aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/tools/perf/scripts/python/export-to-postgresql.py (unfollow)
AgeCommit message (Collapse)AuthorFilesLines
2024-07-12ubifs: dbg_orphan_check: Fix missed key type checkingZhihao Cheng1-2/+6
When selinux/encryption is enabled, xattr entry node is added into TNC before host inode when creating new file. So it is possible to find xattr entry without host inode from TNC. Orphan debug checking is called by ubifs_orphan_end_commit(), at that time, the commit semaphore is already unlock, so the new creation won't be blocked. Fixes: d7f0b70d30ff ("UBIFS: Add security.* XATTR support for the UBIFS") Fixes: d475a507457b ("ubifs: Add skeleton for fscrypto") Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubifs: Fix unattached inode when powercut happens in creatingZhihao Cheng3-26/+51
For selinux or encryption scenarios, UBIFS could become inconsistent while creating new files in powercut case. Encryption/selinux related xattrs will be created before creating file dentry, which makes creation process is not atomic, details are shown as: Encryption case: ubifs_create ubifs_new_inode fscrypt_set_context ubifs_xattr_set create_xattr ubifs_jnl_update // Disk: xentry xinode inode(LAST_OF_NODE_GROUP) >> power cut << ubifs_jnl_update // Disk: dentry inode parent_inode(LAST_OF_NODE_GROUP) Selinux case: ubifs_create ubifs_new_inode ubifs_init_security security_inode_init_security ubifs_xattr_set create_xattr ubifs_jnl_update // Disk: xentry xinode inode(LAST_OF_NODE_GROUP) >> power cut << ubifs_jnl_update // Disk: dentry inode parent_inode(LAST_OF_NODE_GROUP) Above process will make chk_fs failed in next mounting: UBIFS error (ubi0:0 pid 7995): dbg_check_filesystem [ubifs]: inode 66 nlink is 1, but calculated nlink is 0 Fix it by allocating orphan inode for each non-xattr file creation, then removing orphan list in journal writing process, which ensures that both xattr and dentry be effective in atomic when powercut happens. Fixes: d7f0b70d30ff ("UBIFS: Add security.* XATTR support for the UBIFS") Fixes: d475a507457b ("ubifs: Add skeleton for fscrypto") Link: https://bugzilla.kernel.org/show_bug.cgi?id=218309 Suggested-by: Zhang Yi <yi.zhang@huawei.com> Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubifs: Fix space leak when powercut happens in linking tmpfileZhihao Cheng4-17/+15
There is a potential space leak problem when powercut happens in linking tmpfile, in which case, inode node (with nlink=0) and its' data nodes can be found from tnc (on flash), but there are no dentries related to the inode, so the file is invisible but takes free space. Detailed process is shown as: ubifs_tmpfile ubifs_jnl_update // Add bud A into log area ubifs_add_orphan // Add inode into orphan list P1 P2 ubifs_link ubifs_delete_orphan // Delete inode from orphan list, then inode won't // be written into orphan area, there is no chance // to delete inode by replaying orphan. commit // bud A won't be replayed in next mounting >> powercut << ubifs_jnl_update // Link inode to dentry The root cause is that orphan entry deletion and journal writing(for link) are interrupted by commit, which makes the two operations are not atomic. Fix it by doing ubifs_delete_orphan under the protection of c->commit_sem within ubifs_jnl_update. This is also a preparation to support all creating new files by orphan inode. v1 is https://lore.kernel.org/linux-mtd/20200701093227.674945-1-chengzhihao1@huawei.com/ Fixes: 32fe905c17f0 ("ubifs: Fix O_TMPFILE corner case in ubifs_link()") Link: https://bugzilla.kernel.org/show_bug.cgi?id=208405 Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubifs: Move ui->data initialization after initializing securityZhihao Cheng1-8/+10
Host inode and its' xattr will be written on disk after initializing security when creating symlink or dev, then the host inode and its dentry will be written again in ubifs_jnl_update. There is no need to write inode data in the security initialization pass, just move the ui->data initialization after initializing security. Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubifs: Fix adding orphan entry twice for the same inodeZhihao Cheng1-1/+1
The tmpfile could be added into orphan list twice, first time is creation, the second time is removing after it is linked. The orphan entry could be added twice for tmpfile if following sequence is satisfied: ubifs_tmpfile ubifs_jnl_update ubifs_add_orphan // first time to add orphan entry P1 P2 ubifs_link do_commit ubifs_orphan_start_commit orphan->cmt = 1 ubifs_delete_orphan orphan_delete if (orph->cmt) orph->del = 1; // orphan entry is not deleted from tree return ubifs_unlink ubifs_jnl_update ubifs_add_orphan orphan_add // found old orphan entry, second time to add orphan entry ubifs_err(c, "orphaned twice") return -EINVAL // unlink failed! ubifs_orphan_end_commit erase_deleted // delete old orphan entry rb_erase(&orphan->rb, &c->orph_tree) Fix it by removing orphan entry from orphan tree in advance, rather than remove it from orphan tree in committing process. Fixes: 32fe905c17f0 ("ubifs: Fix O_TMPFILE corner case in ubifs_link()") Link: https://bugzilla.kernel.org/show_bug.cgi?id=218672 Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubifs: Remove insert_dead_orphan from replaying orphan processZhihao Cheng1-49/+0
UBIFS will do commit at the end of mounting process(rw mode), dead orphans(added by insert_dead_orphan in replaying orphan) are deleted by ubifs_orphan_end_commit(). The only reason why dead orphans are added into orphan list is that old orpans may be lost when powercut happens in ubifs_orphan_end_commit(): ubifs_orphan_end_commit // TNC(updated by orphans) is not written yet if (c->cmt_orphans != 0) commit_orphans consolidate // traverse orphan list write_orph_nodes // rewrite all orphans by ubifs_leb_change // If dead orphans are not in list, they will be lost when powercut // happens, then TNC won't be updated by old orphans in next mounting. Luckily, the condition 'c->cmt_orphans != 0' will never be true in mounting process, there can't be new orphans added into orphan list before mounting returned, but commit will be done at the end of mounting. Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12Revert "ubifs: ubifs_symlink: Fix memleak of inode->i_link in error path"Zhihao Cheng1-2/+0
