From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
patches@lists.linux.dev,
kernel test robot <oliver.sang@intel.com>,
Dave Chinner <dchinner@redhat.com>,
"Darrick J. Wong" <djwong@kernel.org>,
Christoph Hellwig <hch@lst.de>,
Dave Chinner <david@fromorbit.com>,
Leah Rumancik <leah.rumancik@gmail.com>,
Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.15 29/83] xfs: avoid unnecessary runtime sibling pointer endian conversions
Date: Mon, 20 Feb 2023 14:36:02 +0100 [thread overview]
Message-ID: <20230220133554.692656858@linuxfoundation.org> (raw)
In-Reply-To: <20230220133553.669025851@linuxfoundation.org>
From: Dave Chinner <dchinner@redhat.com>
[ Upstream commit 5672225e8f2a872a22b0cecedba7a6644af1fb84 ]
Commit dc04db2aa7c9 has caused a small aim7 regression, showing a
small increase in CPU usage in __xfs_btree_check_sblock() as a
result of the extra checking.
This is likely due to the endian conversion of the sibling poitners
being unconditional instead of relying on the compiler to endian
convert the NULL pointer at compile time and avoiding the runtime
conversion for this common case.
Rework the checks so that endian conversion of the sibling pointers
is only done if they are not null as the original code did.
.... and these need to be "inline" because the compiler completely
fails to inline them automatically like it should be doing.
$ size fs/xfs/libxfs/xfs_btree.o*
text data bss dec hex filename
51874 240 0 52114 cb92 fs/xfs/libxfs/xfs_btree.o.orig
51562 240 0 51802 ca5a fs/xfs/libxfs/xfs_btree.o.inline
Just when you think the tools have advanced sufficiently we don't
have to care about stuff like this anymore, along comes a reminder
that *our tools still suck*.
Fixes: dc04db2aa7c9 ("xfs: detect self referencing btree sibling pointers")
Reported-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Signed-off-by: Leah Rumancik <leah.rumancik@gmail.com>
Acked-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
fs/xfs/libxfs/xfs_btree.c | 47 +++++++++++++++++++++++++++------------
1 file changed, 33 insertions(+), 14 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c
index 5bec048343b0c..b4b5bf4bfed7f 100644
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -51,16 +51,31 @@ xfs_btree_magic(
return magic;
}
-static xfs_failaddr_t
+/*
+ * These sibling pointer checks are optimised for null sibling pointers. This
+ * happens a lot, and we don't need to byte swap at runtime if the sibling
+ * pointer is NULL.
+ *
+ * These are explicitly marked at inline because the cost of calling them as
+ * functions instead of inlining them is about 36 bytes extra code per call site
+ * on x86-64. Yes, gcc-11 fails to inline them, and explicit inlining of these
+ * two sibling check functions reduces the compiled code size by over 300
+ * bytes.
+ */
+static inline xfs_failaddr_t
xfs_btree_check_lblock_siblings(
struct xfs_mount *mp,
struct xfs_btree_cur *cur,
int level,
xfs_fsblock_t fsb,
- xfs_fsblock_t sibling)
+ __be64 dsibling)
{
- if (sibling == NULLFSBLOCK)
+ xfs_fsblock_t sibling;
+
+ if (dsibling == cpu_to_be64(NULLFSBLOCK))
return NULL;
+
+ sibling = be64_to_cpu(dsibling);
if (sibling == fsb)
return __this_address;
if (level >= 0) {
@@ -74,17 +89,21 @@ xfs_btree_check_lblock_siblings(
return NULL;
}
-static xfs_failaddr_t
+static inline xfs_failaddr_t
xfs_btree_check_sblock_siblings(
struct xfs_mount *mp,
struct xfs_btree_cur *cur,
int level,
xfs_agnumber_t agno,
xfs_agblock_t agbno,
- xfs_agblock_t sibling)
+ __be32 dsibling)
{
- if (sibling == NULLAGBLOCK)
+ xfs_agblock_t sibling;
+
+ if (dsibling == cpu_to_be32(NULLAGBLOCK))
return NULL;
+
+ sibling = be32_to_cpu(dsibling);
if (sibling == agbno)
return __this_address;
if (level >= 0) {
@@ -136,10 +155,10 @@ __xfs_btree_check_lblock(
fsb = XFS_DADDR_TO_FSB(mp, xfs_buf_daddr(bp));
fa = xfs_btree_check_lblock_siblings(mp, cur, level, fsb,
- be64_to_cpu(block->bb_u.l.bb_leftsib));
+ block->bb_u.l.bb_leftsib);
if (!fa)
fa = xfs_btree_check_lblock_siblings(mp, cur, level, fsb,
- be64_to_cpu(block->bb_u.l.bb_rightsib));
+ block->bb_u.l.bb_rightsib);
return fa;
}
@@ -204,10 +223,10 @@ __xfs_btree_check_sblock(
}
fa = xfs_btree_check_sblock_siblings(mp, cur, level, agno, agbno,
- be32_to_cpu(block->bb_u.s.bb_leftsib));
+ block->bb_u.s.bb_leftsib);
if (!fa)
fa = xfs_btree_check_sblock_siblings(mp, cur, level, agno,
- agbno, be32_to_cpu(block->bb_u.s.bb_rightsib));
+ agbno, block->bb_u.s.bb_rightsib);
return fa;
}
@@ -4517,10 +4536,10 @@ xfs_btree_lblock_verify(
/* sibling pointer verification */
fsb = XFS_DADDR_TO_FSB(mp, xfs_buf_daddr(bp));
fa = xfs_btree_check_lblock_siblings(mp, NULL, -1, fsb,
- be64_to_cpu(block->bb_u.l.bb_leftsib));
+ block->bb_u.l.bb_leftsib);
if (!fa)
fa = xfs_btree_check_lblock_siblings(mp, NULL, -1, fsb,
- be64_to_cpu(block->bb_u.l.bb_rightsib));
+ block->bb_u.l.bb_rightsib);
return fa;
}
@@ -4574,10 +4593,10 @@ xfs_btree_sblock_verify(
agno = xfs_daddr_to_agno(mp, xfs_buf_daddr(bp));
agbno = xfs_daddr_to_agbno(mp, xfs_buf_daddr(bp));
fa = xfs_btree_check_sblock_siblings(mp, NULL, -1, agno, agbno,
- be32_to_cpu(block->bb_u.s.bb_leftsib));
+ block->bb_u.s.bb_leftsib);
if (!fa)
fa = xfs_btree_check_sblock_siblings(mp, NULL, -1, agno, agbno,
