)]}' { "log": [ { "commit": "69cf0fac6052c5bd3fb3469a41d4216e926028f8", "tree": "f403d0731549835fbf406fa4cd0910d9aec1d953", "parents": [ "e14d95f773b7365b6ba2d2105522c92b007c6db1" ], "author": { "name": "Hugh Dickins", "email": "hugh@veritas.com", "time": "Mon Apr 17 22:46:32 2006 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Mon Apr 17 18:22:18 2006 -0700" }, "message": "[PATCH] Fix MADV_REMOVE protection checking\n\nmadvise_remove needs to respect file and mmap protections.\n\nSigned-off-by: Hugh Dickins \u003chugh@veritas.com\u003e\n[ Will the real CVE-2006-1524 stand up, please.. ]\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "f822566165dd46ff5de9bf895cfa6c51f53bb0c4", "tree": "e052f406d5a14140d17f76dc8914d33bbc8e5f1d", "parents": [ "8861da31e3b3e3df7b05e7b157230de3d486e53b" ], "author": { "name": "Michael S. Tsirkin", "email": "mst@mellanox.co.il", "time": "Tue Feb 14 13:53:08 2006 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Tue Feb 14 16:09:34 2006 -0800" }, "message": "[PATCH] madvise MADV_DONTFORK/MADV_DOFORK\n\nCurrently, copy-on-write may change the physical address of a page even if the\nuser requested that the page is pinned in memory (either by mlock or by\nget_user_pages). This happens if the process forks meanwhile, and the parent\nwrites to that page. As a result, the page is orphaned: in case of\nget_user_pages, the application will never see any data hardware DMA\u0027s into\nthis page after the COW. In case of mlock\u0027d memory, the parent is not getting\nthe realtime/security benefits of mlock.\n\nIn particular, this affects the Infiniband modules which do DMA from and into\nuser pages all the time.\n\nThis patch adds madvise options to control whether memory range is inherited\nacross fork. Useful e.g. for when hardware is doing DMA from/into these\npages. Could also be useful to an application wanting to speed up its forks\nby cutting large areas out of consideration.\n\nSigned-off-by: Michael S. Tsirkin \u003cmst@mellanox.co.il\u003e\nAcked-by: Hugh Dickins \u003chugh@veritas.com\u003e\nCc: Michael Kerrisk \u003cmtk-manpages@gmx.net\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "f6b3ec238d12c8cc6cc71490c6e3127988460349", "tree": "b395c1054802760b0e938199231a9de9ac2f358a", "parents": [ "d7339071f6a8b50101d7ba327926b770f22d5d8b" ], "author": { "name": "Badari Pulavarty", "email": "pbadari@us.ibm.com", "time": "Fri Jan 06 00:10:38 2006 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Fri Jan 06 08:33:22 2006 -0800" }, "message": "[PATCH] madvise(MADV_REMOVE): remove pages from tmpfs shm backing store\n\nHere is the patch to implement madvise(MADV_REMOVE) - which frees up a\ngiven range of pages \u0026 its associated backing store. Current\nimplementation supports only shmfs/tmpfs and other filesystems return\n-ENOSYS.\n\n\"Some app allocates large tmpfs files, then when some task quits and some\nclient disconnect, some memory can be released. However the only way to\nrelease tmpfs-swap is to MADV_REMOVE\". - Andrea Arcangeli\n\nDatabases want to use this feature to drop a section of their bufferpool\n(shared memory segments) - without writing back to disk/swap space.\n\nThis feature is also useful for supporting hot-plug memory on UML.\n\nConcerns raised by Andrew Morton:\n\n- \"We have no plan for holepunching! If we _do_ have such a plan (or\n might in the future) then what would the API look like? I think\n sys_holepunch(fd, start, len), so we should start out with that.\"\n\n- Using madvise is very weird, because people will ask \"why do I need to\n mmap my file before I can stick a hole in it?\"\n\n- None of the other madvise operations call into the filesystem in this\n manner. A broad question is: is this capability an MM operation or a\n filesytem operation? truncate, for example, is a filesystem operation\n which sometimes has MM side-effects. madvise is an mm operation and with\n this patch, it gains FS side-effects, only they\u0027re really, really\n significant ones.