This reverts commit 6379b44cdcd67f5f5d986b73953e99700591edfa. Commit 1e022216dcd2 ("ubifs: ubifs_symlink: Fix memleak of inode->i_link in error path") is applied again in commit 6379b44cdcd6 ("ubifs: ubifs_symlink: Fix memleak of inode->i_link in error path"), which changed ubifs_mknod (It won't become a real problem). Just revert it. Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubifs: Don't add xattr inode into orphan areaZhihao Cheng2-74/+14
Now, the entire inode with its' xattrs are removed while replaying orphan nodes. There is no need to add xattr inodes into orphan area, which is based on the fact that xattr entries won't be cleared from disk before deleting xattr inodes, in another words, current logic can make sure that xattr inode be deleted in any cases even UBIFS not record xattr inode into orphan area. Let's looking for possible paths that could clear xattr entries from disk but leave the xattr inode on TNC: 1. unlink/tmpfile -> ubifs_jnl_update: inode(nlink=0) is written into bud LEB and added into orphan list, then: a. powercut: ubifs_tnc_remove_ino(xattr entry/inode can be found from TNC and being deleted) is invoked in replaying journal. b. commit + powercut: inode is written into orphan area, and ubifs_tnc_remove_ino is invoked in replaying orphan nodes. c. evicting + powercut: xattr inode(nlink=0) is written on disk, xattr is removed from TNC, gc could clear xattr entries from disk. ubifs_tnc_remove_ino will apply on inode and xattr inode in replaying journal, so lost xattr entries will make no influence. d. evicting + commit + powercut: xattr inode/entry are removed from index tree(on disk) by ubifs_jnl_write_inode, xattr inode is cleared from orphan area by ubifs_jnl_write_inode + commit. e. commit + evicting + powercut: inode is written into orphan area, then equivalent to c. 2. remove xattr -> ubifs_jnl_delete_xattr: xattr entry(inum=0) and xattr inode(nlink=0) is written into bud LEB, xattr entry/inode are removed from TNC, then: a. powercut: gc could clear xattr entries from disk, which won't affect deleting xattr entry from TNC. ubifs_tnc_remove_ino will apply on xattr inode in replaying journal, ubifs_tnc_remove_nm will apply on xattr entry in replaying journal. b. commit + powercut: xattr entry/inode are removed from index tree (on disk). Tracking xattr inode in orphan list is imported by commit 988bec41318f3f ("ubifs: orphan: Handle xattrs like files"), it aims to fix the similar problem described in commit 7959cf3a7506d4a ("ubifs: journal: Handle xattrs like files"). Actually, the problem only exist in journal case but not the orphan case. So, we can remove the orphan tracking for xattr inodes. Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubifs: Fix unattached xattr inode if powercut happens after deletingZhihao Cheng1-7/+4
When powercut happens after deleting file, the xattr inode could be alone existing in TNC but its' xattr entry cannot be found in TNC. File inode and xattr inode are added into orphan list after deleting file, file inode's nlink is 0 but xattr inode's nlink is not 0 (PS: zero nlink xattr inode is written on disk in evicting process by ubifs_jnl_write_inode). So, following process could happen: 1. touch file 2. setxattr(file) 3. unlink file // inode(nlink=0), xattr inode(nlink=1) are added into orphan list 4. commit // write inode inum and xattr inum into orphan area 5. powercut 6. mount do_kill_orphans // inode(nlink=0) is deleted from TNC by ubifs_tnc_remove_range, // xattr entry is deleted too. // xattr inode(nlink=1) is not deleted from TNC Finally we could see following error while debugging UBIFS: UBIFS error (ubi0:0 pid 1093): dbg_check_filesystem [ubifs]: inode 66 nlink is 1, but calculated nlink is 0 UBIFS (ubi0:0): dump of the inode 66 sitting in LEB 12:2128 node_type 0 (inode node) group_type 1 (in node group) len 197 key (66, inode) size 37 nlink 1 flags 0x20 xattr_cnt 0 xattr_size 0 xattr_names 0 data len 37 Fix it by removing entire inode with it's xattrs while replaying orphan, just replace function ubifs_tnc_remove_range by ubifs_tnc_remove_ino. Fixes: ee1438ce5dc4 ("ubifs: Check link count of inodes when killing orphans.") Link: https://bugzilla.kernel.org/show_bug.cgi?id=218661 Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12mtd: ubi: avoid expensive do_div() on 32-bit machinesArnd Bergmann1-3/+3
The use of do_div() in ubi_nvmem_reg_read() makes calling it on 32-bit machines rather expensive. Since the 'from' variable is known to be a 32-bit quantity, it is clearly never needed and can be optimized into a regular division operation. Fixes: b8a77b9a5f9c ("mtd: ubi: fix NVMEM over UBI volumes on 32-bit systems") Fixes: 3ce485803da1 ("mtd: ubi: provide NVMEM layer over UBI volumes") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12mtd: ubi: make ubi_class constantRicardo B. Marliere2-2/+2
Since commit 43a7206b0963 ("driver core: class: make class_register() take a const *"), the driver core allows for struct class to be in read-only memory, so move the ubi_class structure to be declared at build time placing it into read-only memory, instead of having to be dynamically allocated at boot time. Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net> Reviewed-by: Zhihao Cheng <chengzhihao1@huawei.com> Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-07-12ubi: eba: properly rollback inside self_check_ebaFedor Pchelkin1-1/+2
In case of a memory allocation failure in the volumes loop we can only process the already allocated scan_eba and fm_eba array elements on the error path - others are still uninitialized. Found by Linux Verification Center (linuxtesting.org). Fixes: 00abf3041590 ("UBI: Add self_check_eba()") Cc: stable@vger.kernel.org Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru> Reviewed-by: Zhihao Cheng <chengzhihao1@huawei.com> Signed-off-by: Richard Weinberger <richard@nod.at>
2024-06-30Linux 6.10-rc6Linus Torvalds1-1/+1
2024-06-30ata: ahci: Clean up sysfs file on errorNiklas Cassel1-5/+12
.probe() (ahci_init_one()) calls sysfs_add_file_to_group(), however, if probe() fails after this call, we currently never call sysfs_remove_file_from_group(). (The sysfs_remove_file_from_group() call in .remove() (ahci_remove_one()) does not help, as .remove() is not called on .probe() error.) Thus, if probe() fails after the sysfs_add_file_to_group() call, the next time we insmod the module we will get: sysfs: cannot create duplicate filename '/devices/pci0000:00/0000:00:04.0/remapped_nvme' CPU: 11 PID: 954 Comm: modprobe Not tainted 6.10.0-rc5 #43 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014 Call Trace: <TASK> dump_stack_lvl+0x5d/0x80 sysfs_warn_dup.cold+0x17/0x23 sysfs_add_file_mode_ns+0x11a/0x130 sysfs_add_file_to_group+0x7e/0xc0 ahci_init_one+0x31f/0xd40 [ahci] Fixes: 894fba7f434a ("ata: ahci: Add sysfs attribute to show remapped NVMe device count") Cc: stable@vger.kernel.org Reviewed-by: Damien Le Moal <dlemoal@kernel.org> Reviewed-by: Hannes Reinecke <hare@suse.de> Link: https://lore.kernel.org/r/20240629124210.181537-10-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel@kernel.org>
