All of lore.kernel.org
 help / color / mirror / Atom feed
* [syzbot] [mtd?] BUG: unable to handle kernel paging request in phram_read
@ 2026-11-18  1:49 ` syzbot
  0 siblings, 0 replies; 9+ messages in thread
From: syzbot @ 2026-11-18  1:49 UTC (permalink / raw)
  To: joern, linux-kernel, linux-mtd, miquel.raynal, richard,
	syzkaller-bugs, vigneshr

Hello,

syzbot found the following issue on:

HEAD commit:    ad3fc05c7e65 arm64: suspend: Only track cpuidle context fo..
git tree:       git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git for-kernelci
console output: https://syzkaller.appspot.com/x/log.txt?x=165de94e580000
kernel config:  https://syzkaller.appspot.com/x/.config?x=4e8e6c1701ad8af2
dashboard link: https://syzkaller.appspot.com/bug?extid=ee09ee2112d03c35f1f7
compiler:       Debian clang version 22.1.8 (1~deb13u4), Debian LLD 22.1.8
userspace arch: arm64
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=1796373d580000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=13fbe5c3580000

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/a30beb3b2b82/disk-ad3fc05c.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/2eaac3d8da90/vmlinux-ad3fc05c.xz
kernel image: https://storage.googleapis.com/syzbot-assets/212a161a069b/Image-ad3fc05c.gz.xz

The issue was bisected by AI to candidate commits (most relevant first):

1da177e4c3f4 Linux-2.6.12-rc2
f3d45ac21e46 mtd: phram: Module parameters add writable permissions
285133040e6c arm64: Import latest memcpy()/memmove() implementation

AI reasoning:   https://syzkaller.appspot.com/ai_job?id=c053ce4a-8b95-428d-87ad-63a0065b8ed9

AI-generated root cause summary:
In drivers/mtd/devices/phram.c, registering a phram device via module parameter
or sysfs (/sys/module/phram/parameters/phram) invokes register_device() with
pdev = NULL, which causes new->cached to be false. As a result, phram_map() maps
the physical memory region using ioremap() (mapped as PROT_DEVICE_nGnRE on
arm64) and forces the pointer into mtd->priv, yet phram_read(), phram_write(),
and phram_erase() access it using standard memcpy() and memset() instead of I/O
memory accessors (like memcpy_fromio()). When scanning partition headers (e.g.,
ftl_add_mtd() reading a 68-byte header via mtd_read()), arm64's memcpy
implementation issues pair-load (ldp) instructions at unaligned offsets, which
unconditionally triggers an alignment fault (FSC = 0x21) on Device memory and
crashes the kernel.

full analysis:  https://syzkaller.appspot.com/x/ai_root_cause?x=ee09ee2112d03c35f1f7

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+ee09ee2112d03c35f1f7@syzkaller.appspotmail.com

