| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Tue, 15 Mar 2022 13:05:11 +0800
From: Xiubo Li <xiubli@...hat.com>
To: kernel test robot <lkp@...el.com>
Cc: kbuild-all@...ts.01.org, linux-kernel@...r.kernel.org,
Jeff Layton <jlayton@...nel.org>
Subject: Re: [jlayton:ceph-fscrypt-sparse 66/74] fs/ceph/dir.c:424:31: error:
implicit declaration of function 'ceph_encode_encrypted_dname'; did you mean
'ceph_encode_encrypted_fname'?
It failed without the FSCRYPT config, and the following patch will fix it:
diff --git a/fs/ceph/crypto.h b/fs/ceph/crypto.h
index ee593b8ad07c..1ed318565d86 100644
--- a/fs/ceph/crypto.h
+++ b/fs/ceph/crypto.h
@@ -172,6 +172,13 @@ static inline void
ceph_fscrypt_as_ctx_to_req(struct ceph_mds_request *req,
{
}
+static inline int ceph_encode_encrypted_dname(const struct inode *parent,
+ struct qstr *d_name, char
*buf)
+{
+ memcpy(buf, d_name->name, d_name->len);
+ return d_name->len;
+}
+
static inline int ceph_encode_encrypted_fname(const struct inode *parent,
struct dentry *dentry,
char *buf)
{
On 3/15/22 12:18 PM, kernel test robot wrote:
> tree: https://git.kernel.org/pub/scm/linux/kernel/git/jlayton/linux.git ceph-fscrypt-sparse
> head: 0d355bda666e44314f11974167e8d9bb010a539e
> commit: 4378b10dd7df64ef085e3fe2a2251034ee98c4dc [66/74] ceph: dencrypt the dentry names early and once for readdir
> config: openrisc-randconfig-m031-20220314 (https://download.01.org/0day-ci/archive/20220315/202203151243.JKEqZf3z-lkp@intel.com/config)
> compiler: or1k-linux-gcc (GCC) 11.2.0
> reproduce (this is a W=1 build):
> wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
> chmod +x ~/bin/make.cross
> # https://git.kernel.org/pub/scm/linux/kernel/git/jlayton/linux.git/commit/?id=4378b10dd7df64ef085e3fe2a2251034ee98c4dc
> git remote add jlayton https://git.kernel.org/pub/scm/linux/kernel/git/jlayton/linux.git
> git fetch --no-tags jlayton ceph-fscrypt-sparse
> git checkout 4378b10dd7df64ef085e3fe2a2251034ee98c4dc
> # save the config file to linux build tree
> mkdir build_dir
> COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-11.2.0 make.cross O=build_dir ARCH=openrisc SHELL=/bin/bash fs/ceph/
>
> If you fix the issue, kindly add following tag as appropriate
> Reported-by: kernel test robot <lkp@...el.com>
>
> All errors (new ones prefixed by >>):
>
> fs/ceph/dir.c: In function 'ceph_readdir':
>>> fs/ceph/dir.c:424:31: error: implicit declaration of function 'ceph_encode_encrypted_dname'; did you mean 'ceph_encode_encrypted_fname'? [-Werror=implicit-function-declaration]
> 424 | err = ceph_encode_encrypted_dname(inode, &d_name, req->r_path2);
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~
> | ceph_encode_encrypted_fname
> cc1: some warnings being treated as errors
>
>
> vim +424 fs/ceph/dir.c
>
> 307
> 308 static int ceph_readdir(struct file *file, struct dir_context *ctx)
> 309 {
> 310 struct ceph_dir_file_info *dfi = file->private_data;
> 311 struct inode *inode = file_inode(file);
> 312 struct ceph_inode_info *ci = ceph_inode(inode);
