Since some time now, printk() adds carriage return, leading to
unusable xmon output:
[ 54.288722] sysrq: Entering xmon
[ 54.292209] Vector: 0 at [cace3d2c]
[ 54.292274] pc:
[ 54.292331] c0023650
[ 54.292468] : xmon+0x28/0x58
[ 54.292519]
[ 54.292574] lr:
[ 54.292630] c0023724
[ 54.292749] : sysrq_handle_xmon+0xa4/0xfc
[ 54.292801]
[ 54.292867] sp: cace3de8
[ 54.292931] msr: 9032
[ 54.292999] current = 0xc28d0000
[ 54.293072] pid = 377, comm = sh
[ 54.293157] Linux version 5.10.0-rc6-s3k-dev-01364-gedf13f0ccd76-dirty ([email protected]) (powerpc64-linux-gcc (GCC) 10.1.0, GNU ld (GNU Binutils) 2.34) #4211 PREEMPT Fri Dec 4 09:32:11 UTC 2020
[ 54.293287] enter ? for help
[ 54.293470] [cace3de8]
[ 54.293532] c0023724
[ 54.293654] sysrq_handle_xmon+0xa4/0xfc
[ 54.293711] (unreliable)
[ 54.293859] [cace3e18]
[ 54.293918] c03885a8
[ 54.294056] __handle_sysrq+0xe4/0x1d4
[ 54.294108]
[ 54.294255] [cace3e48]
[ 54.294314] c0388704
[ 54.294441] write_sysrq_trigger+0x34/0x74
[ 54.294493]
[ 54.294641] [cace3e68]
[ 54.294700] c01f65d0
[ 54.294822] proc_reg_write+0xac/0x11c
[ 54.294875]
[ 54.295023] [cace3e88]
[ 54.295082] c0179910
[ 54.295198] vfs_write+0x134/0x46c
[ 54.295250]
[ 54.295396] [cace3f08]
[ 54.295455] c0179de8
[ 54.295567] ksys_write+0x78/0x11c
[ 54.295619]
[ 54.295766] [cace3f38]
[ 54.295825] c00110d0
[ 54.295951] ret_from_syscall+0x0/0x34
[ 54.296002]
[ 54.296159] --- Exception: c01 (System Call) at
[ 54.296217] 0fd4e784
[ 54.296303]
[ 54.296375] SP (7fca6ff0) is in userspace
[ 54.296431] mon>
[ 54.296484] <no input ...>
Use pr_cont() instead.
Signed-off-by: Christophe Leroy <[email protected]>
Fixes: 4bcc595ccd80 ("printk: reinstate KERN_CONT for printing continuation lines")
Cc: [email protected]
---
arch/powerpc/xmon/nonstdio.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/powerpc/xmon/nonstdio.c b/arch/powerpc/xmon/nonstdio.c
index 5c1a50912229..9b0d85bff021 100644
--- a/arch/powerpc/xmon/nonstdio.c
+++ b/arch/powerpc/xmon/nonstdio.c
@@ -178,7 +178,7 @@ void xmon_printf(const char *format, ...)
if (n && rc == 0) {
/* No udbg hooks, fallback to printk() - dangerous */
- printk("%s", xmon_outbuf);
+ pr_cont("%s", xmon_outbuf);
}
}
--
2.25.0
Christophe Leroy <[email protected]> writes:
> Since some time now, printk() adds carriage return, leading to
> unusable xmon output:
>
> [ 54.288722] sysrq: Entering xmon
> [ 54.292209] Vector: 0 at [cace3d2c]
> [ 54.292274] pc:
> [ 54.292331] c0023650
...
> diff --git a/arch/powerpc/xmon/nonstdio.c b/arch/powerpc/xmon/nonstdio.c
> index 5c1a50912229..9b0d85bff021 100644
> --- a/arch/powerpc/xmon/nonstdio.c
> +++ b/arch/powerpc/xmon/nonstdio.c
> @@ -178,7 +178,7 @@ void xmon_printf(const char *format, ...)
>
> if (n && rc == 0) {
> /* No udbg hooks, fallback to printk() - dangerous */
> - printk("%s", xmon_outbuf);
> + pr_cont("%s", xmon_outbuf);
> }
Ah OK, in the case where there's no udbg backend. We basically always
have a udbg backend on 64-bit, via hvc console. Which explains why we
haven't noticed it.