Unable to handle kernel paging request at virtual address ffff800090ffe024
KASAN: probably user-memory-access in range [0x0000000487ff0120-0x0000000487ff0127]
Mem abort info:
  ESR = 0x0000000096000021
  EC = 0x25: DABT (current EL), IL = 32 bits
  SET = 0, FnV = 0
  EA = 0, S1PTW = 0
  FSC = 0x21: alignment fault
Data abort info:
  ISV = 0, ISS = 0x00000021, ISS2 = 0x00000000
  CM = 0, WnR = 0, TnD = 0, TagAccess = 0
  GCS = 0, Overlay = 0, DirtyBit = 0
swapper pgtable: 4k pages, 48-bit VAs, pgdp=00000002193b5000
[ffff800090ffe024] pgd=0000000000000000, p4d=100000021f6f0003, pud=100000021f6f1003, pmd=10000001013e1403, pte=0068000000000713
Internal error: Oops: 0000000096000021 [#1]  SMP
Modules linked in:
CPU: 0 UID: 0 PID: 4957 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT 
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 11/06/2026
pstate: 23400005 (nzCv daif +PAN -UAO +TCO +DIT -SSBS BTYPE=--)
pc : __pi_memcpy_generic+0x98/0x22c arch/arm64/lib/memcpy.S:114
lr : __asan_memcpy+0x68/0x84 mm/kasan/shadow.c:109
sp : ffff800095836fc0
x29: ffff800095836fc0 x28: 1fffe00019f6102f x27: ffff0000cfb08000
x26: dfff800000000000 x25: ffff800095837150 x24: dfff800000000000
x23: ffff0000cfb08190 x22: ffff800082ab0064 x21: ffff800095837370
x20: ffff800090ffe000 x19: 0000000000000044 x18: 1fffe000359cf228
x17: 4a414d0064746d3d x16: 4d45545359534255 x15: ffff800080d0603c
x14: 0000000000000001 x13: 0000000000000009 x12: 0000000000000004
x11: ffff700012b06e76 x10: ffff700012b06e76 x9 : ffffffffffffffff
x8 : ffffffffffffffff x7 : ffffffffffffffff x6 : ffffffff14000400
x5 : ffff8000958373b4 x4 : ffff800090ffe044 x3 : ffff800082ab0064
x2 : 0000000000000044 x1 : ffff800090ffe000 x0 : ffff800095837370
Call trace:
 __pi_memcpy_generic+0x98/0x22c arch/arm64/lib/memcpy.S:113 (P)
 phram_read+0x64/0x98 drivers/mtd/devices/phram.c:70
 mtd_read_oob_std+0x180/0x19c drivers/mtd/mtdcore.c:1702
 mtd_read_oob+0x1fc/0x4ec drivers/mtd/mtdcore.c:1811
 mtd_read+0xd0/0x1a8 drivers/mtd/mtdcore.c:1603
 scan_header drivers/mtd/ftl.c:164 [inline]
 ftl_add_mtd+0x1e8/0x16d8 drivers/mtd/ftl.c:1017
 blktrans_notify_add+0xa8/0xf8 drivers/mtd/mtd_blkdevs.c:469
 add_mtd_device+0xedc/0x1504 drivers/mtd/mtdcore.c:828
 mtd_device_parse_register+0x660/0x7a0 drivers/mtd/mtdcore.c:1156
 register_device+0x340/0x51c drivers/mtd/devices/phram.c:162
 phram_setup+0x354/0x4fc drivers/mtd/devices/phram.c:324
 phram_param_call+0x34/0x90 drivers/mtd/devices/phram.c:349
 param_attr_store+0x144/0x1bc kernel/params.c:591
 module_attr_store+0x6c/0x94 kernel/params.c:906
 sysfs_kf_write+0xe0/0x108 fs/sysfs/file.c:145
 kernfs_fop_write_iter+0x260/0x3b8 fs/kernfs/file.c:345
 new_sync_write fs/read_write.c:595 [inline]
 vfs_write+0x51c/0xa0c fs/read_write.c:687
 ksys_write+0x12c/0x228 fs/read_write.c:739
 __do_sys_write fs/read_write.c:750 [inline]
 __se_sys_write fs/read_write.c:747 [inline]
 __arm64_sys_write+0x84/0x98 fs/read_write.c:747
 __invoke_syscall arch/arm64/kernel/syscall.c:35 [inline]
 invoke_syscall+0x98/0x244 arch/arm64/kernel/syscall.c:49
 el0_svc_common+0xec/0x23c arch/arm64/kernel/syscall.c:121
 do_el0_svc+0x4c/0x5c arch/arm64/kernel/syscall.c:140
 el0_svc+0x60/0x1d8 arch/arm64/kernel/entry-common.c:797
 el0t_64_sync_handler+0x44/0x104 arch/arm64/kernel/entry-common.c:816
 el0t_64_sync+0x198/0x19c arch/arm64/kernel/entry.S:606
Code: d65f03c0 d503201f a9401c26 a9412428 (a97e2c8a) 
---[ end trace 0000000000000000 ]---
----------------
Code disassembly (best guess):
   0:	d65f03c0 	ret
   4:	d503201f 	nop
   8:	a9401c26 	ldp	x6, x7, [x1]
   c:	a9412428 	ldp	x8, x9, [x1, #16]
* 10:	a97e2c8a 	ldp	x10, x11, [x4, #-32] <-- trapping instruction


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
For information about AI bisection see: https://goo.gl/tpsmEJ#ai-bisection

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

^ permalink raw reply	[flat|nested] 9+ messages in thread

* [syzbot] [mtd?] BUG: unable to handle kernel paging request in phram_read
@ 2026-11-18  1:49 ` syzbot
  0 siblings, 0 replies; 9+ messages in thread
From: syzbot @ 2026-11-18  1:49 UTC (permalink / raw)
  To: joern, linux-kernel, linux-mtd, miquel.raynal, richard,
	syzkaller-bugs, vigneshr

Hello,

syzbot found the following issue on:

HEAD commit:    ad3fc05c7e65 arm64: suspend: Only track cpuidle context fo..
git tree:       git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git for-kernelci
console output: https://syzkaller.appspot.com/x/log.txt?x=165de94e580000
kernel config:  https://syzkaller.appspot.com/x/.config?x=4e8e6c1701ad8af2
dashboard link: https://syzkaller.appspot.com/bug?extid=ee09ee2112d03c35f1f7
compiler:       Debian clang version 22.1.8 (1~deb13u4), Debian LLD 22.1.8
userspace arch: arm64
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=1796373d580000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=13fbe5c3580000

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/a30beb3b2b82/disk-ad3fc05c.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/2eaac3d8da90/vmlinux-ad3fc05c.xz
kernel image: https://storage.googleapis.com/syzbot-assets/212a161a069b/Image-ad3fc05c.gz.xz

The issue was bisected by AI to candidate commits (most relevant first):

1da177e4c3f4 Linux-2.6.12-rc2
f3d45ac21e46 mtd: phram: Module parameters add writable permissions
285133040e6c arm64: Import latest memcpy()/memmove() implementation

AI reasoning:   https://syzkaller.appspot.com/ai_job?id=c053ce4a-8b95-428d-87ad-63a0065b8ed9