\"\n\nComments:\n\n- Andrea suggested the fs operation too but then it\u0027s more efficient to\n have it as a mm operation with fs side effects, because they don\u0027t\n immediatly know fd and physical offset of the range. It\u0027s possible to\n fixup in userland and to use the fs operation but it\u0027s more expensive,\n the vmas are already in the kernel and we can use them.\n\nShort term plan \u0026 Future Direction:\n\n- We seem to need this interface only for shmfs/tmpfs files in the short\n term. We have to add hooks into the filesystem for correctness and\n completeness. This is what this patch does.\n\n- In the future, plan is to support both fs and mmap apis also. This\n also involves (other) filesystem specific functions to be implemented.\n\n- Current patch doesn\u0027t support VM_NONLINEAR - which can be addressed in\n the future.\n\nSigned-off-by: Badari Pulavarty \u003cpbadari@us.ibm.com\u003e\nCc: Hugh Dickins \u003chugh@veritas.com\u003e\nCc: Andrea Arcangeli \u003candrea@suse.de\u003e\nCc: Michael Kerrisk \u003cmtk-manpages@gmx.net\u003e\nCc: Ulrich Drepper \u003cdrepper@redhat.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "6aab341e0a28aff100a09831c5300a2994b8b986", "tree": "1af3908275aa5e1b16e80efee554a9a7504c56d4", "parents": [ "458af5439fe7ae7d95ca14106844e61f0795166c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Mon Nov 28 14:34:23 2005 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Mon Nov 28 14:34:23 2005 -0800" }, "message": "mm: re-architect the VM_UNPAGED logic\n\nThis replaces the (in my opinion horrible) VM_UNMAPPED logic with very\nexplicit support for a \"remapped page range\" aka VM_PFNMAP. It allows a\nVM area to contain an arbitrary range of page table entries that the VM\nnever touches, and never considers to be normal pages.\n\nAny user of \"remap_pfn_range()\" automatically gets this new\nfunctionality, and doesn\u0027t even have to mark the pages reserved or\nindeed mark them any other way. It just works. As a side effect, doing\nmmap() on /dev/mem works for arbitrary ranges.\n\nSparc update from David in the next commit.\n\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "0b14c179a483e71ea41df2aa4a661760063115bd", "tree": "075fc303a3d2fd33f66c0af8f64064cff2b72b79", "parents": [ "664beed0190fae687ac51295694004902ddeb18e" ], "author": { "name": "Hugh Dickins", "email": "hugh@veritas.com", "time": "Mon Nov 21 21:32:15 2005 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Tue Nov 22 09:13:42 2005 -0800" }, "message": "[PATCH] unpaged: VM_UNPAGED\n\nAlthough we tend to associate VM_RESERVED with remap_pfn_range, quite a few\ndrivers set VM_RESERVED on areas which are then populated by nopage. The\nPageReserved removal in 2.6.15-rc1 changed VM_RESERVED not to free pages in\nzap_pte_range, without changing those drivers not to set it: so their pages\njust leak away.\n\nLet\u0027s not change miscellaneous drivers now: introduce VM_UNPAGED at the core,\nto flag the special areas where the ptes may have no struct page, or if they\nhave then it\u0027s not to be touched. Replace most instances of VM_RESERVED in\ncore mm by VM_UNPAGED. Force it on in remap_pfn_range, and the sparc and\nsparc64 io_remap_pfn_range.\n\nRevert addition of VM_RESERVED to powerpc vdso, it\u0027s not needed there. Is it\nneeded anywhere? It still governs the mm-\u003ereserved_vm statistic, and special\nvmas not to be merged, and areas not to be core dumped; but could probably be\neliminated later (the drivers are probably specifying it because in 2.4 it\nkept swapout off the vma, but in 2.6 we work from the LRU, which these pages\ndon\u0027t get on).\n\nUse the VM_SHM slot for VM_UNPAGED, and define VM_SHM to 0: it serves no\npurpose whatsoever, and should be removed from drivers when we clean up.\n\nSigned-off-by: Hugh Dickins \u003chugh@veritas.com\u003e\nAcked-by: William Irwin \u003cwli@holomorphy.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "b5810039a54e5babf428e9a1e89fc1940fabff11", "tree": "835836cb527ec9bd525f93eb7e016f3dfb8c8ae2", "parents": [ "f9c98d0287de42221c624482fd4f8d485c98ab22" ], "author": { "name": "Nick Piggin", "email": "nickpiggin@yahoo.com.au", "time": "Sat Oct 29 18:16:12 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Sat Oct 29 21:40:39 2005 -0700" }, "message": "[PATCH] core remove PageReserved\n\nRemove PageReserved() calls from core code by tightening VM_RESERVED\nhandling in mm/ to cover PageReserved functionality.