb) { - cyg_io_flash_getconfig_erase_t e; - void *err_addr; - Cyg_ErrNo err; - cyg_uint32 len = sizeof (e); - struct super_block *sb = OFNI_BS_2SFFJ(c); + cyg_io_flash_getconfig_erase_t e; + void *err_addr; + Cyg_ErrNo err; + cyg_uint32 len = sizeof (e); + struct super_block *sb = OFNI_BS_2SFFJ(c); - e.offset = jeb->offset; - e.len = c->sector_size; - e.err_address = &err_addr; + e.offset = jeb->offset; + e.len = c->sector_size; + e.err_address = &err_addr; - // D2(printf("FLASH ERASE ENABLED!!!\n")); - // D2(printf("erase address = %x\n", CYGNUM_FS_JFFS2_BASE_ADDRESS + jeb->offset)); - // D2(printf("size = %x\n", c->sector_size)); + // D2(printf("FLASH ERASE ENABLED!!!\n")); + // D2(printf("erase address = %x\n", CYGNUM_FS_JFFS2_BASE_ADDRESS + jeb->offset)); + // D2(printf("size = %x\n", c->sector_size)); - err = cyg_io_get_config(sb->s_dev, CYG_IO_GET_CONFIG_FLASH_ERASE, - &e, &len); + err = cyg_io_get_config(sb->s_dev, CYG_IO_GET_CONFIG_FLASH_ERASE, + &e, &len); - return (err != ENOERR || e.flasherr != 0); + return (err != ENOERR || e.flasherr != 0); } // ------------------------------------------------------------------------- @@ -1988,81 +1995,81 @@ void jffs2_clear_inode (struct inode *in fill in the raw_inode while you're at it. */ struct inode *jffs2_new_inode (struct inode *dir_i, int mode, struct jffs2_raw_inode *ri) { - struct inode *inode; - struct super_block *sb = dir_i->i_sb; - struct jffs2_sb_info *c; - struct jffs2_inode_info *f; - int ret; + struct inode *inode; + struct super_block *sb = dir_i->i_sb; + struct jffs2_sb_info *c; + struct jffs2_inode_info *f; + int ret; - D1(printk(KERN_DEBUG "jffs2_new_inode(): dir_i %ld, mode 0x%x\n", dir_i->i_ino, mode)); + D1(printk(KERN_DEBUG "jffs2_new_inode(): dir_i %ld, mode 0x%x\n", dir_i->i_ino, mode)); - c = JFFS2_SB_INFO(sb); - - inode = new_inode(sb); - - if (!inode) - return ERR_PTR(-ENOMEM); + c = JFFS2_SB_INFO(sb); + + inode = new_inode(sb); + + if (!inode) + return ERR_PTR(-ENOMEM); - f = JFFS2_INODE_INFO(inode); - jffs2_init_inode_info(f); + f = JFFS2_INODE_INFO(inode); + jffs2_init_inode_info(f); - memset(ri, 0, sizeof(*ri)); - /* Set OS-specific defaults for new inodes */ - ri->uid = ri->gid = cpu_to_je16(0); - ri->mode = cpu_to_jemode(mode); - ret = jffs2_do_new_inode (c, f, mode, ri); - if (ret) { - make_bad_inode(inode); - iput(inode); - return ERR_PTR(ret); - } - inode->i_nlink = 1; - inode->i_ino = je32_to_cpu(ri->ino); - inode->i_mode = jemode_to_cpu(ri->mode); - inode->i_gid = je16_to_cpu(ri->gid); - inode->i_uid = je16_to_cpu(ri->uid); - inode->i_atime = inode->i_ctime = inode->i_mtime = cyg_timestamp(); - ri->atime = ri->mtime = ri->ctime = cpu_to_je32(inode->i_mtime); + memset(ri, 0, sizeof(*ri)); + /* Set OS-specific defaults for new inodes */ + ri->uid = ri->gid = cpu_to_je16(0); + ri->mode = cpu_to_jemode(mode); + ret = jffs2_do_new_inode (c, f, mode, ri); + if (ret) { + make_bad_inode(inode); + iput(inode); + return ERR_PTR(ret); + } + inode->i_nlink = 1; + inode->i_ino = je32_to_cpu(ri->ino); + inode->i_mode = jemode_to_cpu(ri->mode); + inode->i_gid = je16_to_cpu(ri->gid); + inode->i_uid = je16_to_cpu(ri->uid); + inode->i_atime = inode->i_ctime = inode->i_mtime = cyg_timestamp(); + ri->atime = ri->mtime = ri->ctime = cpu_to_je32(inode->i_mtime); - inode->i_size = 0; + inode->i_size = 0; - insert_inode_hash(inode); + insert_inode_hash(inode); - return inode; + return inode; } void jffs2_read_inode (struct inode *inode) { - struct jffs2_inode_info *f; - struct jffs2_sb_info *c; - struct jffs2_raw_inode latest_node; - int ret; + struct jffs2_inode_info *f; + struct jffs2_sb_info *c; + struct jffs2_raw_inode latest_node; + int ret; - D1(printk(KERN_DEBUG "jffs2_read_inode(): inode->i_ino == %lu\n", inode->i_ino)); + D1(printk(KERN_DEBUG "jffs2_read_inode(): inode->i_ino == %lu\n", inode->i_ino)); - f = JFFS2_INODE_INFO(inode); - c = JFFS2_SB_INFO(inode->i_sb); + f = JFFS2_INODE_INFO(inode); + c = JFFS2_SB_INFO(inode->i_sb); - jffs2_init_inode_info(f); - - ret = jffs2_do_read_inode(c, f, inode->i_ino, &latest_node); + jffs2_init_inode_info(f); + + ret = jffs2_do_read_inode(c, f, inode->i_ino, &latest_node); - if (ret) { - make_bad_inode(inode); - up(&f->sem); - return; - } - inode->i_mode = jemode_to_cpu(latest_node.mode); - inode->i_uid = je16_to_cpu(latest_node.uid); - inode->i_gid = je16_to_cpu(latest_node.gid); - inode->i_size = je32_to_cpu(latest_node.isize); - inode->i_atime = je32_to_cpu(latest_node.atime); - inode->i_mtime = je32_to_cpu(latest_node.mtime); - inode->i_ctime = je32_to_cpu(latest_node.ctime); + if (ret) { + make_bad_inode(inode); + up(&f->sem); + return; + } + inode->i_mode = jemode_to_cpu(latest_node.mode); + inode->i_uid = je16_to_cpu(latest_node.uid); + inode->i_gid = je16_to_cpu(latest_node.gid); + inode->i_size = je32_to_cpu(latest_node.isize); + inode->i_atime = je32_to_cpu(latest_node.atime); + inode->i_mtime = je32_to_cpu(latest_node.mtime); + inode->i_ctime = je32_to_cpu(latest_node.ctime); - inode->i_nlink = f->inocache->nlink; - up(&f->sem); + inode->i_nlink = f->inocache->nlink; + up(&f->sem); - D1(printk(KERN_DEBUG "jffs2_read_inode() returning\n")); + D1(printk(KERN_DEBUG "jffs2_read_inode() returning\n")); }
--- a/packages/fs/jffs2/current/src/malloc-ecos.c +++ b/packages/fs/jffs2/current/src/malloc-ecos.c @@ -12,8 +12,13 @@ */ #include <linux/kernel.h> +#include <cyg/hal/drv_api.h> #include "nodelist.h" +#if !defined(CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE) +# define CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE 0 +#endif + struct jffs2_full_dirent *jffs2_alloc_full_dirent(int namesize) { return malloc(sizeof(struct jffs2_full_dirent) + namesize); @@ -64,6 +69,27 @@ void jffs2_free_tmp_dnode_info(struct jf free(x); } +struct jffs2_node_frag *jffs2_alloc_node_frag(void) +{ + return malloc(sizeof(struct jffs2_node_frag)); +} + +void jffs2_free_node_frag(struct jffs2_node_frag *x) +{ + free(x); +} + +#if CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE == 0 + +int jffs2_create_slab_caches(void) +{ + return 0; +} + +void jffs2_destroy_slab_caches(void) +{ +} + struct jffs2_raw_node_ref *jffs2_alloc_raw_node_ref(void) { return malloc(sizeof(struct jffs2_raw_node_ref)); @@ -74,16 +100,54 @@ void jffs2_free_raw_node_ref(struct jffs free(x); } -struct jffs2_node_frag *jffs2_alloc_node_frag(void) +#else // CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE == 0 + +static struct jffs2_raw_node_ref + rnr_pool[CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE] __attribute__ ((aligned (4))), + * first = NULL; +static cyg_drv_mutex_t mutex; + +int jffs2_create_slab_caches(void) { - return malloc(sizeof(struct jffs2_node_frag)); + struct jffs2_raw_node_ref * p; + cyg_drv_mutex_init(&mutex); + for ( + p = rnr_pool; + p < rnr_pool + CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE - 1; + p++ + ) + p->next_phys = p + 1; + rnr_pool[CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE - 1].next_phys = NULL; + first = &rnr_pool[0]; + return 0; } -void jffs2_free_node_frag(struct jffs2_node_frag *x) +void jffs2_destroy_slab_caches(void) +{ +} + +struct jffs2_raw_node_ref *jffs2_alloc_raw_node_ref(void) { - free(x); + struct jffs2_raw_node_ref * p; + + cyg_drv_mutex_lock(&mutex); + p = first; + if (p != NULL) + first = p->next_phys; + cyg_drv_mutex_unlock(&mutex); + return p; } +void jffs2_free_raw_node_ref(struct jffs2_raw_node_ref *x) +{ + cyg_drv_mutex_lock(&mutex); + x->next_phys = first; + first = x; + cyg_drv_mutex_unlock(&mutex); +} + +#endif // CYGNUM_FS_JFFS2_RAW_NODE_REF_CACHE_POOL_SIZE == 0 + struct jffs2_inode_cache *jffs2_alloc_inode_cache(void) { struct jffs2_inode_cache *ret = malloc(sizeof(struct jffs2_inode_cache));