Fix "kmem -n|-p" options display wrong values

Release crash-7.3.0-3

Resolves: rhbz#1975621

Signed-off-by: Lianbo Jiang <lijiang@redhat.com>
This commit is contained in:
Lianbo Jiang 2021-07-05 14:34:27 +08:00
parent a3ae8fcfd1
commit ccbee1fd3e
3 changed files with 157 additions and 1 deletions

View File

@ -0,0 +1,101 @@
From 0b5435e10161345cf713ed447a155a611a1b408b Mon Sep 17 00:00:00 2001
From: Kazuhito Hagio <k-hagio-ab@nec.com>
Date: Wed, 26 May 2021 17:33:13 +0900
Subject: [PATCH 1/2] memory: Add support for SECTION_TAINT_ZONE_DEVICE flag
Fix for "kmem -n|-p" options on Linux 5.12-rc1 and later kernels
that contain commit 1f90a3477df3f ("mm: teach pfn_to_online_page()
about ZONE_DEVICE section collisions"). Without the patch, the
"kmem -n" option incorrectly shows mem_map addresses containing the
flag in bit 5 as part of the virtual address, and also the "kmem -p"
option shows page structures at wrong position. With the patch,
the "kmem -n" option displays the new "D" state flag.
Without the patch:
crash> kmem -n
...
NR SECTION CODED_MEM_MAP MEM_MAP STATE PFN
1040 ffff9edf3ffd4100 ffffe2bcc0000010 ffffe2bd42000010 PMOE 34078720
^ ^
crash> kmem -p
PAGE PHYSICAL MAPPING INDEX CNT FLAGS
ffffe2bd42000010 2080000000 400040 1ffffffff 9961471 dead000000000122 referenced,active,error
ffffe2bd42000050 2080001000 800080 1ffffffff 9961471 dead000000000122 referenced,active,error
ffffe2bd42000090 2080002000 0 1ffffffff 9961471 dead000000000122 referenced,active,error
^^
With the patch:
crash> kmem -n
...
NR SECTION CODED_MEM_MAP MEM_MAP STATE PFN
1040 ffff9edf3ffd4100 ffffe2bcc0000000 ffffe2bd42000000 PMOED 34078720
crash> kmem -p
PAGE PHYSICAL MAPPING INDEX CNT FLAGS
ffffe2bd42000000 2080000000 ffff9ebfc0044100 0 1 97ffffc0000200 slab
ffffe2bd42000040 2080001000 ffff9ebfc0044400 0 1 97ffffc0000200 slab
ffffe2bd42000080 2080002000 0 0 1 97ffffc0000000
Signed-off-by: Kazuhito Hagio <k-hagio-ab@nec.com>
---
help.c | 11 +++++++----
memory.c | 15 +++++++++------
2 files changed, 16 insertions(+), 10 deletions(-)
diff --git a/help.c b/help.c
index e0c84087add3..9649cc81fa36 100644
--- a/help.c
+++ b/help.c
@@ -6584,10 +6584,13 @@ char *help_kmem[] = {
" kernels, the vm_zone_stat, vm_node_stat and vm_numa_stat tables,",
" the cumulative page_states counter values if they exist, and/or ",
" the cumulative, vm_event_states counter values if they exist.",
-" -n display memory node, memory section, and memory block data",
-" and state; the state of each memory section state is encoded",
-" as \"P\", \"M\", \"O\" and/or \"E\", meaning SECTION_MARKED_PRESENT,",
-" SECTION_HAS_MEM_MAP, SECTION_IS_ONLINE and SECTION_IS_EARLY.",
+" -n display memory node, memory section, memory block data and state;",
+" the state of each memory section is shown as the following flags:",
+" \"P\": SECTION_MARKED_PRESENT",
+" \"M\": SECTION_HAS_MEM_MAP",
+" \"O\": SECTION_IS_ONLINE",
+" \"E\": SECTION_IS_EARLY",
+" \"D\": SECTION_TAINT_ZONE_DEVICE",
" -z displays per-zone memory statistics.",
" -o displays each cpu's offset value that is added to per-cpu symbol",
" values to translate them into kernel virtual addresses.",
diff --git a/memory.c b/memory.c
index a3cf8a86728d..2c4f9790f498 100644
--- a/memory.c
+++ b/memory.c
@@ -17270,12 +17270,13 @@ nr_to_section(ulong nr)
* which results in PFN_SECTION_SHIFT equal 6.
* To sum it up, at least 6 bits are available.
*/
-#define SECTION_MARKED_PRESENT (1UL<<0)
-#define SECTION_HAS_MEM_MAP (1UL<<1)
-#define SECTION_IS_ONLINE (1UL<<2)
-#define SECTION_IS_EARLY (1UL<<3)
-#define SECTION_MAP_LAST_BIT (1UL<<4)
-#define SECTION_MAP_MASK (~(SECTION_MAP_LAST_BIT-1))
+#define SECTION_MARKED_PRESENT (1UL<<0)
+#define SECTION_HAS_MEM_MAP (1UL<<1)
+#define SECTION_IS_ONLINE (1UL<<2)
+#define SECTION_IS_EARLY (1UL<<3)
+#define SECTION_TAINT_ZONE_DEVICE (1UL<<4)
+#define SECTION_MAP_LAST_BIT (1UL<<5)
+#define SECTION_MAP_MASK (~(SECTION_MAP_LAST_BIT-1))
int
@@ -17373,6 +17374,8 @@ fill_mem_section_state(ulong state, char *buf)
bufidx += sprintf(buf + bufidx, "%s", "O");
if (state & SECTION_IS_EARLY)
bufidx += sprintf(buf + bufidx, "%s", "E");
+ if (state & SECTION_TAINT_ZONE_DEVICE)
+ bufidx += sprintf(buf + bufidx, "%s", "D");
}
void
--
2.30.2