Will pick up the patch.
cheers
On Fri, 2020-12-04 at 21:56 +1100, Michael Ellerman wrote:
> Christophe Leroy <[email protected]> writes:
> > Since some time now, printk() adds carriage return, leading to
> > unusable xmon output:
> >
> > [ 54.288722] sysrq: Entering xmon
> > [ 54.292209] Vector: 0 at [cace3d2c]
> > [ 54.292274] pc:
> > [ 54.292331] c0023650
>
> ...
>
> > diff --git a/arch/powerpc/xmon/nonstdio.c b/arch/powerpc/xmon/nonstdio.c
> > index 5c1a50912229..9b0d85bff021 100644
> > --- a/arch/powerpc/xmon/nonstdio.c
> > +++ b/arch/powerpc/xmon/nonstdio.c
> > @@ -178,7 +178,7 @@ void xmon_printf(const char *format, ...)
> > ?
> >
> > ? if (n && rc == 0) {
> > ? /* No udbg hooks, fallback to printk() - dangerous */
> > - printk("%s", xmon_outbuf);
> > + pr_cont("%s", xmon_outbuf);
> > ? }
>
> Ah OK, in the case where there's no udbg backend. We basically always
> have a udbg backend on 64-bit, via hvc console. Which explains why we
> haven't noticed it.
>
> Will pick up the patch.
>
> cheers
Perhaps all of these bare printks should be inspected for defects:
$ git grep -P -n '\bprintk\s*\(\s*(?!KERN_\w+)"[^\\n]*"' arch/powerpc
arch/powerpc/kernel/process.c:1475: printk("NIP: "REG" LR: "REG" CTR: "REG"\n",
arch/powerpc/kernel/process.c:1479: printk("MSR: "REG" ", regs->msr);
arch/powerpc/kernel/process.c:1513: printk("NIP ["REG"] %pS\n", regs->nip, (void *)regs->nip);
arch/powerpc/kernel/process.c:1514: printk("LR ["REG"] %pS\n", regs->link, (void *)regs->link);
arch/powerpc/kernel/process.c:2157: printk("%s["REG"] ["REG"] %pS",
arch/powerpc/kernel/traps.c:621: printk("Caused by (from MCSR=%lx): ", reason);
arch/powerpc/kernel/traps.c:726: printk("Caused by (from MCSR=%lx): ", reason);
arch/powerpc/kernel/traps.c:766: printk("Caused by (from MCSR=%lx): ", reason);
arch/powerpc/kernel/traps.c:791: printk("Caused by (from SRR1=%lx): ", reason);
arch/powerpc/kernel/udbg.c:95: printk("%s", s);
arch/powerpc/math-emu/fabs.c:13: printk("%s: D %p, B %p: ", __func__, frD, frB);
arch/powerpc/math-emu/fctiw.c:22: printk("%s: D %p, B %p: ", __func__, frD, frB);
arch/powerpc/math-emu/fctiwz.c:29: printk("%s: D %p, B %p: ", __func__, frD, frB);
arch/powerpc/math-emu/fmr.c:13: printk("%s: D %p, B %p: ", __func__, frD, frB);
arch/powerpc/math-emu/fnabs.c:13: printk("%s: D %p, B %p: ", __func__, frD, frB);
arch/powerpc/math-emu/fneg.c:13: printk("%s: D %p, B %p: ", __func__, frD, frB);
arch/powerpc/math-emu/lfd.c:15: printk("%s: D %p, ea %p: ", __func__, frD, ea);
arch/powerpc/math-emu/stfd.c:11: printk("%s: S %p, ea %p: ", __func__, frS, ea);
arch/powerpc/mm/nohash/44x.c:192: printk("%d ", i);
arch/powerpc/platforms/4xx/machine_check.c:19: printk("Data");
arch/powerpc/platforms/chrp/pci.c:256: printk(" at %llx", (unsigned long long)r.start);