\n\nPageReserved special casing is removed from get_page and put_page.\n\nAll setting and clearing of PageReserved is retained, and it is now flagged\nin the page_alloc checks to help ensure we don\u0027t introduce any refcount\nbased freeing of Reserved pages.\n\nMAP_PRIVATE, PROT_WRITE of VM_RESERVED regions is tentatively being\ndeprecated. We never completely handled it correctly anyway, and is be\nreintroduced in future if required (Hugh has a proof of concept).\n\nOnce PageReserved() calls are removed from kernel/power/swsusp.c, and all\narch/ and driver code, the Set and Clear calls, and the PG_reserved bit can\nbe trivially removed.\n\nLast real user of PageReserved is swsusp, which uses PageReserved to\ndetermine whether a struct page points to valid memory or not. This still\nneeds to be addressed (a generic page_is_ram() should work).\n\nA last caveat: the ZERO_PAGE is now refcounted and managed with rmap (and\nthus mapcounted and count towards shared rss). These writes to the struct\npage could cause excessive cacheline bouncing on big systems. There are a\nnumber of ways this could be addressed if it is an issue.\n\nSigned-off-by: Nick Piggin \u003cnpiggin@suse.de\u003e\n\nRefcount bug fix for filemap_xip.c\n\nSigned-off-by: Carsten Otte \u003ccotte@de.ibm.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "1bef40032992320dd25a266fc166bfb8fa3f2f59", "tree": "e02d654b4db45b752e3201c3f6eec3f4c719b69f", "parents": [ "22c1ea44f0d33eda532883858b6cdabc5f265b66" ], "author": { "name": "Suzuki", "email": "suzuki@in.ibm.com", "time": "Tue Oct 11 08:29:06 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Tue Oct 11 09:46:54 2005 -0700" }, "message": "[PATCH] madvise: Avoid returning error code -EBADF for anonymous mappings\n\nRevert this recent correctness change: Douglas Crosher \u003cdcrosher@scieneer.com\u003e\nreported that it broke an existing application, and that madvise() works\nwithout error on anonymous mappings on Solaris.\n\nThis means that madvise() will remain non-standards-compliant: we should\nreturn -EBADF for all requests against non-file-backed vma\u0027s, but Linux only\ndoes this for MADV_WILLNEED requests.\n\nSigned-off-by: Suzuki K P \u003csuzuki@in.ibm.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "836d5ffd34550901ea024347693e689273ded8aa", "tree": "b4dbbbe436eae38aa2f0f5d333608f64c6338cd8", "parents": [ "53e9a6159fdc6419874ce4d86d3577dbedc77b62" ], "author": { "name": "Hugh Dickins", "email": "hugh@veritas.com", "time": "Sat Sep 03 15:54:53 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@evo.osdl.org", "time": "Mon Sep 05 00:05:44 2005 -0700" }, "message": "[PATCH] mm: fix madvise vma merging\n\nBetter late than never, I\u0027ve at last reviewed the madvise vma merging\ngoing into 2.6.13. Remove a pointless check and fix two little bugs -\na simple test (with /proc/\u003cpid\u003e/maps hacked to show ReadHints) showed\nboth mismerges in practice: though being madvise, neither was disastrous.\n\n1. Correct placement of the success label in madvise_behavior: as in\n mprotect_fixup and mlock_fixup, it is necessary to update vm_flags\n when vma_merge succeeds (to handle the exceptional Case 8 noted in\n the comments above vma_merge itself).\n\n2. Correct initial value of prev when starting part way into a vma: as\n in sys_mprotect and do_mlock, it needs to be set to vma in this case\n (vma_merge handles only that minimum of cases shown in its comments).\n\n3. If find_vma_prev sets prev, then the vma it returns is prev-\u003evm_next,\n so it\u0027s pointless to make that same assignment again in sys_madvise.\n\nSigned-off-by: Hugh Dickins \u003chugh@veritas.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "165cd40235732644b1856a5ed5e158c9b93f6010", "tree": "c68180b6d700953530c8a7852a095f9ac4840345", "parents": [ "4bfdf37830111321e2cd1fe0102dd776ce93194d" ], "author": { "name": "suzuki", "email": "suzuki@in.ibm.com", "time": "Wed Jul 27 11:43:59 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@g5.osdl.org", "time": "Wed Jul 27 16:25:54 2005 -0700" }, "message": "[PATCH] madvise() does not always return -EBADF on non-file mapped area\n\nThe madvise() system call returns -EBADF for areas which does not map to\nfiles, only for *behaviour* request MADV_WILLNEED.