> 313 struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
> 314 struct ceph_mds_client *mdsc = fsc->mdsc;
> 315 int i;
> 316 int err;
> 317 unsigned frag = -1;
> 318 struct ceph_mds_reply_info_parsed *rinfo;
> 319
> 320 dout("readdir %p file %p pos %llx\n", inode, file, ctx->pos);
> 321 if (dfi->file_info.flags & CEPH_F_ATEND)
> 322 return 0;
> 323
> 324 /* always start with . and .. */
> 325 if (ctx->pos == 0) {
> 326 dout("readdir off 0 -> '.'\n");
> 327 if (!dir_emit(ctx, ".", 1, ceph_present_inode(inode),
> 328 inode->i_mode >> 12))
> 329 return 0;
> 330 ctx->pos = 1;
> 331 }
> 332 if (ctx->pos == 1) {
> 333 u64 ino;
> 334 struct dentry *dentry = file->f_path.dentry;
> 335
> 336 spin_lock(&dentry->d_lock);
> 337 ino = ceph_present_inode(dentry->d_parent->d_inode);
> 338 spin_unlock(&dentry->d_lock);
> 339
> 340 dout("readdir off 1 -> '..'\n");
> 341 if (!dir_emit(ctx, "..", 2, ino, inode->i_mode >> 12))
> 342 return 0;
> 343 ctx->pos = 2;
> 344 }
> 345
> 346 err = fscrypt_prepare_readdir(inode);
> 347 if (err)
> 348 return err;
> 349
> 350 spin_lock(&ci->i_ceph_lock);
> 351 /* request Fx cap. if have Fx, we don't need to release Fs cap
> 352 * for later create/unlink. */
> 353 __ceph_touch_fmode(ci, mdsc, CEPH_FILE_MODE_WR);
> 354 /* can we use the dcache? */
> 355 if (ceph_test_mount_opt(fsc, DCACHE) &&
> 356 !ceph_test_mount_opt(fsc, NOASYNCREADDIR) &&
> 357 ceph_snap(inode) != CEPH_SNAPDIR &&
> 358 __ceph_dir_is_complete_ordered(ci) &&
> 359 __ceph_caps_issued_mask_metric(ci, CEPH_CAP_FILE_SHARED, 1)) {
> 360 int shared_gen = atomic_read(&ci->i_shared_gen);
> 361
> 362 spin_unlock(&ci->i_ceph_lock);
> 363 err = __dcache_readdir(file, ctx, shared_gen);
> 364 if (err != -EAGAIN)
> 365 return err;
> 366 } else {
> 367 spin_unlock(&ci->i_ceph_lock);
> 368 }
> 369
> 370 /* proceed with a normal readdir */
> 371 more:
> 372 /* do we have the correct frag content buffered? */
> 373 if (need_send_readdir(dfi, ctx->pos)) {
> 374 struct ceph_mds_request *req;
> 375 int op = ceph_snap(inode) == CEPH_SNAPDIR ?
> 376 CEPH_MDS_OP_LSSNAP : CEPH_MDS_OP_READDIR;
> 377
> 378 /* discard old result, if any */
> 379 if (dfi->last_readdir) {
> 380 ceph_mdsc_put_request(dfi->last_readdir);
> 381 dfi->last_readdir = NULL;
> 382 }
> 383
> 384 if (is_hash_order(ctx->pos)) {
> 385 /* fragtree isn't always accurate. choose frag
> 386 * based on previous reply when possible. */
> 387 if (frag == (unsigned)-1)
> 388 frag = ceph_choose_frag(ci, fpos_hash(ctx->pos),
> 389 NULL, NULL);
> 390 } else {
> 391 frag = fpos_frag(ctx->pos);
> 392 }
> 393
> 394 dout("readdir fetching %llx.%llx frag %x offset '%s'\n",
> 395 ceph_vinop(inode), frag, dfi->last_name);
> 396 req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
> 397 if (IS_ERR(req)) {
> 398 err = PTR_ERR(req);
> 399 goto out;
> 400 }
> 401 err = ceph_alloc_readdir_reply_buffer(req, inode);
> 402 if (err) {
> 403 ceph_mdsc_put_request(req);
> 404 goto out;
> 405 }
> 406 /* hints to request -> mds selection code */
> 407 req->r_direct_mode = USE_AUTH_MDS;