AI-generated root cause summary:
In drivers/mtd/devices/phram.c, registering a phram device via module parameter
or sysfs (/sys/module/phram/parameters/phram) invokes register_device() with
pdev = NULL, which causes new->cached to be false. As a result, phram_map() maps
the physical memory region using ioremap() (mapped as PROT_DEVICE_nGnRE on
arm64) and forces the pointer into mtd->priv, yet phram_read(), phram_write(),
and phram_erase() access it using standard memcpy() and memset() instead of I/O
memory accessors (like memcpy_fromio()). When scanning partition headers (e.g.,
ftl_add_mtd() reading a 68-byte header via mtd_read()), arm64's memcpy
implementation issues pair-load (ldp) instructions at unaligned offsets, which
unconditionally triggers an alignment fault (FSC = 0x21) on Device memory and
crashes the kernel.

full analysis:  https://syzkaller.appspot.com/x/ai_root_cause?x=ee09ee2112d03c35f1f7

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+ee09ee2112d03c35f1f7@syzkaller.appspotmail.com

Unable to handle kernel paging request at virtual address ffff800090ffe024
KASAN: probably user-memory-access in range [0x0000000487ff0120-0x0000000487ff0127]
Mem abort info:
  ESR = 0x0000000096000021
  EC = 0x25: DABT (current EL), IL = 32 bits
  SET = 0, FnV = 0
  EA = 0, S1PTW = 0
  FSC = 0x21: alignment fault
Data abort info:
  ISV = 0, ISS = 0x00000021, ISS2 = 0x00000000
  CM = 0, WnR = 0, TnD = 0, TagAccess = 0
  GCS = 0, Overlay = 0, DirtyBit = 0
swapper pgtable: 4k pages, 48-bit VAs, pgdp=00000002193b5000
[ffff800090ffe024] pgd=0000000000000000, p4d=100000021f6f0003, pud=100000021f6f1003, pmd=10000001013e1403, pte=0068000000000713
Internal error: Oops: 0000000096000021 [#1]  SMP
Modules linked in:
CPU: 0 UID: 0 PID: 4957 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT 
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 11/06/2026
pstate: 23400005 (nzCv daif +PAN -UAO +TCO +DIT -SSBS BTYPE=--)
pc : __pi_memcpy_generic+0x98/0x22c arch/arm64/lib/memcpy.S:114
lr : __asan_memcpy+0x68/0x84 mm/kasan/shadow.c:109
sp : ffff800095836fc0
x29: ffff800095836fc0 x28: 1fffe00019f6102f x27: ffff0000cfb08000
x26: dfff800000000000 x25: ffff800095837150 x24: dfff800000000000
x23: ffff0000cfb08190 x22: ffff800082ab0064 x21: ffff800095837370
x20: ffff800090ffe000 x19: 0000000000000044 x18: 1fffe000359cf228
x17: 4a414d0064746d3d x16: 4d45545359534255 x15: ffff800080d0603c
x14: 0000000000000001 x13: 0000000000000009 x12: 0000000000000004
x11: ffff700012b06e76 x10: ffff700012b06e76 x9 : ffffffffffffffff
x8 : ffffffffffffffff x7 : ffffffffffffffff x6 : ffffffff14000400
x5 : ffff8000958373b4 x4 : ffff800090ffe044 x3 : ffff800082ab0064
x2 : 0000000000000044 x1 : ffff800090ffe000 x0 : ffff800095837370
Call trace:
 __pi_memcpy_generic+0x98/0x22c arch/arm64/lib/memcpy.S:113 (P)
 phram_read+0x64/0x98 drivers/mtd/devices/phram.c:70
 mtd_read_oob_std+0x180/0x19c drivers/mtd/mtdcore.c:1702
 mtd_read_oob+0x1fc/0x4ec drivers/mtd/mtdcore.c:1811
 mtd_read+0xd0/0x1a8 drivers/mtd/mtdcore.c:1603
 scan_header drivers/mtd/ftl.c:164 [inline]
 ftl_add_mtd+0x1e8/0x16d8 drivers/mtd/ftl.c:1017
 blktrans_notify_add+0xa8/0xf8 drivers/mtd/mtd_blkdevs.c:469
 add_mtd_device+0xedc/0x1504 drivers/mtd/mtdcore.c:828
 mtd_device_parse_register+0x660/0x7a0 drivers/mtd/mtdcore.c:1156
 register_device+0x340/0x51c drivers/mtd/devices/phram.c:162
 phram_setup+0x354/0x4fc drivers/mtd/devices/phram.c:324
 phram_param_call+0x34/0x90 drivers/mtd/devices/phram.c:349
 param_attr_store+0x144/0x1bc kernel/params.c:591
 module_attr_store+0x6c/0x94 kernel/params.c:906
 sysfs_kf_write+0xe0/0x108 fs/sysfs/file.c:145
 kernfs_fop_write_iter+0x260/0x3b8 fs/kernfs/file.c:345
 new_sync_write fs/read_write.c:595 [inline]
 vfs_write+0x51c/0xa0c fs/read_write.c:687
 ksys_write+0x12c/0x228 fs/read_write.c:739
 __do_sys_write fs/read_write.c:750 [inline]
 __se_sys_write fs/read_write.c:747 [inline]
 __arm64_sys_write+0x84/0x98 fs/read_write.c:747
 __invoke_syscall arch/arm64/kernel/syscall.c:35 [inline]
 invoke_syscall+0x98/0x244 arch/arm64/kernel/syscall.c:49
 el0_svc_common+0xec/0x23c arch/arm64/kernel/syscall.c:121
 do_el0_svc+0x4c/0x5c arch/arm64/kernel/syscall.c:140
 el0_svc+0x60/0x1d8 arch/arm64/kernel/entry-common.c:797
 el0t_64_sync_handler+0x44/0x104 arch/arm64/kernel/entry-common.c:816
 el0t_64_sync+0x198/0x19c arch/arm64/kernel/entry.S:606
Code: d65f03c0 d503201f a9401c26 a9412428 (a97e2c8a) 
---[ end trace 0000000000000000 ]---
----------------
Code disassembly (best guess):
   0:	d65f03c0 	ret
   4:	d503201f 	nop
   8:	a9401c26 	ldp	x6, x7, [x1]
   c:	a9412428 	ldp	x8, x9, [x1, #16]
* 10:	a97e2c8a 	ldp	x10, x11, [x4, #-32] <-- trapping instruction