2024-06-30ata: libata-core: Fix double free on errorNiklas Cassel1-4/+4
If e.g. the ata_port_alloc() call in ata_host_alloc() fails, we will jump to the err_out label, which will call devres_release_group(). devres_release_group() will trigger a call to ata_host_release(). ata_host_release() calls kfree(host), so executing the kfree(host) in ata_host_alloc() will lead to a double free: kernel BUG at mm/slub.c:553! Oops: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI CPU: 11 PID: 599 Comm: (udev-worker) Not tainted 6.10.0-rc5 #47 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014 RIP: 0010:kfree+0x2cf/0x2f0 Code: 5d 41 5e 41 5f 5d e9 80 d6 ff ff 4d 89 f1 41 b8 01 00 00 00 48 89 d9 48 89 da RSP: 0018:ffffc90000f377f0 EFLAGS: 00010246 RAX: ffff888112b1f2c0 RBX: ffff888112b1f2c0 RCX: ffff888112b1f320 RDX: 000000000000400b RSI: ffffffffc02c9de5 RDI: ffff888112b1f2c0 RBP: ffffc90000f37830 R08: 0000000000000000 R09: 0000000000000000 R10: ffffc90000f37610 R11: 617461203a736b6e R12: ffffea00044ac780 R13: ffff888100046400 R14: ffffffffc02c9de5 R15: 0000000000000006 FS: 00007f2f1cabe980(0000) GS:ffff88813b380000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f2f1c3acf75 CR3: 0000000111724000 CR4: 0000000000750ef0 PKRU: 55555554 Call Trace: <TASK> ? __die_body.cold+0x19/0x27 ? die+0x2e/0x50 ? do_trap+0xca/0x110 ? do_error_trap+0x6a/0x90 ? kfree+0x2cf/0x2f0 ? exc_invalid_op+0x50/0x70 ? kfree+0x2cf/0x2f0 ? asm_exc_invalid_op+0x1a/0x20 ? ata_host_alloc+0xf5/0x120 [libata] ? ata_host_alloc+0xf5/0x120 [libata] ? kfree+0x2cf/0x2f0 ata_host_alloc+0xf5/0x120 [libata] ata_host_alloc_pinfo+0x14/0xa0 [libata] ahci_init_one+0x6c9/0xd20 [ahci] Ensure that we will not call kfree(host) twice, by performing the kfree() only if the devres_open_group() call failed. Fixes: dafd6c496381 ("libata: ensure host is free'd on error exit paths") Cc: stable@vger.kernel.org Reviewed-by: Damien Le Moal <dlemoal@kernel.org> Reviewed-by: Hannes Reinecke <hare@suse.de> Link: https://lore.kernel.org/r/20240629124210.181537-9-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel@kernel.org>
2024-06-30ata,scsi: libata-core: Do not leak memory for ata_port struct membersNiklas Cassel4-14/+19
libsas is currently not freeing all the struct ata_port struct members, e.g. ncq_sense_buf for a driver supporting Command Duration Limits (CDL). Add a function, ata_port_free(), that is used to free a ata_port, including its struct members. It makes sense to keep the code related to freeing a ata_port in its own function, which will also free all the struct members of struct ata_port. Fixes: 18bd7718b5c4 ("scsi: ata: libata: Handle completion of CDL commands using policy 0xD") Reviewed-by: John Garry <john.g.garry@oracle.com> Link: https://lore.kernel.org/r/20240629124210.181537-8-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel@kernel.org>
2024-06-30ata: libata-core: Fix null pointer dereference on errorNiklas Cassel1-0/+3
If the ata_port_alloc() call in ata_host_alloc() fails, ata_host_release() will get called. However, the code in ata_host_release() tries to free ata_port struct members unconditionally, which can lead to the following: BUG: unable to handle page fault for address: 0000000000003990 PGD 0 P4D 0 Oops: Oops: 0000 [#1] PREEMPT SMP NOPTI CPU: 10 PID: 594 Comm: (udev-worker) Not tainted 6.10.0-rc5 #44 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014 RIP: 0010:ata_host_release.cold+0x2f/0x6e [libata] Code: e4 4d 63 f4 44 89 e2 48 c7 c6 90 ad 32 c0 48 c7 c7 d0 70 33 c0 49 83 c6 0e 41 RSP: 0018:ffffc90000ebb968 EFLAGS: 00010246 RAX: 0000000000000041 RBX: ffff88810fb52e78 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffff88813b3218c0 RDI: ffff88813b3218c0 RBP: ffff88810fb52e40 R08: 0000000000000000 R09: 6c65725f74736f68 R10: ffffc90000ebb738 R11: 73692033203a746e R12: 0000000000000004 R13: 0000000000000000 R14: 0000000000000011 R15: 0000000000000006 FS: 00007f6cc55b9980(0000) GS:ffff88813b300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000003990 CR3: 00000001122a2000 CR4: 0000000000750ef0 PKRU: 55555554 Call Trace: <TASK> ? __die_body.cold+0x19/0x27 ? page_fault_oops+0x15a/0x2f0 ? exc_page_fault+0x7e/0x180 ? asm_exc_page_fault+0x26/0x30 ? ata_host_release.cold+0x2f/0x6e [libata] ? ata_host_release.cold+0x2f/0x6e [libata] release_nodes+0x35/0xb0 devres_release_group+0x113/0x140 ata_host_alloc+0xed/0x120 [libata] ata_host_alloc_pinfo+0x14/0xa0 [libata] ahci_init_one+0x6c9/0xd20 [ahci] Do not access ata_port struct members unconditionally. Fixes: 633273a3ed1c ("libata-pmp: hook PMP support and enable it") Cc: stable@vger.kernel.org Reviewed-by: Damien Le Moal <dlemoal@kernel.org> Reviewed-by: Hannes Reinecke <hare@suse.de> Reviewed-by: John Garry <john.g.garry@oracle.com> Link: https://lore.kernel.org/r/20240629124210.181537-7-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel@kernel.org>
2024-06-30x86-32: fix cmpxchg8b_emu build error with clangLinus Torvalds1-7/+5
The kernel test robot reported that clang no longer compiles the 32-bit x86 kernel in some configurations due to commit 95ece48165c1 ("locking/atomic/x86: Rewrite x86_32 arch_atomic64_{,fetch}_{and,or,xor}() functions"). The build fails with arch/x86/include/asm/cmpxchg_32.h:149:9: error: inline assembly requires more registers than available and the reason seems to be that not only does the cmpxchg8b instruction need four fixed registers (EDX:EAX and ECX:EBX), with the emulation fallback the inline asm also wants a fifth fixed register for the address (it uses %esi for that, but that's just a software convention with cmpxchg8b_emu). Avoiding using another pointer input to the asm (and just forcing it to use the "0(%esi)" addressing that we end up requiring for the sw fallback) seems to fix the issue. Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202406230912.F6XFIyA6-lkp@intel.com/ Fixes: 95ece48165c1 ("locking/atomic/x86: Rewrite x86_32 arch_atomic64_{,fetch}_{and,or,xor}() functions") Link: https://lore.kernel.org/all/202406230912.F6XFIyA6-lkp@intel.com/ Suggested-by: Uros Bizjak <ubizjak@gmail.com> Reviewed-and-Tested-by: Uros Bizjak <ubizjak@gmail.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2024-06-28x86: stop playing stack games in profile_pc()Linus Torvalds1-19/+1