> 408 if (op == CEPH_MDS_OP_READDIR) {
> 409 req->r_direct_hash = ceph_frag_value(frag);
> 410 __set_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);
> 411 req->r_inode_drop = CEPH_CAP_FILE_EXCL;
> 412 }
> 413 if (dfi->last_name) {
> 414 struct qstr d_name = { .name = dfi->last_name,
> 415 .len = strlen(dfi->last_name) };
> 416
> 417 req->r_path2 = kzalloc(NAME_MAX + 1, GFP_KERNEL);
> 418 if (!req->r_path2) {
> 419 ceph_mdsc_put_request(req);
> 420 err = -ENOMEM;
> 421 goto out;
> 422 }
> 423
> > 424 err = ceph_encode_encrypted_dname(inode, &d_name, req->r_path2);
> 425 if (err < 0) {
> 426 ceph_mdsc_put_request(req);
> 427 goto out;
> 428 }
> 429 } else if (is_hash_order(ctx->pos)) {
> 430 req->r_args.readdir.offset_hash =
> 431 cpu_to_le32(fpos_hash(ctx->pos));
> 432 }
> 433
> 434 req->r_dir_release_cnt = dfi->dir_release_count;
> 435 req->r_dir_ordered_cnt = dfi->dir_ordered_count;
> 436 req->r_readdir_cache_idx = dfi->readdir_cache_idx;
> 437 req->r_readdir_offset = dfi->next_offset;
> 438 req->r_args.readdir.frag = cpu_to_le32(frag);
> 439 req->r_args.readdir.flags =
> 440 cpu_to_le16(CEPH_READDIR_REPLY_BITFLAGS);
> 441
> 442 req->r_inode = inode;
> 443 ihold(inode);
> 444 req->r_dentry = dget(file->f_path.dentry);
> 445 err = ceph_mdsc_do_request(mdsc, NULL, req);
> 446 if (err < 0) {
> 447 ceph_mdsc_put_request(req);
> 448 goto out;
> 449 }
> 450 dout("readdir got and parsed readdir result=%d on "
> 451 "frag %x, end=%d, complete=%d, hash_order=%d\n",
> 452 err, frag,
> 453 (int)req->r_reply_info.dir_end,
> 454 (int)req->r_reply_info.dir_complete,
> 455 (int)req->r_reply_info.hash_order);
> 456
> 457 rinfo = &req->r_reply_info;
> 458 if (le32_to_cpu(rinfo->dir_dir->frag) != frag) {
> 459 frag = le32_to_cpu(rinfo->dir_dir->frag);
> 460 if (!rinfo->hash_order) {
> 461 dfi->next_offset = req->r_readdir_offset;
> 462 /* adjust ctx->pos to beginning of frag */
> 463 ctx->pos = ceph_make_fpos(frag,
> 464 dfi->next_offset,
> 465 false);
> 466 }
> 467 }
> 468
> 469 dfi->frag = frag;
> 470 dfi->last_readdir = req;
> 471
> 472 if (test_bit(CEPH_MDS_R_DID_PREPOPULATE, &req->r_req_flags)) {
> 473 dfi->readdir_cache_idx = req->r_readdir_cache_idx;
> 474 if (dfi->readdir_cache_idx < 0) {
> 475 /* preclude from marking dir ordered */
> 476 dfi->dir_ordered_count = 0;
> 477 } else if (ceph_frag_is_leftmost(frag) &&
> 478 dfi->next_offset == 2) {
> 479 /* note dir version at start of readdir so
> 480 * we can tell if any dentries get dropped */
> 481 dfi->dir_release_count = req->r_dir_release_cnt;
> 482 dfi->dir_ordered_count = req->r_dir_ordered_cnt;
> 483 }
> 484 } else {
> 485 dout("readdir !did_prepopulate\n");
> 486 /* disable readdir cache */
> 487 dfi->readdir_cache_idx = -1;
> 488 /* preclude from marking dir complete */
> 489 dfi->dir_release_count = 0;
> 490 }
> 491
> 492 /* note next offset and last dentry name */
> 493 if (rinfo->dir_nr > 0) {
> 494 struct ceph_mds_reply_dir_entry *rde =
> 495 rinfo->dir_entries + (rinfo->dir_nr-1);
> 496 unsigned next_offset = req->r_reply_info.dir_end ?