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
For information about AI bisection see: https://goo.gl/tpsmEJ#ai-bisection

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

^ permalink raw reply	[flat|nested] 9+ messages in thread

* [PATCH] mtd: Use I/O memory accessors when in an uncached
  2026-11-18  1:49 ` syzbot
@ 2026-11-20  4:37   ` Edward Adam Davis
  -1 siblings, 0 replies; 9+ messages in thread
From: Edward Adam Davis @ 2026-11-20  4:37 UTC (permalink / raw)
  To: syzbot+ee09ee2112d03c35f1f7
  Cc: joern, linux-kernel, linux-mtd, miquel.raynal, richard,
	syzkaller-bugs, vigneshr

When registering a phram device via module parameter or sysfs invokes
register_device() with pdev = NULL, which causes new->cached to be false.
As a result, phram_map() maps the physical memory region using ioremap()
(mapped as PROT_DEVICE_nGnRE on arm64) and forces the pointer into mtd->
priv, yet phram_read(), phram_write(), and phram_erase() access it using
standard memcpy() and memset() instead of I/O memory accessors (like
memcpy_fromio()). When scanning partition headers (e.g., ftl_add_mtd()
reading a 68-byte header via mtd_read()), arm64's memcpy implementation
issues pair-load (ldp) instructions at unaligned offsets, which triggers
an alignment fault (FSC = 0x21) on Device memory and crashes the kernel.

In phram_read/write/erase(), standard memory accessors are used when in
a cached; otherwise, I/O memory accessors are used.

Fixes: f3d45ac21e46 ("mtd: phram: Module parameters add writable permissions")
Fixes: 285133040e6c ("arm64: Import latest memcpy()/memmove() implementation")
Reported-by: syzbot+ee09ee2112d03c35f1f7@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ee09ee2112d03c35f1f7
Signed-off-by: Edward Adam Davis <eadavis@sina.com>
---
 drivers/mtd/devices/phram.c | 21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

diff --git a/drivers/mtd/devices/phram.c b/drivers/mtd/devices/phram.c
index b42cadcd76b1..95bd5a9d7e48 100644
--- a/drivers/mtd/devices/phram.c
+++ b/drivers/mtd/devices/phram.c
@@ -43,8 +43,13 @@ static LIST_HEAD(phram_list);
 static int phram_erase(struct mtd_info *mtd, struct erase_info *instr)
 {
 	u_char *start = mtd->priv;
+	struct device_node *np = mtd_get_of_node(mtd);
+	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
 
-	memset(start + instr->addr, 0xff, instr->len);
+	if (cached)
+		memset(start + instr->addr, 0xff, instr->len);
+	else
+		memset_io(start + instr->addr, 0xff, instr->len);
 
 	return 0;
 }
@@ -66,8 +71,13 @@ static int phram_read(struct mtd_info *mtd, loff_t from, size_t len,
 		size_t *retlen, u_char *buf)
 {
 	u_char *start = mtd->priv;
+	struct device_node *np = mtd_get_of_node(mtd);
+	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
 
-	memcpy(buf, start + from, len);
+	if (cached)
+		memcpy(buf, start + from, len);
+	else
+		memcpy_fromio(buf, start + from, len);
 	*retlen = len;
 	return 0;
 }
@@ -76,8 +86,13 @@ static int phram_write(struct mtd_info *mtd, loff_t to, size_t len,
 		size_t *retlen, const u_char *buf)
 {
 	u_char *start = mtd->priv;
+	struct device_node *np = mtd_get_of_node(mtd);
+	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
 
-	memcpy(start + to, buf, len);
+	if (cached)
+		memcpy(start + to, buf, len);
+	else
+		memcpy_fromio(start + to, buf, len);
 	*retlen = len;
 	return 0;
 }
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 9+ messages in thread

* [PATCH] mtd: Use I/O memory accessors when in an uncached
@ 2026-11-20  4:37   ` Edward Adam Davis
  0 siblings, 0 replies; 9+ messages in thread
From: Edward Adam Davis @ 2026-11-20  4:37 UTC (permalink / raw)
  To: syzbot+ee09ee2112d03c35f1f7
  Cc: joern, linux-kernel, linux-mtd, miquel.raynal, richard,
	syzkaller-bugs, vigneshr

When registering a phram device via module parameter or sysfs invokes
register_device() with pdev = NULL, which causes new->cached to be false.
As a result, phram_map() maps the physical memory region using ioremap()
(mapped as PROT_DEVICE_nGnRE on arm64) and forces the pointer into mtd->
priv, yet phram_read(), phram_write(), and phram_erase() access it using
standard memcpy() and memset() instead of I/O memory accessors (like
memcpy_fromio()). When scanning partition headers (e.g., ftl_add_mtd()
reading a 68-byte header via mtd_read()), arm64's memcpy implementation
issues pair-load (ldp) instructions at unaligned offsets, which triggers
an alignment fault (FSC = 0x21) on Device memory and crashes the kernel.