The 'profile_pc()' function is used for timer-based profiling, which isn't really all that relevant any more to begin with, but it also ends up making assumptions based on the stack layout that aren't necessarily valid. Basically, the code tries to account the time spent in spinlocks to the caller rather than the spinlock, and while I support that as a concept, it's not worth the code complexity or the KASAN warnings when no serious profiling is done using timers anyway these days. And the code really does depend on stack layout that is only true in the simplest of cases. We've lost the comment at some point (I think when the 32-bit and 64-bit code was unified), but it used to say: Assume the lock function has either no stack frame or a copy of eflags from PUSHF. which explains why it just blindly loads a word or two straight off the stack pointer and then takes a minimal look at the values to just check if they might be eflags or the return pc: Eflags always has bits 22 and up cleared unlike kernel addresses but that basic stack layout assumption assumes that there isn't any lock debugging etc going on that would complicate the code and cause a stack frame. It causes KASAN unhappiness reported for years by syzkaller [1] and others [2]. With no real practical reason for this any more, just remove the code. Just for historical interest, here's some background commits relating to this code from 2006: 0cb91a229364 ("i386: Account spinlocks to the caller during profiling for !FP kernels") 31679f38d886 ("Simplify profile_pc on x86-64") and a code unification from 2009: ef4512882dbe ("x86: time_32/64.c unify profile_pc") but the basics of this thing actually goes back to before the git tree. Link: https://syzkaller.appspot.com/bug?extid=84fe685c02cd112a2ac3 [1] Link: https://lore.kernel.org/all/CAK55_s7Xyq=nh97=K=G1sxueOFrJDAvPOJAL4TPTCAYvmxO9_A@mail.gmail.com/ [2] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2024-06-28i2c: testunit: discard write requests while old command is runningWolfram Sang1-0/+3
When clearing registers on new write requests was added, the protection for currently running commands was missed leading to concurrent access to the testunit registers. Check the flag beforehand. Fixes: b39ab96aa894 ("i2c: testunit: add support for block process calls") Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Reviewed-by: Andi Shyti <andi.shyti@kernel.org>
2024-06-28i2c: testunit: don't erase registers after STOPWolfram Sang1-1/+1
STOP fallsthrough to WRITE_REQUESTED but this became problematic when clearing the testunit registers was added to the latter. Actually, there is no reason to clear the testunit state after STOP. Doing it when a new WRITE_REQUESTED arrives is enough. So, no need to fallthrough, at all. Fixes: b39ab96aa894 ("i2c: testunit: add support for block process calls") Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Reviewed-by: Andi Shyti <andi.shyti@kernel.org>
2024-06-28tty: mxser: Remove __counted_by from mxser_board.ports[]Nathan Chancellor1-1/+1
Work for __counted_by on generic pointers in structures (not just flexible array members) has started landing in Clang 19 (current tip of tree). During the development of this feature, a restriction was added to __counted_by to prevent the flexible array member's element type from including a flexible array member itself such as: struct foo { int count; char buf[]; }; struct bar { int count; struct foo data[] __counted_by(count); }; because the size of data cannot be calculated with the standard array size formula: sizeof(struct foo) * count This restriction was downgraded to a warning but due to CONFIG_WERROR, it can still break the build. The application of __counted_by on the ports member of 'struct mxser_board' triggers this restriction, resulting in: drivers/tty/mxser.c:291:2: error: 'counted_by' should not be applied to an array with element of unknown size because 'struct mxser_port' is a struct type with a flexible array member. This will be an error in a future compiler version [-Werror,-Wbounds-safety-counted-by-elt-type-unknown-size] 291 | struct mxser_port ports[] __counted_by(nports); | ^~~~~~~~~~~~~~~~~~~~~~~~~ 1 error generated. Remove this use of __counted_by to fix the warning/error. However, rather than remove it altogether, leave it commented, as it may be possible to support this in future compiler releases. Cc: <stable@vger.kernel.org> Closes: https://github.com/ClangBuiltLinux/linux/issues/2026 Fixes: f34907ecca71 ("mxser: Annotate struct mxser_board with __counted_by") Signed-off-by: Nathan Chancellor <nathan@kernel.org> Link: https://lore.kernel.org/r/20240529-drop-counted-by-ports-mxser-board-v1-1-0ab217f4da6d@kernel.org Signed-off-by: Kees Cook <kees@kernel.org>
2024-06-28randomize_kstack: Remove non-functional per-arch entropy filteringKees Cook3-19/+14
An unintended consequence of commit 9c573cd31343 ("randomize_kstack: Improve entropy diffusion") was that the per-architecture entropy size filtering reduced how many bits were being added to the mix, rather than how many bits were being used during the offsetting. All architectures fell back to the existing default of 0x3FF (10 bits), which will consume at most 1KiB of stack space. It seems that this is working just fine, so let's avoid the confusion and update everything to use the default. The prior intent of the per-architecture limits were: arm64: capped at 0x1FF (9 bits), 5 bits effective powerpc: uncapped (10 bits), 6 or 7 bits effective riscv: uncapped (10 bits), 6 bits effective x86: capped at 0xFF (8 bits), 5 (x86_64) or 6 (ia32) bits effective s390: capped at 0xFF (8 bits), undocumented effective entropy Current discussion has led to just dropping the original per-architecture filters. The additional entropy appears to be safe for arm64, x86, and s390. Quoting Arnd, "There is no point pretending that 15.75KB is somehow safe to use while 15.00KB is not." Co-developed-by: Yuntao Liu <liuyuntao12@huawei.com> Signed-off-by: Yuntao Liu <liuyuntao12@huawei.com> Fixes: 9c573cd31343 ("randomize_kstack: Improve entropy diffusion") Link: https://lore.kernel.org/r/20240617133721.377540-1-liuyuntao12@huawei.com Reviewed-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Mark Rutland <mark.rutland@arm.com> Acked-by: Heiko Carstens <hca@linux.ibm.com> # s390 Link: https://lore.kernel.org/r/20240619214711.work.953-kees@kernel.org Signed-off-by: Kees Cook <kees@kernel.org>