- be32_to_cpu(block->bb_u.s.bb_rightsib));
+ block->bb_u.s.bb_rightsib);
return fa;
}
--
2.39.0
next prev parent reply other threads:[~2023-02-20 13:52 UTC|newest]
Thread overview: 91+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-20 13:35 [PATCH 5.15 00/83] 5.15.95-rc1 review Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 01/83] mptcp: fix locking for in-kernel listener creation Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 02/83] kprobes: treewide: Cleanup the error messages for kprobes Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 03/83] riscv: kprobe: Fixup misaligned load text Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 04/83] ACPI / x86: Add support for LPS0 callback handler Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 05/83] ASoC: Intel: sof_rt5682: always set dpcm_capture for amplifiers Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 06/83] ASoC: Intel: sof_cs42l42: " Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 07/83] selftests/bpf: Verify copy_register_state() preserves parent/live fields Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 08/83] ALSA: hda: Do not unset preset when cleaning up codec Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 09/83] bpf, sockmap: Dont let sock_map_{close,destroy,unhash} call itself Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 10/83] ASoC: cs42l56: fix DT probe Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 11/83] tools/virtio: fix the vringh test for virtio ring changes Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 12/83] net/rose: Fix to not accept on connected socket Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 13/83] net: stmmac: do not stop RX_CLK in Rx LPI state for qcs404 SoC Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 14/83] drm/nouveau/devinit/tu102-: wait for GFW_BOOT_PROGRESS == COMPLETED Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 15/83] net: sched: sch: Bounds check priority Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 16/83] s390/decompressor: specify __decompress() buf len to avoid overflow Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 17/83] nvme-fc: fix a missing queue put in nvmet_fc_ls_create_association Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 18/83] drm/amd/display: Properly handle additional cases where DCN is not supported Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 19/83] platform/x86: touchscreen_dmi: Add Chuwi Vi8 (CWI501) DMI match Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 20/83] nvmem: core: add error handling for dev_set_name Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 21/83] nvmem: core: fix cleanup after dev_set_name() Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 22/83] nvmem: core: fix registration vs use race Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 23/83] nvmem: core: fix return value Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 24/83] xfs: zero inode fork buffer at allocation Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 25/83] xfs: fix potential log item leak Greg Kroah-Hartman
2023-02-20 13:35 ` [PATCH 5.15 26/83] xfs: detect self referencing btree sibling pointers Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 27/83] xfs: set XFS_FEAT_NLINK correctly Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 28/83] xfs: validate v5 feature fields Greg Kroah-Hartman
2023-02-20 13:36 ` Greg Kroah-Hartman [this message]
2023-02-20 13:36 ` [PATCH 5.15 30/83] xfs: dont assert fail on perag references on teardown Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 31/83] xfs: assert in xfs_btree_del_cursor should take into account error Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 32/83] xfs: purge dquots after inode walk fails during quotacheck Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 33/83] xfs: dont leak btree cursor when insrec fails after a split Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 34/83] mptcp: do not wait for bare sockets timeout Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 35/83] aio: fix mremap after fork null-deref Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 36/83] drm/amd/display: Fail atomic_check early on normalize_zpos error Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 37/83] platform/x86: amd-pmc: Export Idlemask values based on the APU Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 38/83] platform/x86: amd-pmc: Fix compilation when CONFIG_DEBUGFS is disabled Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 39/83] platform/x86: amd-pmc: Correct usage of SMU version Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 40/83] platform/x86/amd: pmc: Disable IRQ1 wakeup for RN/CZN Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 41/83] netfilter: nft_tproxy: restrict to prerouting hook Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 42/83] tcp: Fix listen() regression in 5.15.88 Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 43/83] mmc: jz4740: Work around bug on JZ4760(B) Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 44/83] mmc: sdio: fix possible resource leaks in some error paths Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 45/83] mmc: mmc_spi: fix error handling in