In phram_read/write/erase(), standard memory accessors are used when in
a cached; otherwise, I/O memory accessors are used.

Fixes: f3d45ac21e46 ("mtd: phram: Module parameters add writable permissions")
Fixes: 285133040e6c ("arm64: Import latest memcpy()/memmove() implementation")
Reported-by: syzbot+ee09ee2112d03c35f1f7@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ee09ee2112d03c35f1f7
Signed-off-by: Edward Adam Davis <eadavis@sina.com>
---
 drivers/mtd/devices/phram.c | 21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

diff --git a/drivers/mtd/devices/phram.c b/drivers/mtd/devices/phram.c
index b42cadcd76b1..95bd5a9d7e48 100644
--- a/drivers/mtd/devices/phram.c
+++ b/drivers/mtd/devices/phram.c
@@ -43,8 +43,13 @@ static LIST_HEAD(phram_list);
 static int phram_erase(struct mtd_info *mtd, struct erase_info *instr)
 {
 	u_char *start = mtd->priv;
+	struct device_node *np = mtd_get_of_node(mtd);
+	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
 
-	memset(start + instr->addr, 0xff, instr->len);
+	if (cached)
+		memset(start + instr->addr, 0xff, instr->len);
+	else
+		memset_io(start + instr->addr, 0xff, instr->len);
 
 	return 0;
 }
@@ -66,8 +71,13 @@ static int phram_read(struct mtd_info *mtd, loff_t from, size_t len,
 		size_t *retlen, u_char *buf)
 {
 	u_char *start = mtd->priv;
+	struct device_node *np = mtd_get_of_node(mtd);
+	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
 
-	memcpy(buf, start + from, len);
+	if (cached)
+		memcpy(buf, start + from, len);
+	else
+		memcpy_fromio(buf, start + from, len);
 	*retlen = len;
 	return 0;
 }
@@ -76,8 +86,13 @@ static int phram_write(struct mtd_info *mtd, loff_t to, size_t len,
 		size_t *retlen, const u_char *buf)
 {
 	u_char *start = mtd->priv;
+	struct device_node *np = mtd_get_of_node(mtd);
+	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
 
-	memcpy(start + to, buf, len);
+	if (cached)
+		memcpy(start + to, buf, len);
+	else
+		memcpy_fromio(start + to, buf, len);
 	*retlen = len;
 	return 0;
 }
-- 
2.43.0


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

^ permalink raw reply related	[flat|nested] 9+ messages in thread

* Re: [PATCH] mtd: Use I/O memory accessors when in an uncached
  2026-11-20  4:37   ` Edward Adam Davis
  (?)
@ 2026-11-20  4:47   ` sashiko-bot
  2026-11-20  5:24     ` [PATCH v2] " Edward Adam Davis
  -1 siblings, 1 reply; 9+ messages in thread
From: sashiko-bot @ 2026-11-20  4:47 UTC (permalink / raw)
  To: Edward Adam Davis, syzbot+ee09ee2112d03c35f1f7
  Cc: joern, linux-mtd, miquel.raynal, richard, syzkaller-bugs,
	vigneshr

Thank you for your contribution! Sashiko AI review found 3 potential
issues to consider:
- [High] mtd: phram: wrong I/O accessor memcpy_fromio() used in
  phram_write()
- [Low] mtd: phram: missing __iomem casts in phram_erase(),
  phram_read(), phram_write()
- [Low] mtd: phram: redundant DT lookup on every I/O instead of using
  phram->cached
--

commit 652c886e91fa425e4122f1ebcb9ef0f2b874fd28
Author: Edward Adam Davis <eadavis@sina.com>

mtd: Use I/O memory accessors when in an uncached

This commit uses I/O memory accessors instead of standard memset() and
memcpy() in phram_erase(), phram_read(), and phram_write() when the phram
mapping is uncached.

> diff --git a/drivers/mtd/devices/phram.c b/drivers/mtd/devices/phram.c
> index b42cadcd76b11..95bd5a9d7e482 100644
> --- a/drivers/mtd/devices/phram.c
> +++ b/drivers/mtd/devices/phram.c
> @@ -43,8 +43,13 @@ static LIST_HEAD(phram_list);
>  static int phram_erase(struct mtd_info *mtd, struct erase_info *instr)
>  {
>  	u_char *start = mtd->priv;
> +	struct device_node *np = mtd_get_of_node(mtd);
> +	bool cached = np ? !of_property_read_bool(np, "no-map") : false;

[Severity: Low]
Can phram_erase(), phram_read(), and phram_write() read
container_of(mtd, struct phram_mtd_list, mtd)->cached instead of calling
of_property_read_bool() on every I/O?  register_device() already computes
this flag once and stores it in new->cached.

