FreeBSD source tree https://cgit-dev.freebsd.org/src-test2/atom?h=release%2F7.3.0_cvs 2010-03-21T01:30:13Z 2010-03-21T01:30:13Z Ken Smith kensmith@FreeBSD.org 2010-03-21T01:30:13Z urn:sha1:9c4d1c1956ecb0a222fb2f4579c320391fb67f06 Approved by: re (implicit) This commit was manufactured to restore the state of the 7.3-RELEASE image. 2009-06-24T09:54:05Z Konstantin Belousov kib@FreeBSD.org 2009-06-24T09:54:05Z urn:sha1:0a0696c5a38d4ab6b6d6d9b37073dc76bde87168 s/a_fdidx/a_fp/ for VOP_OPEN comments that inline struct vop_open_args definition. MFC r193924: Fix r193923 by noting that type of a_fp is struct file *, not int. 2009-06-23T10:59:59Z Konstantin Belousov kib@FreeBSD.org 2009-06-23T10:59:59Z urn:sha1:72debd9b5f1638aa4f1294edb76e31a991eb9c6b Do not use casts (int *)0 and (struct thread *)0 for the arguments of vn_rdwr, use NULL. 2009-06-19T16:27:41Z Stanislav Sedov stas@FreeBSD.org 2009-06-19T16:27:41Z urn:sha1:05590a12f9df07855d3ceda9d1827e4af0c6bbf7 Add support for inode sizes other than 128 bytes. Cleanup. 2009-05-05T10:34:43Z Konstantin Belousov kib@FreeBSD.org 2009-05-05T10:34:43Z urn:sha1:a899912cb6771bc7eed8b56ab93f164212bcd74d Fix two issues with bufdaemon, often causing the processes to hang in the "nbufkv" sleep. Do not retry request for the new block from ffs_bufwrite() when write is done from bufdaemon and there is a buffer shortage. In getnewbuf(), help bufdaemon to flush dirty buffers owned by the vnode locked by curthread. For MFC, default value for sysctl vfs.flushbufqtarget is set to -1, disabling the helpers. The TDP_BUFNEED flag value from HEAD conflicts with TDP_UPCALLING KSE bit, so it is moved to the end of allocated bits. 2009-03-01T11:02:37Z Edward Tomasz Napierala trasz@FreeBSD.org 2009-03-01T11:02:37Z urn:sha1:7e0c02494c43e12adc95f8cb028e7e5c057c3b52 According to phk@, VOP_STRATEGY should never, _ever_, return anything other than 0. Make it so. This fixes "panic: VOP_STRATEGY failed bp=0xc320dd90 vp=0xc3b9f648", encountered when writing to an orphaned filesystem. Reason for the panic was the following assert: KASSERT(i == 0, ("VOP_STRATEGY failed bp=%p vp=%p", bp, bp->b_vp)); at vfs_bio:bufstrategy(). Reviewed by: scottl, phk Approved by: rwatson (mentor) Sponsored by: FreeBSD Foundation 2009-01-29T11:38:28Z Konstantin Belousov kib@FreeBSD.org 2009-01-29T11:38:28Z urn:sha1:e8b10b2a053cc3ab9215d9ce7687d7daf1152667 Garbage-collect vn_write_suspend_wait(). 2009-01-21T15:51:15Z Konstantin Belousov kib@FreeBSD.org 2009-01-21T15:51:15Z urn:sha1:1a037be619deeea5bdabca65dc3324d61cbccaef Initialize birthtime fields in vn_stat() to prevent stat(2) from returning uninitialized birthtime. Most file systems don't initialize birthtime properly in their VOP_GETTATTR(). MFC r183212: Initialize va_flags and va_filerev properly in VOP_GETATTR(). Don't initialize va_vaflags and va_spare because they are not part of the VOP_GETATTR() API. Also don't initialize birthtime to ctime or zero. MFC rr183213: Initialize va_rdev to NODEV and va_fsid to VNOVAL before the VOP_GETATTR() call in vn_stat(). Thus if a file system doesn't initialize those fields in VOP_GETATTR() they will have a sane default value. MFC r183214: Initialize va_rdev to NODEV instead of 0 or VNOVAL in VOP_GETATTR(). NODEV is more appropriate when va_rdev doesn't have a meaningful value. MFC r183215: fdescfs, devfs, mqueuefs, nfs, portalfs, pseudofs, tmpfs and xfs initialize the vattr structure in VOP_GETATTR() with VATTR_NULL(), vattr_null() or by zeroing it. Remove these to allow preinitialization of fields work in vn_stat(). This is needed to get birthtime initialized correctly. 2009-01-18T11:47:39Z Konstantin Belousov kib@FreeBSD.org 2009-01-18T11:47:39Z urn:sha1:a46abf212dd690fbae1c538cc0c8aee529843276 Do not incorrectly add the low 5 bits of the offset to the resulting position of the found zero bit. 2008-08-01T09:46:19Z Konstantin Belousov kib@FreeBSD.org 2008-08-01T09:46:19Z urn:sha1:cb917b839b94adf81f241d51bcd2a1ce82161618 Move the head of byte-level advisory lock list from the filesystem-specific vnode data to the struct vnode. Provide the default implementation for the vop_advlock and vop_advlockasync. Purge the locks on the vnode reclaim by using the lf_purgelocks(). The default implementation is augmented for the nfs and smbfs. In the nfs_advlock, push the Giant inside the nfs_dolock. Before the change, the vop_advlock and vop_advlockasync have taken the unlocked vnode and dereferenced the fs-private inode data, racing with with the vnode reclamation due to forced unmount. Now, the vop_getattr under the shared vnode lock is used to obtain the inode size, and later, in the lf_advlockasync, after locking the vnode interlock, the VI_DOOMED flag is checked to prevent an operation on the doomed vnode. MFC r178247: Fix compilation with LOCKF_DEBUG. As was announced on the stable@, the MFC changes the VFS KBI on RELENG_7. Approved by: re (kensmith)