Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

invalid fd number return when open(2) is used. #11

Closed
thehajime opened this issue Nov 14, 2014 · 2 comments
Closed

invalid fd number return when open(2) is used. #11

thehajime opened this issue Nov 14, 2014 · 2 comments
Assignees
@thehajime

Comments

@thehajime
Copy link
Member

  • How to reproduce it
    • with raw socket

sudo NUSEDEV=eno16777736 nuse-eno16777736=192.168.209.39 ./nuse ../nginx-1.6.1/objs/nginx -c >pwd/../nginx-1.6.1/conf/nginx.conf -p ./

  • with tap

sudo NUSEDEV=eth1 NUSEVIF=TAP nuse-eth1=192.168.0.2 ./nuse nginx -c pwd/../nginx-1.6.1/conf/nginx.conf -p ./

both are closing socket for VIF, resulting no VIF channel after that (no RX/TX).
@thehajime thehajime self-assigned this Nov 14, 2014
@thehajime
Copy link
Member Author

(renamed the issue title)

https://github.com/thehajime/net-next-nuse/blob/nuse/arch/sim/nuse-glue.c#L431

int open (const char *pathname, int flags)
{
curfd++;
return host_open (pathname, flags);
}

this should NOT return a 'real' fd number, but pseudo fd num (curfd).

@thehajime thehajime changed the title read(2) in vif-raw/vif-tap fails with nginx-1.6.1 invalid fd number return when open(2) is used. Nov 14, 2014
thehajime added a commit that referenced this issue Nov 16, 2014
@thehajime
fix the issue #11. invalid fd number return when open(2) is used.
4d3692d 
 introduce 'struct nuse_fd' to handle real fd (file descriptor), epoll fd, and nuse fd. it's coverage needs to work more.
thehajime added a commit that referenced this issue Nov 17, 2014
@thehajime
fix the issue #11 take 2. invalid fd number return when open(2) is used.
b21e5a8 
unuse curfd variable and assign fake fd number by open ("/"..) for nuse_fd.
@thehajime
Copy link
Member Author

b21e5a8 is more improved version of the fix to this issue.

thehajime pushed a commit that referenced this issue Apr 8, 2015
@davem330
Merge branch 'dsa-next'
99949a7 
Andrew Lunn says:

====================
DSA Mavell drivers refactoring and cleanup

v1->v2:
 * Add missing signed-of-by: For patches authored by Guenter Roeck.
 * Add Reviewed by from Guenter Roack to patch #5.

This is a collection of patches again net-next from today containing
refactoring and consolidate of code, cleanups and using #define's to
replace register numbers.

Patch #1 Swaps the 6131 driver to use the consolidated setup code.

Patch #2 Moves the Switch IDs used during probe into a central
         location.  We need these later so that we can differentiate
         the different features the devices have.

Patch #3 Makes the 6131 driver set the number of ports in the private
         state structure. It then uses this, rather than hard coded
         maximum number of ports.

Patch #4 Similar to Patch #3, but for the 6123_61_65 driver.

Patch #5 Similar to Patch #3, and #4, but for all the remaining
         drivers.  This greatly increases the similarity of the code
         between drivers, allow further patches to consolidate the
         duplicated code.

Patch #6 Consolidate the switch reset code, which has two minor
         variants. Removes around 35 lines per driver.

Patch #7 Moves phy page access functions out of the 6352 driver into
         the shared code. Currently only the 6352 driver uses this,
         but it is likely other devices will come along wanting this
         functionality.

Patch #8 Consolidates the code used to access phy registers. Removes
         around 40 lines of code per driver.

Patch #9 Fixes missing mutex locking in the EEE code, and refactors
         the code a bit to make it more understandable with respect to
         locks.

Patch #10 Consolidates reading statistics. This is very similar code
          for all devices, but the number of available statistics
          differ, which can be determined from the product ID. Removes
	  around 65 lines per driver.

Patch #11 Add #defines for registers, and bits within the
          registers. For the moment, this is limited to the shared
          code. The individual drivers will be converted once the
          remaining duplicated code is consolidated

Patch #12 Fix broken statistic counters on the 6172. The 6352 family
          requires the port number is poked into a different set of
          bits in the register compared to other devices.

Many thanks to Guenter Roeck for repeatedly reviewing the patches and
testing them on his hardware.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
thehajime pushed a commit that referenced this issue Aug 14, 2015
@davem330
fq_codel: explicitly reset flows in ->reset()
3d0e0af 
Alex reported the following crash when using fq_codel
with htb:

  crash> bt
  PID: 630839  TASK: ffff8823c990d280  CPU: 14  COMMAND: "tc"
   [... snip ...]
   #8 [ffff8820ceec17a0] page_fault at ffffffff8160a8c2
      [exception RIP: htb_qlen_notify+24]
      RIP: ffffffffa0841718  RSP: ffff8820ceec1858  RFLAGS: 00010282
      RAX: 0000000000000000  RBX: 0000000000000000  RCX: ffff88241747b400
      RDX: ffff88241747b408  RSI: 0000000000000000  RDI: ffff8811fb27d000
      RBP: ffff8820ceec1868   R8: ffff88120cdeff24   R9: ffff88120cdeff30
      R10: 0000000000000bd4  R11: ffffffffa0840919  R12: ffffffffa0843340
      R13: 0000000000000000  R14: 0000000000000001  R15: ffff8808dae5c2e8
      ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
   #9 [...] qdisc_tree_decrease_qlen at ffffffff81565375
  #10 [...] fq_codel_dequeue at ffffffffa084e0a0 [sch_fq_codel]
  #11 [...] fq_codel_reset at ffffffffa084e2f8 [sch_fq_codel]
  #12 [...] qdisc_destroy at ffffffff81560d2d
  #13 [...] htb_destroy_class at ffffffffa08408f8 [sch_htb]
  #14 [...] htb_put at ffffffffa084095c [sch_htb]
  #15 [...] tc_ctl_tclass at ffffffff815645a3
  #16 [...] rtnetlink_rcv_msg at ffffffff81552cb0
  [... snip ...]

