From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Al Viro <viro@zeniv.linux.org.uk>
Subject: [PATCH 5.4 54/57] epoll: replace ->visited/visited_list with generation count
Date: Mon, 5 Oct 2020 17:27:06 +0200 [thread overview]
Message-ID: <20201005142112.398368208@linuxfoundation.org> (raw)
In-Reply-To: <20201005142109.796046410@linuxfoundation.org>
From: Al Viro <viro@zeniv.linux.org.uk>
commit 18306c404abe18a0972587a6266830583c60c928 upstream.
removes the need to clear it, along with the races.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
fs/eventpoll.c | 26 +++++++-------------------
1 file changed, 7 insertions(+), 19 deletions(-)
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -218,8 +218,7 @@ struct eventpoll {
struct file *file;
/* used to optimize loop detection check */
- int visited;
- struct list_head visited_list_link;
+ u64 gen;
#ifdef CONFIG_NET_RX_BUSY_POLL
/* used to track busy poll napi_id */
@@ -269,6 +268,8 @@ static long max_user_watches __read_most
*/
static DEFINE_MUTEX(epmutex);
+static u64 loop_check_gen = 0;
+
/* Used to check for epoll file descriptor inclusion loops */
static struct nested_calls poll_loop_ncalls;
@@ -278,9 +279,6 @@ static struct kmem_cache *epi_cache __re
/* Slab cache used to allocate "struct eppoll_entry" */
static struct kmem_cache *pwq_cache __read_mostly;
-/* Visited nodes during ep_loop_check(), so we can unset them when we finish */
-static LIST_HEAD(visited_list);
-
/*
* List of files with newly added links, where we may need to limit the number
* of emanating paths. Protected by the epmutex.
@@ -1968,13 +1966,12 @@ static int ep_loop_check_proc(void *priv
struct epitem *epi;
mutex_lock_nested(&ep->mtx, call_nests + 1);
- ep->visited = 1;
- list_add(&ep->visited_list_link, &visited_list);
+ ep->gen = loop_check_gen;
for (rbp = rb_first_cached(&ep->rbr); rbp; rbp = rb_next(rbp)) {
epi = rb_entry(rbp, struct epitem, rbn);
if (unlikely(is_file_epoll(epi->ffd.file))) {
ep_tovisit = epi->ffd.file->private_data;
- if (ep_tovisit->visited)
+ if (ep_tovisit->gen == loop_check_gen)
continue;
error = ep_call_nested(&poll_loop_ncalls,
ep_loop_check_proc, epi->ffd.file,
@@ -2015,18 +2012,8 @@ static int ep_loop_check_proc(void *priv
*/
static int ep_loop_check(struct eventpoll *ep, struct file *file)
{
- int ret;
- struct eventpoll *ep_cur, *ep_next;
-
- ret = ep_call_nested(&poll_loop_ncalls,
+ return ep_call_nested(&poll_loop_ncalls,
ep_loop_check_proc, file, ep, current);
- /* clear visited list */
- list_for_each_entry_safe(ep_cur, ep_next, &visited_list,
- visited_list_link) {
- ep_cur->visited = 0;
- list_del(&ep_cur->visited_list_link);
- }
- return ret;
}
static void clear_tfile_check_list(void)
@@ -2248,6 +2235,7 @@ SYSCALL_DEFINE4(epoll_ctl, int, epfd, in
error_tgt_fput:
if (full_check) {
clear_tfile_check_list();
+ loop_check_gen++;
mutex_unlock(&epmutex);
}
next prev parent reply other threads:[~2020-10-05 15:39 UTC|newest]
Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-10-05 15:26 [PATCH 5.4 00/57] 5.4.70-rc1 review Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 01/57] btrfs: fix filesystem corruption after a device replace Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 02/57] mmc: sdhci: Workaround broken command queuing on Intel GLK based IRBIS models Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 03/57] USB: gadget: f_ncm: Fix NDP16 datagram validation Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 04/57] gpio: siox: explicitly support only threaded irqs Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 05/57] gpio: mockup: fix resource leak in error path Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 06/57] gpio: tc35894: fix up tc35894 interrupt configuration Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 07/57] clk: socfpga: stratix10: fix the divider for the emac_ptp_free_clk Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 08/57] vsock/virtio: add transport parameter to the virtio_transport_reset_no_sock() Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 09/57] net: virtio_vsock: Enhance connection semantics Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 10/57] xfs: trim IO to found COW extent limit Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 11/57] Input: i8042 - add nopnp quirk for Acer Aspire 5 A515 Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 12/57] iio: adc: qcom-spmi-adc5: fix driver name Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 13/57] ftrace: Move RCU is watching check after recursion check Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 14/57] memstick: Skip allocating card when removing host Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 15/57] drm/amdgpu: restore proper ref count in amdgpu_display_crtc_set_config Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 16/57] clocksource/drivers/timer-gx6605s: Fixup counter reload Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 17/57] libbpf: Remove arch-specific include path in