arch/powerpc/platforms/embedded6xx/ls_uart.c:47: printk("%c", in_8(avr_addr + UART_RX));
arch/powerpc/platforms/powermac/pfunc_core.c:83: printk("%s", title);
arch/powerpc/platforms/powermac/pfunc_core.c:85: printk("%02x ", *((u8 *)blob));
arch/powerpc/platforms/powernv/pci-ioda.c:81: printk("%spci %s: [PE# %.2x] %pV",
arch/powerpc/sysdev/tsi108_pci.c:63: printk("PCI CFG write : ");
arch/powerpc/sysdev/tsi108_pci.c:64: printk("%d:0x%x:0x%x ", bus->number, devfunc, offset);
arch/powerpc/sysdev/tsi108_pci.c:65: printk("%d ADDR=0x%08x ", len, (uint) cfg_addr);
arch/powerpc/sysdev/tsi108_pci.c:164: printk("PCI CFG read : ");
arch/powerpc/sysdev/tsi108_pci.c:165: printk("%d:0x%x:0x%x ", bus->number, devfn, offset);
arch/powerpc/sysdev/tsi108_pci.c:166: printk("%d ADDR=0x%08x ", len, (uint) cfg_addr);
arch/powerpc/sysdev/tsi108_pci.c:315: printk("cfg_ctl=0x%08x ", temp);
arch/powerpc/xmon/nonstdio.c:181: printk("%s", xmon_outbuf);
On Fri, 4 Dec 2020 10:35:38 +0000 (UTC), Christophe Leroy wrote:
> Since some time now, printk() adds carriage return, leading to
> unusable xmon output:
>
> [ 54.288722] sysrq: Entering xmon
> [ 54.292209] Vector: 0 at [cace3d2c]
> [ 54.292274] pc:
> [ 54.292331] c0023650
> [ 54.292468] : xmon+0x28/0x58
> [ 54.292519]
> [ 54.292574] lr:
> [ 54.292630] c0023724
> [ 54.292749] : sysrq_handle_xmon+0xa4/0xfc
> [ 54.292801]
> [ 54.292867] sp: cace3de8
> [ 54.292931] msr: 9032
> [ 54.292999] current = 0xc28d0000
> [ 54.293072] pid = 377, comm = sh
> [ 54.293157] Linux version 5.10.0-rc6-s3k-dev-01364-gedf13f0ccd76-dirty ([email protected]) (powerpc64-linux-gcc (GCC) 10.1.0, GNU ld (GNU Binutils) 2.34) #4211 PREEMPT Fri Dec 4 09:32:11 UTC 2020
> [ 54.293287] enter ? for help
> [ 54.293470] [cace3de8]
> [ 54.293532] c0023724
> [ 54.293654] sysrq_handle_xmon+0xa4/0xfc
> [ 54.293711] (unreliable)
> [ 54.293859] [cace3e18]
> [ 54.293918] c03885a8
> [ 54.294056] __handle_sysrq+0xe4/0x1d4
> [ 54.294108]
> [ 54.294255] [cace3e48]
> [ 54.294314] c0388704
> [ 54.294441] write_sysrq_trigger+0x34/0x74
> [ 54.294493]
> [ 54.294641] [cace3e68]
> [ 54.294700] c01f65d0
> [ 54.294822] proc_reg_write+0xac/0x11c
> [ 54.294875]
> [ 54.295023] [cace3e88]
> [ 54.295082] c0179910
> [ 54.295198] vfs_write+0x134/0x46c
> [ 54.295250]
> [ 54.295396] [cace3f08]
> [ 54.295455] c0179de8
> [ 54.295567] ksys_write+0x78/0x11c
> [ 54.295619]
> [ 54.295766] [cace3f38]
> [ 54.295825] c00110d0
> [ 54.295951] ret_from_syscall+0x0/0x34
> [ 54.296002]
> [ 54.296159] --- Exception: c01 (System Call) at
> [ 54.296217] 0fd4e784
> [ 54.296303]
> [ 54.296375] SP (7fca6ff0) is in userspace
> [ 54.296431] mon>
> [ 54.296484] <no input ...>
>
> [...]
Applied to powerpc/next.
[1/1] powerpc/xmon: Change printk() to pr_cont()
https://git.kernel.org/powerpc/c/7c6c86b36a36dd4a13d30bba07718e767aa2e7a1
cheers