> -	memset(start + instr->addr, 0xff, instr->len);
> +	if (cached)
> +		memset(start + instr->addr, 0xff, instr->len);
> +	else
> +		memset_io(start + instr->addr, 0xff, instr->len);

[Severity: Low]
Since start is u_char *, passing start + offset to memset_io(),
memcpy_fromio(), and memcpy_toio() without a (void __iomem *) cast
triggers sparse address-space warnings across phram_erase(), phram_read(),
and phram_write().

>  
>  	return 0;
>  }
> @@ -66,8 +71,13 @@ static int phram_read(struct mtd_info *mtd, loff_t from, size_t len,
>  		size_t *retlen, u_char *buf)
>  {
>  	u_char *start = mtd->priv;
> +	struct device_node *np = mtd_get_of_node(mtd);
> +	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
>  
> -	memcpy(buf, start + from, len);
> +	if (cached)
> +		memcpy(buf, start + from, len);
> +	else
> +		memcpy_fromio(buf, start + from, len);
>  	*retlen = len;
>  	return 0;
>  }
> @@ -76,8 +86,13 @@ static int phram_write(struct mtd_info *mtd, loff_t to, size_t len,
>  		size_t *retlen, const u_char *buf)
>  {
>  	u_char *start = mtd->priv;
> +	struct device_node *np = mtd_get_of_node(mtd);
> +	bool cached = np ? !of_property_read_bool(np, "no-map") : false;
>  
> -	memcpy(start + to, buf, len);
> +	if (cached)
> +		memcpy(start + to, buf, len);
> +	else
> +		memcpy_fromio(start + to, buf, len);

[Severity: High]
Should phram_write() call memcpy_toio() instead of memcpy_fromio()?

memcpy_fromio(dst, src, count) treats src (buf) as __iomem and dst
(start + to) as normal RAM: it aligns src rather than dst and writes to
dst via put_unaligned().  When start + to is not aligned to sizeof(long),
writing to an uncached phram device issues unaligned word stores to
Device memory and triggers an alignment fault on arm64.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261120043709.99845-1-eadavis@sina.com?part=1

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

^ permalink raw reply	[flat|nested] 9+ messages in thread

* [PATCH v2] mtd: Use I/O memory accessors when in an uncached
  2026-11-20  4:47   ` sashiko-bot
@ 2026-11-20  5:24     ` Edward Adam Davis
  2026-11-20  5:35       ` sashiko-bot
  0 siblings, 1 reply; 9+ messages in thread
From: Edward Adam Davis @ 2026-11-20  5:24 UTC (permalink / raw)
  To: sashiko-bot
  Cc: eadavis, joern, linux-mtd, miquel.raynal, richard,
	sashiko-reviews, syzbot+ee09ee2112d03c35f1f7, syzkaller-bugs,
	vigneshr

When registering a phram device via module parameter or sysfs invokes
register_device() with pdev = NULL, which causes new->cached to be false.
As a result, phram_map() maps the physical memory region using ioremap()
(mapped as PROT_DEVICE_nGnRE on arm64) and forces the pointer into mtd->
priv, yet phram_read(), phram_write(), and phram_erase() access it using
standard memcpy() and memset() instead of I/O memory accessors (like
memcpy_fromio()). When scanning partition headers (e.g., ftl_add_mtd()
reading a 68-byte header via mtd_read()), arm64's memcpy implementation
issues pair-load (ldp) instructions at unaligned offsets, which triggers
an alignment fault (FSC = 0x21) on Device memory and crashes the kernel.

In phram_read/write/erase(), standard memory accessors are used when in
a cached; otherwise, I/O memory accessors are used.

Fixes: f3d45ac21e46 ("mtd: phram: Module parameters add writable permissions")
Fixes: 285133040e6c ("arm64: Import latest memcpy()/memmove() implementation")
Reported-by: syzbot+ee09ee2112d03c35f1f7@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ee09ee2112d03c35f1f7
Signed-off-by: Edward Adam Davis <eadavis@sina.com>
---
v1 -> v2: use phram_mtd_list cached, add void __iomem* and write use toio

 drivers/mtd/devices/phram.c | 22 +++++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/drivers/mtd/devices/phram.c b/drivers/mtd/devices/phram.c
index b42cadcd76b1..802da960d840 100644
--- a/drivers/mtd/devices/phram.c
+++ b/drivers/mtd/devices/phram.c
@@ -43,8 +43,14 @@ static LIST_HEAD(phram_list);
 static int phram_erase(struct mtd_info *mtd, struct erase_info *instr)
 {
 	u_char *start = mtd->priv;
+	struct phram_mtd_list *phram =
+		container_of(mtd, struct phram_mtd_list, mtd);
 
-	memset(start + instr->addr, 0xff, instr->len);
+	if (phram->cached)
+		memset(start + instr->addr, 0xff, instr->len);
+	else
+		memset_io((void __iomem *)start + instr->addr, 0xff,
+			  instr->len);
 
 	return 0;
 }
@@ -66,8 +72,13 @@ static int phram_read(struct mtd_info *mtd, loff_t from, size_t len,
 		size_t *retlen, u_char *buf)
 {
 	u_char *start = mtd->priv;
+	struct phram_mtd_list *phram =
+		container_of(mtd, struct phram_mtd_list, mtd);
 