\n\nAccording to man pages, madvise returns :\n\nEBADF - the map exists, but the area maps something that isn\u0027t a file.\n\nFixes bug 2995.\n\nSigned-off-by: Suzuki K P \u003csuzuki@in.ibm.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "fe77ba6f4f97690baa4c756611a07f3cc033f6ae", "tree": "01abe2ed3f1ed6f9340f7d9cbad461cbedb47e65", "parents": [ "eb6fe0c388e43b02e261f0fdee60e42f6298d7f7" ], "author": { "name": "Carsten Otte", "email": "cotte@de.ibm.com", "time": "Thu Jun 23 22:05:29 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@ppc970.osdl.org", "time": "Fri Jun 24 00:06:42 2005 -0700" }, "message": "[PATCH] xip: madvice/fadvice: execute in place\n\nMake sys_madvice/fadvice return sane with xip.\n\nSigned-off-by: Carsten Otte \u003ccotte@de.ibm.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "3bc1ee3e8f1c05c0f64a479c6d56eb34a6190599", "tree": "d69ea17b71b309151914ef722d2159e0c780312c", "parents": [ "be5b4fbd017d12e0d09ea0528a5839ce2ed2c8c8" ], "author": { "name": "Pekka Enberg", "email": "penberg@cs.helsinki.fi", "time": "Thu Jun 23 00:09:09 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@ppc970.osdl.org", "time": "Thu Jun 23 09:45:19 2005 -0700" }, "message": "[PATCH] remove redundant vm_flags clearing from madvise.c\n\nThis patch removes redundant VM_ClearReadHint from mm/madvice.c which was\nleft there by Prasanna\u0027s patch.\n\nSigned-off-by: Pekka Enberg \u003cpenberg@cs.helsinki.fi\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "05b7438475ddbac47e75506913d44550f0e75938", "tree": "3871b9c9c0cfa3699de9581a9283dc9fee608f2c", "parents": [ "e798c6e87b64d9fdbd5e9f757b1c033223763d9f" ], "author": { "name": "Prasanna Meda", "email": "pmeda@akamai.com", "time": "Tue Jun 21 17:14:37 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@ppc970.osdl.org", "time": "Tue Jun 21 18:46:13 2005 -0700" }, "message": "[PATCH] madvise: merge the maps\n\nThis attempts to merge back the split maps. This code is mostly copied\nfrom Chrisw\u0027s mlock merging from post 2.6.11 trees. The only difference is\nin munmapped_error handling. Also passed prev to willneed/dontneed,\neventhogh they do not handle it now, since I felt it will be cleaner,\ninstead of handling prev in madvise_vma in some cases and in subfunction in\nsome cases.\n\nSigned-off-by: Prasanna Meda \u003cpmeda@akamai.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "e798c6e87b64d9fdbd5e9f757b1c033223763d9f", "tree": "da4cdd7c60acff82d2fcc826cdbae43440a6df44", "parents": [ "b15e0905f2b9964fc7426fecab57445e96021b61" ], "author": { "name": "Prasanna Meda", "email": "pmeda@akamai.com", "time": "Tue Jun 21 17:14:36 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@ppc970.osdl.org", "time": "Tue Jun 21 18:46:13 2005 -0700" }, "message": "[PATCH] madvise: do not split the maps\n\nThis attempts to avoid splittings when it is not needed, that is when\nvm_flags are same as new flags. The idea is from the \u003c2.6.11 mlock_fixup\nand others. This will provide base for the next madvise merging patch.\n\nSigned-off-by: Prasanna Meda \u003cpmeda@akamai.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n" }, { "commit": "1da177e4c3f41524e886b7f1b8a0c1fc7321cac2", "tree": "0bba044c4ce775e45a88a51686b5d9f90697ea9d", "parents": [], "author": { "name": "Linus Torvalds", "email": "torvalds@ppc970.osdl.org", "time": "Sat Apr 16 15:20:36 2005 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@ppc970.osdl.org", "time": "Sat Apr 16 15:20:36 2005 -0700" }, "message": "Linux-2.6.12-rc2\n\nInitial git repository build. I\u0027m not bothering with the full history,\neven though we have it. We can create a separate \"historical\" git\narchive of that later if we want to, and in the meantime it\u0027s about\n3.2GB when imported into git - space that would just make the early\ngit days unnecessarily complicated, when we don\u0027t have a lot of good\ninfrastructure for it.\n\nLet it rip!\n" } ] }