mmc_spi_probe() Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 46/83] ALSA: hda/conexant: add a new hda codec SN6180 Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 47/83] ALSA: hda/realtek - fixed wrong gpio assigned Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 48/83] sched/psi: Fix use-after-free in ep_remove_wait_queue() Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 49/83] hugetlb: check for undefined shift on 32 bit architectures Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 50/83] of: reserved_mem: Have kmemleak ignore dynamically allocated reserved mem Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 51/83] selftest/lkdtm: Skip stack-entropy test if lkdtm is not available Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 52/83] Revert "mm: Always release pages to the buddy allocator in memblock_free_late()." Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 53/83] net: Fix unwanted sign extension in netdev_stats_to_stats64() Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 54/83] revert "squashfs: harden sanity check in squashfs_read_xattr_id_table" Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 55/83] ixgbe: allow to increase MTU to 3K with XDP enabled Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 56/83] i40e: add double of VLAN header when computing the max MTU Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 57/83] net: bgmac: fix BCM5358 support by setting correct flags Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 58/83] net: ethernet: ti: am65-cpsw: Add RX DMA Channel Teardown Quirk Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 59/83] sctp: sctp_sock_filter(): avoid list_entry() on possibly empty list Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 60/83] net/sched: tcindex: update imperfect hash filters respecting rcu Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 61/83] dccp/tcp: Avoid negative sk_forward_alloc by ipv6_pinfo.pktoptions Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 62/83] net/usb: kalmia: Dont pass act_len in usb_bulk_msg error path Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 63/83] net: openvswitch: fix possible memory leak in ovs_meter_cmd_set() Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 64/83] net: stmmac: fix order of dwmac5 FlexPPS parametrization sequence Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 65/83] bnxt_en: Fix mqprio and XDP ring checking logic Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 66/83] net: stmmac: Restrict warning on disabling DMA store and fwd mode Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 67/83] net: mpls: fix stale pointer if allocation fails during device rename Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 68/83] ixgbe: add double of VLAN header when computing the max MTU Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 69/83] ipv6: Fix datagram socket connection with DSCP Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 70/83] ipv6: Fix tcp " Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 71/83] nilfs2: fix underflow in second superblock position calculations Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 72/83] mm/filemap: fix page end in filemap_get_read_batch Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 73/83] drm/i915/gen11: Moving WAs to icl_gt_workarounds_init() Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 74/83] drm/i915/gen11: Wa_1408615072/Wa_1407596294 should be on GT list Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 75/83] flow_offload: fill flags to action structure Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 76/83] net/sched: act_ctinfo: use percpu stats Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 77/83] i40e: Add checking for null for nlmsg_find_attr() Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 78/83] net/sched: tcindex: search key must be 16 bits Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 79/83] kvm: initialize all of the kvm_debugregs structure before sending it to userspace Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 80/83] alarmtimer: Prevent starvation by small intervals and SIG_IGN Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 81/83] ASoC: SOF: Intel: hda-dai: fix possible stream_tag leak Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 82/83] net: sched: sch: Fix off by one in htb_activate_prios() Greg Kroah-Hartman
2023-02-20 13:36 ` [PATCH 5.15 83/83] platform/x86/amd: pmc: add CONFIG_SERIO dependency Greg Kroah-Hartman
2023-02-21 4:10 ` [PATCH 5.15 00/83] 5.15.95-rc1 review Bagas Sanjaya
2023-02-21 7:09 ` Naresh Kamboju
2023-02-21 12:04 ` Ron Economos
2023-02-21 14:57 ` Sudip Mukherjee (Codethink)
2023-02-21 16:21 ` Guenter Roeck
2023-02-21 20:10 ` Florian Fainelli
2023-02-21 23:48 ` Shuah Khan
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=20230220133554.692656858@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=david@fromorbit.com \
--cc=dchinner@redhat.com \
--cc=djwong@kernel.org \
--cc=hch@lst.de \
--cc=leah.rumancik@gmail.com \
--cc=oliver.sang@intel.com \
--cc=patches@lists.linux.dev \
--cc=sashal@kernel.org \
--cc=stable@vger.kernel.org \
/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).