2024-06-28string: kunit: add missing MODULE_DESCRIPTION() macrosJeff Johnson2-0/+2
make allmodconfig && make W=1 C=1 reports: WARNING: modpost: missing MODULE_DESCRIPTION() in lib/string_kunit.o WARNING: modpost: missing MODULE_DESCRIPTION() in lib/string_helpers_kunit.o Add the missing invocation of the MODULE_DESCRIPTION() macro. Signed-off-by: Jeff Johnson <quic_jjohnson@quicinc.com> Link: https://lore.kernel.org/r/20240531-md-lib-string-v1-1-2738cf057d94@quicinc.com Signed-off-by: Kees Cook <kees@kernel.org>
2024-06-28ata: libata-core: Add ATA_HORKAGE_NOLPM for all Crucial BX SSD1 modelsNiklas Cassel1-2/+1
We got another report that CT1000BX500SSD1 does not work with LPM. If you look in libata-core.c, we have six different Crucial devices that are marked with ATA_HORKAGE_NOLPM. This model would have been the seventh. (This quirk is used on Crucial models starting with both CT* and Crucial_CT*) It is obvious that this vendor does not have a great history of supporting LPM properly, therefore, add the ATA_HORKAGE_NOLPM quirk for all Crucial BX SSD1 models. Fixes: 7627a0edef54 ("ata: ahci: Drop low power policy board type") Cc: stable@vger.kernel.org Reported-by: Alessandro Maggio <alex.tkd.alex@gmail.com> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218832 Reviewed-by: Damien Le Moal <dlemoal@kernel.org> Link: https://lore.kernel.org/r/20240627105551.4159447-2-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel@kernel.org>
2024-06-28MAINTAINERS: Update IOMMU tree locationJoerg Roedel1-4/+4
Update the maintainers entries to the new location of the IOMMU tree. Signed-off-by: Joerg Roedel <jroedel@suse.de>
2024-06-27tools/power turbostat: Add local build_bug.h header for snapshot targetPatryk Wlazlyn2-2/+6
Fixes compilation errors for Makefile snapshot target described in: commit 231ce08b662a ("tools/power turbostat: Add "snapshot:" Makefile target") Signed-off-by: Patryk Wlazlyn <patryk.wlazlyn@linux.intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2024-06-27tools/power turbostat: Fix unc freq columns not showing with '-q' or '-l'Adam Hawley1-8/+8
Commit 78464d7681f7 ("tools/power turbostat: Add columns for clustered uncore frequency") introduced 'probe_intel_uncore_frequency_cluster()' in a way which prevents printing uncore frequency columns if either of the '-q' or '-l' options are used. Systems which do not have multiple uncore frequencies per package are unaffected by this regression. Fix the function so that uncore frequency columns are shown when either the '-l' or '-q' option is used by checking if 'quiet' is true after adding counters for the uncore frequency columns. Fixes: 78464d7681f7 ("tools/power turbostat: Add columns for clustered uncore frequency") Signed-off-by: Adam Hawley <adam.james.hawley@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2024-06-27tools/power turbostat: option '-n' is ambiguousDavid Arcari1-1/+1
In some cases specifying the '-n' command line argument will cause turbostat to fail. For instance 'turbostat -n 1' works fine; however, 'turbostat -n 1 -d' will fail. This is the result of the first call to getopt_long_only() where "MP" is specified as the optstring. This can be easily fixed by changing the optstring from "MP" to "MPn:" to remove ambiguity between the arguments. tools/power turbostat: option '-n' is ambiguous; possibilities: '-num_iterations' '-no-msr' '-no-perf' Fixes: a0e86c90b83c ("tools/power turbostat: Add --no-perf option") Signed-off-by: David Arcari <darcari@redhat.com> Signed-off-by: Len Brown <len.brown@intel.com>
2024-06-28drm/drm_file: Fix pid refcounting raceJann Horn1-5/+3
<maarten.lankhorst@linux.intel.com>, Maxime Ripard <mripard@kernel.org>, Thomas Zimmermann <tzimmermann@suse.de> filp->pid is supposed to be a refcounted pointer; however, before this patch, drm_file_update_pid() only increments the refcount of a struct pid after storing a pointer to it in filp->pid and dropping the dev->filelist_mutex, making the following race possible: process A process B ========= ========= begin drm_file_update_pid mutex_lock(&dev->filelist_mutex) rcu_replace_pointer(filp->pid, <pid B>, 1) mutex_unlock(&dev->filelist_mutex) begin drm_file_update_pid mutex_lock(&dev->filelist_mutex) rcu_replace_pointer(filp->pid, <pid A>, 1) mutex_unlock(&dev->filelist_mutex) get_pid(<pid A>) synchronize_rcu() put_pid(<pid B>) *** pid B reaches refcount 0 and is freed here *** get_pid(<pid B>) *** UAF *** synchronize_rcu() put_pid(<pid A>) As far as I know, this race can only occur with CONFIG_PREEMPT_RCU=y because it requires RCU to detect a quiescent state in code that is not explicitly calling into the scheduler. This race leads to use-after-free of a "struct pid". It is probably somewhat hard to hit because process A has to pass through a synchronize_rcu() operation while process B is between mutex_unlock() and get_pid(). Fix it by ensuring that by the time a pointer to the current task's pid is stored in the file, an extra reference to the pid has been taken. This fix also removes the condition for synchronize_rcu(); I think that optimization is unnecessary complexity, since in that case we would usually have bailed out on the lockless check above. Fixes: 1c7a387ffef8 ("drm: Update file owner during use") Cc: <stable@vger.kernel.org> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2024-06-27kallsyms: rework symbol lookup return codesArnd Bergmann7-53/+49
