[Linux-cachefs] 3.0.3 64-bit Crash running fscache/cachefilesd

[David Howells]

You'll probably need to add the attached patch also.  Turns out there were
some bits I'd missed.

David
---
From: David Howells <dhowells at redhat.com>
Subject: [PATCH] CacheFiles: Add missing retrieval completions

CacheFiles is missing some calls to fscache_retrieval_complete() in the error
handling/collision paths of its reader functions.

This can be seen by the following assertion tripping in fscache_put_operation()
whereby the operation being destroyed is still in the in-progress state and has
not been cancelled or completed:

FS-Cache: Assertion failed
3 == 5 is false
------------[ cut here ]------------
kernel BUG at fs/fscache/operation.c:408!
invalid opcode: 0000 [#1] SMP
CPU 2
Modules linked in: xfs ioatdma dca loop joydev evdev
psmouse dcdbas pcspkr serio_raw i5000_edac edac_core i5k_amb shpchp
pci_hotplug sg sr_mod]

Pid: 8062, comm: httpd Not tainted 3.1.0-rc8 #1 Dell Inc. PowerEdge 1950/0DT097
RIP: 0010:[<ffffffff81197b24>]  [<ffffffff81197b24>] fscache_put_operation+0x304/0x330
RSP: 0018:ffff880062f739d8  EFLAGS: 00010296
RAX: 0000000000000025 RBX: ffff8800c5122e84 RCX: ffffffff81ddf040
RDX: 00000000ffffffff RSI: 0000000000000082 RDI: ffffffff81ddef30
RBP: ffff880062f739f8 R08: 0000000000000005 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000003 R12: ffff8800c5122e40
R13: ffff880037a2cd20 R14: ffff880087c7a058 R15: ffff880087c7a000
FS:  00007f63dcf636e0(0000) GS:ffff88022fc80000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f0c0a91f000 CR3: 0000000062ec2000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process httpd (pid: 8062, threadinfo ffff880062f72000, task ffff880087e58000)
Stack:
 ffff880062f73bf8 0000000000000000 ffff880062f73bf8 ffff880037a2cd20
 ffff880062f73a68 ffffffff8119aa7e ffff88006540e000 ffff880062f73ad4
 ffff88008e9a4308 ffff880037a2cd20 ffff880062f73a48 ffff8800c5122e40
Call Trace:
 [<ffffffff8119aa7e>] __fscache_read_or_alloc_pages+0x1fe/0x530
 [<ffffffff81250780>] __nfs_readpages_from_fscache+0x70/0x1c0
 [<ffffffff8123142a>] nfs_readpages+0xca/0x1e0
 [<ffffffff815f3c06>] ? rpc_do_put_task+0x36/0x50
 [<ffffffff8122755b>] ? alloc_nfs_open_context+0x4b/0x110
 [<ffffffff815ecd1a>] ? rpc_call_sync+0x5a/0x70
 [<ffffffff810e7e9a>] __do_page_cache_readahead+0x1ca/0x270
 [<ffffffff810e7f61>] ra_submit+0x21/0x30
 [<ffffffff810e818d>] ondemand_readahead+0x11d/0x250
 [<ffffffff810e83b6>] page_cache_sync_readahead+0x36/0x60
 [<ffffffff810dffa4>] generic_file_aio_read+0x454/0x770
 [<ffffffff81224ce1>] nfs_file_read+0xe1/0x130
 [<ffffffff81121bd9>] do_sync_read+0xd9/0x120
 [<ffffffff8114088f>] ? mntput+0x1f/0x40
 [<ffffffff811238cb>] ? fput+0x1cb/0x260
 [<ffffffff81122938>] vfs_read+0xc8/0x180
 [<ffffffff81122af5>] sys_read+0x55/0x90

Reported-by: Mark Moseley <moseleymark at gmail.com>
Signed-off-by: David Howells <dhowells at redhat.com>
---

 fs/cachefiles/rdwr.c |   14 ++++++++++----
 fs/fscache/page.c    |    2 ++
 2 files changed, 12 insertions(+), 4 deletions(-)


diff --git a/fs/cachefiles/rdwr.c b/fs/cachefiles/rdwr.c
index 637a27d..eb9ab4b 100644
--- a/fs/cachefiles/rdwr.c
+++ b/fs/cachefiles/rdwr.c
@@ -361,8 +361,10 @@ out:
 
 read_error:
 	_debug("read error %d", ret);
-	if (ret == -ENOMEM)
+	if (ret == -ENOMEM) {
+		fscache_retrieval_complete(op, 1);
 		goto out;
+	}
 io_error:
 	cachefiles_io_error_obj(object, "Page read error on backing file");
 	fscache_retrieval_complete(op, 1);
@@ -551,6 +553,7 @@ static int cachefiles_read_backing_file(struct cachefiles_object *object,
 		if (ret < 0) {
 			if (ret == -EEXIST) {
 				page_cache_release(netpage);
+				fscache_retrieval_complete(op, 1);
 				continue;
 			}
 			goto nomem;
@@ -627,6 +630,7 @@ static int cachefiles_read_backing_file(struct cachefiles_object *object,
 		if (ret < 0) {
 			if (ret == -EEXIST) {
 				page_cache_release(netpage);
+				fscache_retrieval_complete(op, 1);
 				continue;
 			}
 			goto nomem;
@@ -645,9 +649,9 @@ static int cachefiles_read_backing_file(struct cachefiles_object *object,
 
 		/* the netpage is unlocked and marked up to date here */
 		fscache_end_io(op, netpage, 0);
-		fscache_retrieval_complete(op, 1);
 		page_cache_release(netpage);
 		netpage = NULL;
+		fscache_retrieval_complete(op, 1);
 		continue;
 	}
 
@@ -682,15 +686,17 @@ out:
 nomem:
 	_debug("nomem");
 	ret = -ENOMEM;
-	goto out;
+	goto record_page_complete;
 
 read_error:
 	_debug("read error %d", ret);
 	if (ret == -ENOMEM)
-		goto out;
+		goto record_page_complete;
 io_error:
 	cachefiles_io_error_obj(object, "Page read error on backing file");
 	ret = -ENOBUFS;
+record_page_complete:
+	fscache_retrieval_complete(op, 1);
 	goto out;
 }
 
diff --git a/fs/fscache/page.c b/fs/fscache/page.c
index a30c157..00a5ed9 100644
--- a/fs/fscache/page.c
+++ b/fs/fscache/page.c
@@ -329,6 +329,8 @@ check_if_dead:
 		return -ENOBUFS;
 	}
 	if (unlikely(fscache_object_is_dead(object))) {
+		pr_err("%s() = -ENOBUFS [obj dead %d]", __func__, op->op.state);
+		fscache_cancel_op(&op->op);
 		fscache_stat(stat_object_dead);
 		return -ENOBUFS;
 	}