-	memcpy(buf, start + from, len);
+	if (phram->cached)
+		memcpy(buf, start + from, len);
+	else
+		memcpy_fromio(buf, (void __iomem *)start + from, len);
 	*retlen = len;
 	return 0;
 }
@@ -76,8 +87,13 @@ static int phram_write(struct mtd_info *mtd, loff_t to, size_t len,
 		size_t *retlen, const u_char *buf)
 {
 	u_char *start = mtd->priv;
+	struct phram_mtd_list *phram =
+		container_of(mtd, struct phram_mtd_list, mtd);
 
-	memcpy(start + to, buf, len);
+	if (phram->cached)
+		memcpy(start + to, buf, len);
+	else
+		memcpy_toio((void __iomem *)start + to, buf, len);
 	*retlen = len;
 	return 0;
 }
-- 
2.43.0


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

^ permalink raw reply related	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] mtd: Use I/O memory accessors when in an uncached
  2026-11-20  5:24     ` [PATCH v2] " Edward Adam Davis
@ 2026-11-20  5:35       ` sashiko-bot
  2026-11-20  8:25         ` Miquel Raynal
  0 siblings, 1 reply; 9+ messages in thread
From: sashiko-bot @ 2026-11-20  5:35 UTC (permalink / raw)
  To: Edward Adam Davis
  Cc: joern, linux-mtd, miquel.raynal, richard,
	syzbot+ee09ee2112d03c35f1f7, syzkaller-bugs, vigneshr

Thank you for your contribution! Sashiko AI review found 2 potential
issues to consider:
- [High] mtd: phram: uncached ioremap() pointer exposed via
  phram_point()
- [Low] mtd: phram: inaccurate Fixes trailer and truncated commit
  message
--

commit 54b14d07ef014deae72f6fbc65c025f6ac913af6
Author: Edward Adam Davis <eadavis@sina.com>

> mtd: Use I/O memory accessors when in an uncached
>
> When registering a phram device via module parameter or sysfs invokes
> register_device() with pdev = NULL, which causes new->cached to be false.
> As a result, phram_map() maps the physical memory region using ioremap()
> (mapped as PROT_DEVICE_nGnRE on arm64) and forces the pointer into mtd->
> priv, yet phram_read(), phram_write(), and phram_erase() access it using
> standard memcpy() and memset() instead of I/O memory accessors (like
> memcpy_fromio()). When scanning partition headers (e.g., ftl_add_mtd()
> reading a 68-byte header via mtd_read()), arm64's memcpy implementation
> issues pair-load (ldp) instructions at unaligned offsets, which triggers
> an alignment fault (FSC = 0x21) on Device memory and crashes the kernel.
>
> In phram_read/write/erase(), standard memory accessors are used when in
> a cached; otherwise, I/O memory accessors are used.
>
> Fixes: f3d45ac21e46 ("mtd: phram: Module parameters add writable permissions")
> Fixes: 285133040e6c ("arm64: Import latest memcpy()/memmove() implementation")
> Reported-by: syzbot+ee09ee2112d03c35f1f7@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/bug?extid=ee09ee2112d03c35f1f7
> Signed-off-by: Edward Adam Davis <eadavis@sina.com>

[Severity: Low]
Are the Fixes tags accurate here?  Calling memcpy() on an __iomem pointer
from ioremap() is a phram bug rather than a bug in 285133040e6c ("arm64:
Import latest memcpy()/memmove() implementation"), and this patch branches on
phram->cached, which was added in 9401911f2d9f ("mtd: phram: Allow cached
mappings").

Also, the subject ("mtd: Use I/O memory accessors when in an uncached") and
body ("when in a cached") are missing the "mtd: phram:" prefix and the
trailing noun (such as "mapping").

> diff --git a/drivers/mtd/devices/phram.c b/drivers/mtd/devices/phram.c
> index b42cadcd76b11..802da960d840a 100644
> --- a/drivers/mtd/devices/phram.c
> +++ b/drivers/mtd/devices/phram.c
> @@ -43,8 +43,14 @@ static LIST_HEAD(phram_list);
>  static int phram_erase(struct mtd_info *mtd, struct erase_info *instr)
>  {
>  	u_char *start = mtd->priv;
> +	struct phram_mtd_list *phram =
> +		container_of(mtd, struct phram_mtd_list, mtd);
>  
> -	memset(start + instr->addr, 0xff, instr->len);
> +	if (phram->cached)
> +		memset(start + instr->addr, 0xff, instr->len);
> +	else
> +		memset_io((void __iomem *)start + instr->addr, 0xff,
> +			  instr->len);
>  
>  	return 0;
>  }
> @@ -66,8 +72,13 @@ static int phram_read(struct mtd_info *mtd, loff_t from, size_t len,
>  		size_t *retlen, u_char *buf)
>  {
>  	u_char *start = mtd->priv;
> +	struct phram_mtd_list *phram =
> +		container_of(mtd, struct phram_mtd_list, mtd);
>  
> -	memcpy(buf, start + from, len);
> +	if (phram->cached)
> +		memcpy(buf, start + from, len);
> +	else
> +		memcpy_fromio(buf, (void __iomem *)start + from, len);
>  	*retlen = len;
>  	return 0;
>  }

[Severity: High]
Does phram_point() still expose the uncached ioremap() pointer in mtd->priv
to callers that use normal memcpy() on it?