Building with W=1 in some configurations produces a false positive warning for kallsyms: kernel/kallsyms.c: In function '__sprint_symbol.isra': kernel/kallsyms.c:503:17: error: 'strcpy' source argument is the same as destination [-Werror=restrict] 503 | strcpy(buffer, name); | ^~~~~~~~~~~~~~~~~~~~ This originally showed up while building with -O3, but later started happening in other configurations as well, depending on inlining decisions. The underlying issue is that the local 'name' variable is always initialized to the be the same as 'buffer' in the called functions that fill the buffer, which gcc notices while inlining, though it could see that the address check always skips the copy. The calling conventions here are rather unusual, as all of the internal lookup functions (bpf_address_lookup, ftrace_mod_address_lookup, ftrace_func_address_lookup, module_address_lookup and kallsyms_lookup_buildid) already use the provided buffer and either return the address of that buffer to indicate success, or NULL for failure, but the callers are written to also expect an arbitrary other buffer to be returned. Rework the calling conventions to return the length of the filled buffer instead of its address, which is simpler and easier to follow as well as avoiding the warning. Leave only the kallsyms_lookup() calling conventions unchanged, since that is called from 16 different functions and adapting this would be a much bigger change. Link: https://lore.kernel.org/lkml/20200107214042.855757-1-arnd@arndb.de/ Link: https://lore.kernel.org/lkml/20240326130647.7bfb1d92@gandalf.local.home/ Tested-by: Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by: Luis Chamberlain <mcgrof@kernel.org> Acked-by: Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2024-06-27gpiolib: cdev: Ignore reconfiguration without directionKent Gibson1-5/+7
linereq_set_config() behaves badly when direction is not set. The configuration validation is borrowed from linereq_create(), where, to verify the intent of the user, the direction must be set to in order to effect a change to the electrical configuration of a line. But, when applied to reconfiguration, that validation does not allow for the unset direction case, making it possible to clear flags set previously without specifying the line direction. Adding to the inconsistency, those changes are not immediately applied by linereq_set_config(), but will take effect when the line value is next get or set. For example, by requesting a configuration with no flags set, an output line with GPIO_V2_LINE_FLAG_ACTIVE_LOW and GPIO_V2_LINE_FLAG_OPEN_DRAIN set could have those flags cleared, inverting the sense of the line and changing the line drive to push-pull on the next line value set. Skip the reconfiguration of lines for which the direction is not set, and only reconfigure the lines for which direction is set. Fixes: a54756cb24ea ("gpiolib: cdev: support GPIO_V2_LINE_SET_CONFIG_IOCTL") Signed-off-by: Kent Gibson <warthog618@gmail.com> Link: https://lore.kernel.org/r/20240626052925.174272-3-warthog618@gmail.com Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
2024-06-27gpiolib: cdev: Disallow reconfiguration without direction (uAPI v1)Kent Gibson1-6/+10
linehandle_set_config() behaves badly when direction is not set. The configuration validation is borrowed from linehandle_create(), where, to verify the intent of the user, the direction must be set to in order to effect a change to the electrical configuration of a line. But, when applied to reconfiguration, that validation does not allow for the unset direction case, making it possible to clear flags set previously without specifying the line direction. Adding to the inconsistency, those changes are not immediately applied by linehandle_set_config(), but will take effect when the line value is next get or set. For example, by requesting a configuration with no flags set, an output line with GPIOHANDLE_REQUEST_ACTIVE_LOW and GPIOHANDLE_REQUEST_OPEN_DRAIN requested could have those flags cleared, inverting the sense of the line and changing the line drive to push-pull on the next line value set. Ensure the intent of the user by disallowing configurations which do not have direction set, returning an error to userspace to indicate that the configuration is invalid. And, for clarity, use lflags, a local copy of gcnf.flags, throughout when dealing with the requested flags, rather than a mixture of both. Fixes: e588bb1eae31 ("gpio: add new SET_CONFIG ioctl() to gpio chardev") Signed-off-by: Kent Gibson <warthog618@gmail.com> Link: https://lore.kernel.org/r/20240626052925.174272-2-warthog618@gmail.com Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
2024-06-27usb: dwc3: core: Workaround for CSR read timeoutJos Wang1-1/+19
This is a workaround for STAR 4846132, which only affects DWC_usb31 version2.00a operating in host mode. There is a problem in DWC_usb31 version 2.00a operating in host mode that would cause a CSR read timeout When CSR read coincides with RAM Clock Gating Entry. By disable Clock Gating, sacrificing power consumption for normal operation. Cc: stable <stable@kernel.org> # 5.10.x: 1e43c86d: usb: dwc3: core: Add DWC31 version 2.00a controller Signed-off-by: Jos Wang <joswang@lenovo.com> Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com> Link: https://lore.kernel.org/r/20240619114529.3441-1-joswang1221@gmail.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2024-06-27Revert "usb: gadget: u_ether: Replace netif_stop_queue with netif_device_detach"Ferry Toth1-1/+1
This reverts commit f49449fbc21e7e9550a5203902d69c8ae7dfd918. This commit breaks u_ether on some setups (at least Merrifield). The fix "usb: gadget: u_ether: Re-attach netif device to mirror detachment" party restores u-ether. However the netif usb: remains up even usb is switched from device to host mode. This creates problems for user space as the interface remains in the routing table while not realy present and network managers (connman) not detecting a network change. Various attempts to find the root cause were unsuccesful up to now. Therefore revert until a solution is found. Link: https://lore.kernel.org/linux-usb/20231006141231.7220-1-hgajjar@de.adit-jv.com/ Reported-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reported-by: Ferry Toth <fntoth@gmail.com> Fixes: f49449fbc21e ("usb: gadget: u_ether: Replace netif_stop_queue with netif_device_detach") Cc: stable@vger.kernel.org Signed-off-by: Ferry Toth <fntoth@gmail.com> Link: https://lore.kernel.org/r/20240620204832.24518-3-ftoth@exalondelft.nl Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2024-06-27Revert "usb: gadget: u_ether: Re-attach netif device to mirror detachment"Ferry Toth1-2/+0
This reverts commit 76c945730cdffb572c7767073cc6515fd3f646b4. Prerequisite revert for the reverting of the original commit f49449fbc21e. Fixes: 76c945730cdf ("usb: gadget: u_ether: Re-attach netif device to mirror detachment") Fixes: f49449fbc21e ("usb: gadget: u_ether: Replace netif_stop_queue with netif_device_detach") Reported-by: Ferry Toth <fntoth@gmail.com> Cc: stable@vger.kernel.org Signed-off-by: Ferry Toth <fntoth@gmail.com> Link: https://lore.kernel.org/r/20240620204832.24518-2-ftoth@exalondelft.nl Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2024-06-27staging: vchiq_debugfs: Fix build if CONFIG_DEBUG_FS is not setBernhard Rosenkränzer1-1/+1