As Jamal pointed out, there is actually no need to call dequeue
to purge the queued skb's in reset, data structures can be just
reset explicitly. Therefore, we reset everything except config's
and stats, so that we would have a fresh start after device flipping.

Fixes: 4b549a2 ("fq_codel: Fair Queue Codel AQM")
Reported-by: Alex Gartrell <agartrell@fb.com>
Cc: Alex Gartrell <agartrell@fb.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
[xiyou.wangcong@gmail.com: added codel_vars_init() and qdisc_qstats_backlog_dec()]
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
thehajime pushed a commit that referenced this issue Aug 14, 2015
@torvalds
mm, vmscan: Do not wait for page writeback for GFP_NOFS allocations
ecf5fc6 
Nikolay has reported a hang when a memcg reclaim got stuck with the
following backtrace:

PID: 18308  TASK: ffff883d7c9b0a30  CPU: 1   COMMAND: "rsync"
  #0 __schedule at ffffffff815ab152
  #1 schedule at ffffffff815ab76e
  #2 schedule_timeout at ffffffff815ae5e5
  #3 io_schedule_timeout at ffffffff815aad6a
  #4 bit_wait_io at ffffffff815abfc6
  #5 __wait_on_bit at ffffffff815abda5
  #6 wait_on_page_bit at ffffffff8111fd4f
  #7 shrink_page_list at ffffffff81135445
  #8 shrink_inactive_list at ffffffff81135845
  #9 shrink_lruvec at ffffffff81135ead
 #10 shrink_zone at ffffffff811360c3
 #11 shrink_zones at ffffffff81136eff
 #12 do_try_to_free_pages at ffffffff8113712f
 #13 try_to_free_mem_cgroup_pages at ffffffff811372be
 #14 try_charge at ffffffff81189423
 #15 mem_cgroup_try_charge at ffffffff8118c6f5
 #16 __add_to_page_cache_locked at ffffffff8112137d
 #17 add_to_page_cache_lru at ffffffff81121618
 #18 pagecache_get_page at ffffffff8112170b
 #19 grow_dev_page at ffffffff811c8297
 #20 __getblk_slow at ffffffff811c91d6
 #21 __getblk_gfp at ffffffff811c92c1
 #22 ext4_ext_grow_indepth at ffffffff8124565c
 #23 ext4_ext_create_new_leaf at ffffffff81246ca8
 #24 ext4_ext_insert_extent at ffffffff81246f09
 #25 ext4_ext_map_blocks at ffffffff8124a848
 #26 ext4_map_blocks at ffffffff8121a5b7
 #27 mpage_map_one_extent at ffffffff8121b1fa
 #28 mpage_map_and_submit_extent at ffffffff8121f07b
 #29 ext4_writepages at ffffffff8121f6d5
 #30 do_writepages at ffffffff8112c490
 #31 __filemap_fdatawrite_range at ffffffff81120199
 #32 filemap_flush at ffffffff8112041c
 #33 ext4_alloc_da_blocks at ffffffff81219da1
 #34 ext4_rename at ffffffff81229b91
 #35 ext4_rename2 at ffffffff81229e32
 #36 vfs_rename at ffffffff811a08a5
 #37 SYSC_renameat2 at ffffffff811a3ffc
 #38 sys_renameat2 at ffffffff811a408e
 #39 sys_rename at ffffffff8119e51e
 #40 system_call_fastpath at ffffffff815afa89

Dave Chinner has properly pointed out that this is a deadlock in the
reclaim code because ext4 doesn't submit pages which are marked by
PG_writeback right away.

The heuristic was introduced by commit e62e384 ("memcg: prevent OOM
with too many dirty pages") and it was applied only when may_enter_fs
was specified.  The code has been changed by c3b94f4 ("memcg:
further prevent OOM with too many dirty pages") which has removed the
__GFP_FS restriction with a reasoning that we do not get into the fs
code.  But this is not sufficient apparently because the fs doesn't
necessarily submit pages marked PG_writeback for IO right away.

ext4_bio_write_page calls io_submit_add_bh but that doesn't necessarily
submit the bio.  Instead it tries to map more pages into the bio and
mpage_map_one_extent might trigger memcg charge which might end up
waiting on a page which is marked PG_writeback but hasn't been submitted
yet so we would end up waiting for something that never finishes.

Fix this issue by replacing __GFP_IO by may_enter_fs check (for case 2)
before we go to wait on the writeback.  The page fault path, which is
the only path that triggers memcg oom killer since 3.12, shouldn't
require GFP_NOFS and so we shouldn't reintroduce the premature OOM
killer issue which was originally addressed by the heuristic.

As per David Chinner the xfs is doing similar thing since 2.6.15 already
so ext4 is not the only affected filesystem.  Moreover he notes:

: For example: IO completion might require unwritten extent conversion
: which executes filesystem transactions and GFP_NOFS allocations. The
: writeback flag on the pages can not be cleared until unwritten
: extent conversion completes. Hence memory reclaim cannot wait on
: page writeback to complete in GFP_NOFS context because it is not
: safe to do so, memcg reclaim or otherwise.

Cc: stable@vger.kernel.org # 3.9+
[tytso@mit.edu: corrected the control flow]
Fixes: c3b94f4 ("memcg: further prevent OOM with too many dirty pages")
Reported-by: Nikolay Borisov <kernel@kyup.com>
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@thehajime thehajime

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@thehajime