Between phram_erase() and phram_read(), phram_point() remains unchanged and
is unconditionally wired up in register_device():

drivers/mtd/devices/phram.c:phram_point() {
	*virt = mtd->priv + from;
	*retlen = len;
	return 0;
}

Callers of mtd_point() expect normal direct-accessible memory and bypass
mtd_read().  For example, mounting cramfs or jffs2 on an uncached phram
device succeeds in mtd_point() and runs memcpy() directly on the returned
pointer (such as the 76-byte superblock read in cramfs_read_super()),
triggering the same alignment fault on arm64 Device memory.  Should
_point and _unpoint only be registered (or phram_point() return
-EOPNOTSUPP) when phram->cached is false?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261120052434.101613-1-eadavis@sina.com?part=1

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] mtd: Use I/O memory accessors when in an uncached
  2026-11-20  5:35       ` sashiko-bot
@ 2026-11-20  8:25         ` Miquel Raynal
  2026-11-20  9:44           ` Edward Adam Davis
  0 siblings, 1 reply; 9+ messages in thread
From: Miquel Raynal @ 2026-11-20  8:25 UTC (permalink / raw)
  To: sashiko-bot
  Cc: Edward Adam Davis, sashiko-reviews, joern, linux-mtd, richard,
	syzbot+ee09ee2112d03c35f1f7, syzkaller-bugs, vigneshr

Hi Edward, 

On 20/11/2026 at 05:35:32 GMT, sashiko-bot@kernel.org wrote:

> Thank you for your contribution! Sashiko AI review found 2 potential
> issues to consider:
> - [High] mtd: phram: uncached ioremap() pointer exposed via
>   phram_point()
> - [Low] mtd: phram: inaccurate Fixes trailer and truncated commit
>   message

These issues look plausible, I'll let you answer them.

Thanks,
Miquèl

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

^ permalink raw reply	[flat|nested] 9+ messages in thread

* [PATCH v2] mtd: Use I/O memory accessors when in an uncached
  2026-11-20  8:25         ` Miquel Raynal
@ 2026-11-20  9:44           ` Edward Adam Davis
  0 siblings, 0 replies; 9+ messages in thread
From: Edward Adam Davis @ 2026-11-20  9:44 UTC (permalink / raw)
  To: miquel.raynal
  Cc: eadavis, joern, linux-mtd, richard, sashiko-bot, sashiko-reviews,
	syzbot+ee09ee2112d03c35f1f7, syzkaller-bugs, vigneshr

On Fri, 20 Nov 2026 09:25:02 +0100, Miquel Raynal wrote:
> > Thank you for your contribution! Sashiko AI review found 2 potential
> > issues to consider:
> > - [High] mtd: phram: uncached ioremap() pointer exposed via
> >   phram_point()
One point needs correction:
_point and _unpoint should only be registered when phram->cached is true,
because the I/O map is used exclusively when the mapping is uncached.
> > - [Low] mtd: phram: inaccurate Fixes trailer and truncated commit
> >   message
I'll update the subject in the v3;

285133040e6c ("arm64: Import latest memcpy()/memmove() implementation")
In ARM64's __pi_memcpy_generic (arch/arm64/lib/memcpy.S), copies between
33 and 128 bytes branch to L(copy32_128):

L(copy32_128):
    ldp A_l, A_h, [src]
    ldp B_l, B_h, [src, 16]
    ldp C_l, C_h, [srcend, -32]

Here:

    src = 0xffff80009594b000
    count = 0x44
    srcend = src + count = 0xffff80009594b044
    srcend - 32 = 0xffff80009594b024

The instruction ldp x10, x11, [x4, #-32] attempts a 16-byte pair load
from address 0xffff80009594b024, which is only 4-byte aligned (offset
0x24). Because the address points into Device-nGnRE memory, the MMU
generates a Data Abort alignment fault (ESR = 0x96000021, FSC = 0x21),
leading to the reported kernel Oops.

It was introduced by commit 285133040e6c, so I add the Fixes tag for
285133040e6c.
> 
> These issues look plausible, I'll let you answer them.

BR,
Edward

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2026-11-20  9:45 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-18  1:49 [syzbot] [mtd?] BUG: unable to handle kernel paging request in phram_read syzbot
2026-11-18  1:49 ` syzbot
2026-11-20  4:37 ` [PATCH] mtd: Use I/O memory accessors when in an uncached Edward Adam Davis
2026-11-20  4:37   ` Edward Adam Davis
2026-11-20  4:47   ` sashiko-bot
2026-11-20  5:24     ` [PATCH v2] " Edward Adam Davis
2026-11-20  5:35       ` sashiko-bot
2026-11-20  8:25         ` Miquel Raynal
2026-11-20  9:44           ` Edward Adam Davis

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.