> 497 2 : (fpos_off(rde->offset) + 1);
> 498 err = note_last_dentry(dfi, rde->name, rde->name_len,
> 499 next_offset);
> 500 if (err) {
> 501 ceph_mdsc_put_request(dfi->last_readdir);
> 502 dfi->last_readdir = NULL;
> 503 goto out;
> 504 }
> 505 } else if (req->r_reply_info.dir_end) {
> 506 dfi->next_offset = 2;
> 507 /* keep last name */
> 508 }
> 509 }
> 510
> 511 rinfo = &dfi->last_readdir->r_reply_info;
> 512 dout("readdir frag %x num %d pos %llx chunk first %llx\n",
> 513 dfi->frag, rinfo->dir_nr, ctx->pos,
> 514 rinfo->dir_nr ? rinfo->dir_entries[0].offset : 0LL);
> 515
> 516 i = 0;
> 517 /* search start position */
> 518 if (rinfo->dir_nr > 0) {
> 519 int step, nr = rinfo->dir_nr;
> 520 while (nr > 0) {
> 521 step = nr >> 1;
> 522 if (rinfo->dir_entries[i + step].offset < ctx->pos) {
> 523 i += step + 1;
> 524 nr -= step + 1;
> 525 } else {
> 526 nr = step;
> 527 }
> 528 }
> 529 }
> 530 for (; i < rinfo->dir_nr; i++) {
> 531 struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
> 532
> 533 BUG_ON(rde->offset < ctx->pos);
> 534 BUG_ON(!rde->inode.in);
> 535
> 536 ctx->pos = rde->offset;
> 537 dout("readdir (%d/%d) -> %llx '%.*s' %p\n",
> 538 i, rinfo->dir_nr, ctx->pos,
> 539 rde->name_len, rde->name, &rde->inode.in);
> 540
> 541 if (!dir_emit(ctx, rde->name, rde->name_len,
> 542 ceph_present_ino(inode->i_sb, le64_to_cpu(rde->inode.in->ino)),
> 543 le32_to_cpu(rde->inode.in->mode) >> 12)) {
> 544 /*
> 545 * NOTE: Here no need to put the 'dfi->last_readdir',
> 546 * because when dir_emit stops us it's most likely
> 547 * doesn't have enough memory, etc. So for next readdir
> 548 * it will continue.
> 549 */
> 550 dout("filldir stopping us...\n");
> 551 err = 0;
> 552 goto out;
> 553 }
> 554
> 555 /* Reset the lengths to their original allocated vals */
> 556 ctx->pos++;
> 557 }
> 558
> 559 ceph_mdsc_put_request(dfi->last_readdir);
> 560 dfi->last_readdir = NULL;
> 561
> 562 if (dfi->next_offset > 2) {
> 563 frag = dfi->frag;
> 564 goto more;
> 565 }
> 566
> 567 /* more frags? */
> 568 if (!ceph_frag_is_rightmost(dfi->frag)) {
> 569 frag = ceph_frag_next(dfi->frag);
> 570 if (is_hash_order(ctx->pos)) {
> 571 loff_t new_pos = ceph_make_fpos(ceph_frag_value(frag),
> 572 dfi->next_offset, true);
> 573 if (new_pos > ctx->pos)
> 574 ctx->pos = new_pos;
> 575 /* keep last_name */
> 576 } else {
> 577 ctx->pos = ceph_make_fpos(frag, dfi->next_offset,
> 578 false);
> 579 kfree(dfi->last_name);
> 580 dfi->last_name = NULL;
> 581 }
> 582 dout("readdir next frag is %x\n", frag);
> 583 goto more;
> 584 }
> 585 dfi->file_info.flags |= CEPH_F_ATEND;
> 586
> 587 /*
> 588 * if dir_release_count still matches the dir, no dentries
> 589 * were released during the whole readdir, and we should have
> 590 * the complete dir contents in our cache.
> 591 */
> 592 if (atomic64_read(&ci->i_release_count) ==
> 593 dfi->dir_release_count) {
> 594 spin_lock(&ci->i_ceph_lock);
> 595 if (dfi->dir_ordered_count ==
> 596 atomic64_read(&ci->i_ordered_count)) {
> 597 dout(" marking %p complete and ordered\n", inode);
> 598 /* use i_size to track number of entries in
> 599 * readdir cache */
> 600 BUG_ON(dfi->readdir_cache_idx < 0);
> 601 i_size_write(inode, dfi->readdir_cache_idx *
> 602 sizeof(struct dentry*));
> 603 } else {
> 604 dout(" marking %p complete\n", inode);
> 605 }
> 606 __ceph_dir_set_complete(ci, dfi->dir_release_count,
> 607 dfi->dir_ordered_count);
> 608 spin_unlock(&ci->i_ceph_lock);
> 609 }
> 610 err = 0;
> 611 dout("readdir %p file %p done.\n", inode, file);
> 612 out:
> 613 return err;
> 614 }
> 615
>
> ---
> 0-DAY CI Kernel Test Service
> https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
>
Powered by blists - more mailing lists