Makefile Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 18/57] drivers/net/wan/hdlc_fr: Add needed_headroom for PVC devices Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 19/57] drm/sun4i: mixer: Extend regmap max_register Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 20/57] net: dec: de2104x: Increase receive ring size for Tulip Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 21/57] rndis_host: increase sleep time in the query-response loop Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 22/57] nvme-core: get/put ctrl and transport module in nvme_dev_open/release() Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 23/57] fuse: fix the ->direct_IO() treatment of iov_iter Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 24/57] drivers/net/wan/lapbether: Make skb->protocol consistent with the header Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 25/57] drivers/net/wan/hdlc: Set skb->protocol before transmitting Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 26/57] mac80211: Fix radiotap header channel flag for 6GHz band Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 27/57] mac80211: do not allow bigger VHT MPDUs than the hardware supports Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 28/57] tracing: Make the space reserved for the pid wider Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 29/57] tools/io_uring: fix compile breakage Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 30/57] spi: fsl-espi: Only process interrupts for expected events Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 31/57] nvme-pci: fix NULL req in completion handler Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 32/57] nvme-fc: fail new connections to a deleted host or remote port Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 33/57] gpio: sprd: Clear interrupt when setting the type as edge Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 34/57] phy: ti: am654: Fix a leak in serdes_am654_probe() Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 35/57] pinctrl: mvebu: Fix i2c sda definition for 98DX3236 Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 36/57] nfs: Fix security label length not being reset Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 37/57] clk: tegra: Always program PLL_E when enabled Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 38/57] clk: samsung: exynos4: mark chipid clock as CLK_IGNORE_UNUSED Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 39/57] iommu/exynos: add missing put_device() call in exynos_iommu_of_xlate() Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 40/57] gpio/aspeed-sgpio: enable access to all 80 input & output sgpios Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 41/57] gpio/aspeed-sgpio: dont enable all interrupts by default Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 42/57] gpio: aspeed: fix ast2600 bank properties Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 43/57] i2c: cpm: Fix i2c_ram structure Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 44/57] Input: trackpoint - enable Synaptics trackpoints Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 45/57] scripts/dtc: only append to HOST_EXTRACFLAGS instead of overwriting Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 46/57] random32: Restore __latent_entropy attribute on net_rand_state Greg Kroah-Hartman
2020-10-05 15:26 ` [PATCH 5.4 47/57] block/diskstats: more accurate approximation of io_ticks for slow disks Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 48/57] mm: replace memmap_context by meminit_context Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 49/57] mm: dont rely on system state to detect hot-plug operations Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 50/57] nvme: Cleanup and rename nvme_block_nr() Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 51/57] nvme: Introduce nvme_lba_to_sect() Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 52/57] nvme: consolidate chunk_sectors settings Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 53/57] epoll: do not insert into poll queues until all sanity checks are done Greg Kroah-Hartman
2020-10-05 15:27 ` Greg Kroah-Hartman [this message]
2020-10-05 15:27 ` [PATCH 5.4 55/57] epoll: EPOLL_CTL_ADD: close the race in decision to take fast path Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 56/57] ep_create_wakeup_source(): dentry name can change under you Greg Kroah-Hartman
2020-10-05 15:27 ` [PATCH 5.4 57/57] netfilter: ctnetlink: add a range check for l3/l4 protonum Greg Kroah-Hartman
2020-10-05 17:50 ` [PATCH 5.4 00/57] 5.4.70-rc1 review Jon Hunter
2020-10-06 0:22 ` Shuah Khan
2020-10-06 5:54 ` Naresh Kamboju
2020-10-06 8:25 ` Naresh Kamboju
2020-10-06 14:49 ` Ben Hutchings
2020-10-06 18:17 ` Guenter Roeck
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20201005142112.398368208@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
--cc=viro@zeniv.linux.org.uk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).