View File

@ -0,0 +1,48 @@
From ec44b902d3467e7b86ee39e2d7d472b9cb202148 Mon Sep 17 00:00:00 2001
From: Kazuhito Hagio <k-hagio-ab@nec.com>
Date: Mon, 31 May 2021 14:08:28 +0900
Subject: [PATCH 2/2] memory: Fix for "kmem -n" option to display NID correctly
The nid member of struct memory_block is a 4-byte integer, but read
and printed as a 8-byte integer on 64-bit machines. Without the
patch, the option displays wrong NIDs.
crash> kmem -n
...
MEM_BLOCK NAME PHYSICAL RANGE NODE STATE START_SECTION_NO
ffff9edeff2b9400 memory0 0 - 7fffffff 14195095130662240256 ONLINE 0
ffff9edeff2bb400 memory2 100000000 - 17fffffff 14195094718345379840 ONLINE 32
The issue seems to appear on Linux 5.12 and later kernels that contain
commit e9a2e48e8704c ("drivers/base/memory: don't store phys_device
in memory blocks"), which changed the arrangement of the members of
struct memory_block.
Signed-off-by: Kazuhito Hagio <k-hagio-ab@nec.com>
---
memory.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/memory.c b/memory.c
index 2c4f9790f498..cbe90eebe748 100644
--- a/memory.c
+++ b/memory.c
@@ -17568,13 +17568,13 @@ print_memory_block(ulong memory_block)
if (MEMBER_EXISTS("memory_block", "nid")) {
readmem(memory_block + OFFSET(memory_block_nid), KVADDR, &nid,
- sizeof(void *), "memory_block nid", FAULT_ON_ERROR);
+ sizeof(int), "memory_block nid", FAULT_ON_ERROR);
fprintf(fp, " %s %s %s %s %s %s\n",
mkstring(buf1, VADDR_PRLEN, LJUST|LONG_HEX,
MKSTR(memory_block)),
mkstring(buf2, 12, CENTER, name),
parangebuf,
- mkstring(buf5, strlen("NODE"), CENTER|LONG_DEC,
+ mkstring(buf5, strlen("NODE"), CENTER|INT_DEC,
MKSTR(nid)),
mkstring(buf6, strlen("OFFLINE"), LJUST,
statebuf),
--
2.30.2

View File

@ -4,7 +4,7 @@
Summary: Kernel analysis utility for live systems, netdump, diskdump, kdump, LKCD or mcore dumpfiles
Name: crash
Version: 7.3.0
Release: 2%{?dist}
Release: 3%{?dist}
License: GPLv3
Source0: https://github.com/crash-utility/crash/archive/crash-%{version}.tar.gz
Source1: http://ftp.gnu.org/gnu/gdb/gdb-7.6.tar.gz
@ -19,6 +19,8 @@ Provides: bundled(libiberty)
Provides: bundled(gdb) = 7.6
Patch0: lzo_snappy.patch
Patch1: 0001-Fix-for-kmem-s-S-option-on-Linux-5.7-and-later-kerne.patch
Patch2: 0001-memory-Add-support-for-SECTION_TAINT_ZONE_DEVICE-fla.patch
Patch3: 0002-memory-Fix-for-kmem-n-option-to-display-NID-correctl.patch
%description
The core analysis suite is a self-contained tool that can be used to
@ -40,6 +42,8 @@ offered by Mission Critical Linux, or the LKCD kernel patch.
%setup -n %{name}-%{version} -q
%patch0 -p1 -b lzo_snappy.patch
%patch1 -p1
%patch2 -p1
%patch3 -p1
%build
# This package has an internal copy of GDB which has broken configure code for
@ -71,6 +75,9 @@ cp -p defs.h %{buildroot}%{_includedir}/crash
%{_includedir}/*
%changelog
* Mon Jul 05 2021 Lianbo Jiang <lijiang@redhat.com> - 7.3.0-3
- Fix "kmem -n|-p" options display wrong values.
* Fri Jun 11 2021 Lianbo Jiang <lijiang@redhat.com> - 7.3.0-2
- Fix for "kmem -s|-S" option on Linux 5.7 and later kernels