Commit 42a2f6664e18 ("staging: vc04_services: Move global g_state to vchiq_state") adds a parameter to vchiq_debugfs_init, but leaves the dummy implementation in the !CONFIG_DEBUG_FS case untouched, causing a compile time error. Fixes: c3552ab19aeb ("staging: vchiq_debugfs: Fix NPD in vchiq_dump_state") Signed-off-by: Bernhard Rosenkränzer <bero@baylibre.com> Reviewed-by: Stefan Wahren <wahrenst@gmx.net> Link: https://lore.kernel.org/r/20240627124419.2498642-1-bero@baylibre.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2024-06-27net: mana: Fix possible double free in error handling pathMa Ke1-0/+2
When auxiliary_device_add() returns error and then calls auxiliary_device_uninit(), callback function adev_release calls kfree(madev). We shouldn't call kfree(madev) again in the error handling path. Set 'madev' to NULL. Fixes: a69839d4327d ("net: mana: Add support for auxiliary device") Signed-off-by: Ma Ke <make24@iscas.ac.cn> Link: https://patch.msgid.link/20240625130314.2661257-1-make24@iscas.ac.cn Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27iommu/amd: Fix GT feature enablement againVasant Hegde1-0/+1
Current code configures GCR3 even when device is attached to identity domain. So that we can support SVA with identity domain. This means in attach device path it updates Guest Translation related bits in DTE. Commit de111f6b4f6a ("iommu/amd: Enable Guest Translation after reading IOMMU feature register") missed to enable Control[GT] bit in resume path. Its causing certain laptop to fail to resume after suspend. This is because we have inconsistency between between control register (GT is disabled) and DTE (where we have enabled guest translation related bits) in resume path. And IOMMU hardware throws ILLEGAL_DEV_TABLE_ENTRY. Fix it by enabling GT bit in resume path. Reported-by: Błażej Szczygieł <spaz16@wp.pl> Link: https://bugzilla.kernel.org/show_bug.cgi?id=218975 Fixes: de111f6b4f6a ("iommu/amd: Enable Guest Translation after reading IOMMU feature register") Tested-by: Błażej Szczygieł <spaz16@wp.pl> Signed-off-by: Vasant Hegde <vasant.hegde@amd.com> Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com> Link: https://lore.kernel.org/r/20240621101533.20216-1-vasant.hegde@amd.com Signed-off-by: Joerg Roedel <jroedel@suse.de>
2024-06-27iommu/vt-d: Fix missed device TLB cache tagLu Baolu1-10/+10
When a domain is attached to a device, the required cache tags are assigned to the domain so that the related caches can be flushed whenever it is needed. The device TLB cache tag is created based on whether the ats_enabled field of the device's iommu data is set. This creates an ordered dependency between cache tag assignment and ATS enabling. The device TLB cache tag would not be created if device's ATS is enabled after the cache tag assignment. This causes devices with PCI ATS support to malfunction. The ATS control is exclusively owned by the iommu driver. Hence, move cache_tag_assign_domain() after PCI ATS enabling to make sure that the device TLB cache tag is created for the domain. Fixes: 3b1d9e2b2d68 ("iommu/vt-d: Add cache tag assignment interface") Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com> Reviewed-by: Kevin Tian <kevin.tian@intel.com> Link: https://lore.kernel.org/r/20240620062940.201786-1-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel <jroedel@suse.de>
2024-06-27iommu/amd: Invalidate cache before removing device from domain listVasant Hegde1-6/+6
Commit 87a6f1f22c97 ("iommu/amd: Introduce per-device domain ID to fix potential TLB aliasing issue") introduced per device domain ID when domain is configured with v2 page table. And in invalidation path, it uses per device structure (dev_data->gcr3_info.domid) to get the domain ID. In detach_device() path, current code tries to invalidate IOMMU cache after removing dev_data from domain device list. This means when domain is configured with v2 page table, amd_iommu_domain_flush_all() will not be able to invalidate cache as device is already removed from domain device list. This is causing change domain tests (changing domain type from identity to DMA) to fail with IO_PAGE_FAULT issue. Hence invalidate cache and update DTE before updating data structures. Reported-by: FahHean Lee <fahhean.lee@amd.com> Reported-by: Dheeraj Kumar Srivastava <dheerajkumar.srivastava@amd.com> Fixes: 87a6f1f22c97 ("iommu/amd: Introduce per-device domain ID to fix potential TLB aliasing issue") Tested-by: Dheeraj Kumar Srivastava <dheerajkumar.srivastava@amd.com> Tested-by: Sairaj Arun Kodilkar <sairaj.arunkodilkar@amd.com> Tested-by: FahHean Lee <fahhean.lee@amd.com> Signed-off-by: Vasant Hegde <vasant.hegde@amd.com> Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com> Link: https://lore.kernel.org/r/20240620060552.13984-1-vasant.hegde@amd.com Signed-off-by: Joerg Roedel <jroedel@suse.de>
2024-06-27selftest: af_unix: Check SIOCATMARK after every send()/recv() in msg_oob.c.Kuniyuki Iwashima1-0/+72
To catch regression, let's check ioctl(SIOCATMARK) after every send() and recv() calls. Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27af_unix: Fix wrong ioctl(SIOCATMARK) when consumed OOB skb is at the head.Kuniyuki Iwashima2-2/+81
Even if OOB data is recv()ed, ioctl(SIOCATMARK) must return 1 when the OOB skb is at the head of the receive queue and no new OOB data is queued. Without fix: # RUN msg_oob.no_peek.oob ... # msg_oob.c:305:oob:Expected answ[0] (0) == oob_head (1) # oob: Test terminated by assertion # FAIL msg_oob.no_peek.oob not ok 2 msg_oob.no_peek.oob With fix: # RUN msg_oob.no_peek.oob ... # OK msg_oob.no_peek.oob ok 2 msg_oob.no_peek.oob Fixes: 314001f0bf92 ("af_unix: Add OOB support") Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27selftest: af_unix: Check EPOLLPRI after every send()/recv() in msg_oob.cKuniyuki Iwashima1-0/+147
When OOB data is in recvq, we can detect it with epoll by checking EPOLLPRI. This patch add checks for EPOLLPRI after every send() and recv() in all test cases. Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27selftest: af_unix: Check SIGURG after every send() in msg_oob.cKuniyuki Iwashima1-1/+50
When data is sent with MSG_OOB, SIGURG is sent to a process if the receiver socket has set its owner to the process by ioctl(FIOSETOWN) or fcntl(F_SETOWN). This patch adds SIGURG check after every send(MSG_OOB) call. Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27selftest: af_unix: Add SO_OOBINLINE test cases in msg_oob.cKuniyuki Iwashima1-0/+91
When SO_OOBINLINE is enabled on a socket, MSG_OOB can be recv()ed without MSG_OOB flag, and ioctl(SIOCATMARK) will behaves differently. This patch adds some test cases for SO_OOBINLINE. Note the new test cases found two bugs in TCP. 1) After reading OOB data with non-inline mode, we can re-read the data by setting SO_OOBINLINE. # RUN msg_oob.no_peek.inline_oob_ahead_break ... # msg_oob.c:146:inline_oob_ahead_break:AF_UNIX :world # msg_oob.c:147:inline_oob_ahead_break:TCP :oworld # OK msg_oob.no_peek.inline_oob_ahead_break ok 14 msg_oob.no_peek.inline_oob_ahead_break 2) The head OOB data is dropped if SO_OOBINLINE is disabled if a new OOB data is queued. # RUN msg_oob.no_peek.inline_ex_oob_drop ... # msg_oob.c:171:inline_ex_oob_drop:AF_UNIX :x # msg_oob.c:172:inline_ex_oob_drop:TCP :y # msg_oob.c:146:inline_ex_oob_drop:AF_UNIX :y # msg_oob.c:147:inline_ex_oob_drop:TCP :Resource temporarily unavailable # OK msg_oob.no_peek.inline_ex_oob_drop ok 17 msg_oob.no_peek.inline_ex_oob_drop Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27af_unix: Don't stop recv() at consumed ex-OOB skb.Kuniyuki Iwashima2-1/+17
Currently, recv() is stopped at a consumed OOB skb even if a new OOB skb is queued and we can ignore the old OOB skb. >>> from socket import * >>> c1, c2 = socket(AF_UNIX, SOCK_STREAM) >>> c1.send(b'hellowor', MSG_OOB) 8 >>> c2.recv(1, MSG_OOB) # consume OOB data stays at middle of recvq. b'r' >>> c1.send(b'ld', MSG_OOB) 2 >>> c2.recv(10) # recv() stops at the old consumed OOB b'hellowo' # should be 'hellowol' manage_oob() should not stop recv() at the old consumed OOB skb if there is a new OOB data queued. Note that TCP behaviour is apparently wrong in this test case because we can recv() the same OOB data twice. Without fix: # RUN msg_oob.no_peek.ex_oob_ahead_break ... # msg_oob.c:138:ex_oob_ahead_break:AF_UNIX :hellowo # msg_oob.c:139:ex_oob_ahead_break:Expected:hellowol # msg_oob.c:141:ex_oob_ahead_break:Expected ret[0] (7) == expected_len (8) # ex_oob_ahead_break: Test terminated by assertion # FAIL msg_oob.no_peek.ex_oob_ahead_break not ok 11 msg_oob.no_peek.ex_oob_ahead_break With fix: # RUN msg_oob.no_peek.ex_oob_ahead_break ... # msg_oob.c:146:ex_oob_ahead_break:AF_UNIX :hellowol # msg_oob.c:147:ex_oob_ahead_break:TCP :helloworl # OK msg_oob.no_peek.ex_oob_ahead_break ok 11 msg_oob.no_peek.ex_oob_ahead_break Fixes: 314001f0bf92 ("af_unix: Add OOB support") Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27selftest: af_unix: Add non-TCP-compliant test cases in msg_oob.c.Kuniyuki Iwashima1-5/+44
While testing, I found some weird behaviour on the TCP side as well. For example, TCP drops the preceding OOB data when queueing a new OOB data if the old OOB data is at the head of recvq. # RUN msg_oob.no_peek.ex_oob_drop ... # msg_oob.c:146:ex_oob_drop:AF_UNIX :x # msg_oob.c:147:ex_oob_drop:TCP :Resource temporarily unavailable # msg_oob.c:146:ex_oob_drop:AF_UNIX :y # msg_oob.c:147:ex_oob_drop:TCP :Invalid argument # OK msg_oob.no_peek.ex_oob_drop ok 9 msg_oob.no_peek.ex_oob_drop # RUN msg_oob.no_peek.ex_oob_drop_2 ... # msg_oob.c:146:ex_oob_drop_2:AF_UNIX :x # msg_oob.c:147:ex_oob_drop_2:TCP :Resource temporarily unavailable # OK msg_oob.no_peek.ex_oob_drop_2 ok 10 msg_oob.no_peek.ex_oob_drop_2 This patch allows AF_UNIX's MSG_OOB implementation to produce different results from TCP when operations are guarded with tcp_incompliant{}. Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27af_unix: Don't stop recv(MSG_DONTWAIT) if consumed OOB skb is at the head.Kuniyuki Iwashima2-4/+26
Let's say a socket send()s "hello" with MSG_OOB and "world" without flags, >>> from socket import * >>> c1, c2 = socketpair(AF_UNIX) >>> c1.send(b'hello', MSG_OOB) 5 >>> c1.send(b'world') 5 and its peer recv()s "hell" and "o". >>> c2.recv(10) b'hell' >>> c2.recv(1, MSG_OOB) b'o' Now the consumed OOB skb stays at the head of recvq to return a correct value for ioctl(SIOCATMARK), which is broken now and fixed by a later patch. Then, if peer issues recv() with MSG_DONTWAIT, manage_oob() returns NULL, so recv() ends up with -EAGAIN. >>> c2.setblocking(False) # This causes -EAGAIN even with available data >>> c2.recv(5) Traceback (most recent call last): File "<stdin>", line 1, in <module> BlockingIOError: [Errno 11] Resource temporarily unavailable However, next recv() will return the following available data, "world". >>> c2.recv(5) b'world' When the consumed OOB skb is at the head of the queue, we need to fetch the next skb to fix the weird behaviour. Note that the issue does not happen without MSG_DONTWAIT because we can retry after manage_oob(). This patch also adds a test case that covers the issue. Without fix: # RUN msg_oob.no_peek.ex_oob_break ... # msg_oob.c:134:ex_oob_break:AF_UNIX :Resource temporarily unavailable # msg_oob.c:135:ex_oob_break:Expected:ld # msg_oob.c:137:ex_oob_break:Expected ret[0] (-1) == expected_len (2) # ex_oob_break: Test terminated by assertion # FAIL msg_oob.no_peek.ex_oob_break not ok 8 msg_oob.no_peek.ex_oob_break With fix: # RUN msg_oob.no_peek.ex_oob_break ... # OK msg_oob.no_peek.ex_oob_break ok 8 msg_oob.no_peek.ex_oob_break Fixes: 314001f0bf92 ("af_unix: Add OOB support") Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2024-06-27af_unix: Stop recv(MSG_PEEK) at consumed OOB skb.Kuniyuki Iwashima2-5/+29
After consuming OOB data, recv() reading the preceding data must break at the OOB skb regardless of MSG_PEEK. Currently, MSG_PEEK does not stop recv() for AF_UNIX, and the behaviour is not compliant with TCP. >>> from socket import * >>> c1, c2 = socketpair(AF_UNIX) >>> c1.send(b'hello', MSG_OOB) 5 >>> c1.send(b'world') 5 >>> c2.recv(1, MSG_OOB) b'o' >>> c2.recv(9, MSG_PEEK) # This should return b'hell' b'hellworld' # even with enough buffer. Let's fix it by returning NULL for consumed skb and unlinking it only if MSG_PEEK is not specified. This patch also adds test cases that add recv(MSG_PEEK) before each recv(). Without fix: # RUN msg_oob.peek.oob_ahead_break ... # msg_oob.c:134:oob_ahead_break:AF_UNIX :hellworld # msg_oob.c:135:oob_ahead_break:Expected:hell # msg_oob.c:137:oob_ahead_break:Expected ret[0] (9) == expected_len (4) # oob_ahead_break: Test terminated by assertion # FAIL msg_oob.peek.oob_ahead_break not ok 13 msg_oob.peek.oob_ahead_break With fix: # RUN msg_oob.peek.oob_ahead_break ... # OK msg_oob.peek.oob_ahead_break ok 13 msg_oob.peek.oob_ahead_break Fixes: 314001f0bf92 ("af_unix: Add OOB support") Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.