| Message ID | 20250213181610.718343-1-colin.i.king@gmail.com (mailing list archive) |
|---|---|
| State | Not Applicable, archived |
| Headers | show |
| Series | [next] ACPI: OSL: ratelimit ACPICA kernel messages | expand |
+Saket Dumbre On Thu, Feb 13, 2025 at 7:16 PM Colin Ian King <colin.i.king@gmail.com> wrote: > > In cases where the ACPI AML contains errors there can be quite a large > amount of ACPICA kernel log spamming. Except when someone wants to see them all which also happens. And wouldn't this also rate limit debug messages from ACPICA specifically enabled via the kernel command line? If so, I'd rather find a way to tell ACPICA to be less verbose. > Reduce this by rate limiting the messages. > > Signed-off-by: Colin Ian King <colin.i.king@gmail.com> > --- > drivers/acpi/osl.c | 20 ++++++++++++-------- > 1 file changed, 12 insertions(+), 8 deletions(-) > > diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c > index 5ff343096ece..d4ece68e0fd6 100644 > --- a/drivers/acpi/osl.c > +++ b/drivers/acpi/osl.c > @@ -159,17 +159,21 @@ void __printf(1, 0) acpi_os_vprintf(const char *fmt, va_list args) > if (acpi_in_debugger) { > kdb_printf("%s", buffer); > } else { > - if (printk_get_level(buffer)) > - printk("%s", buffer); > - else > - printk(KERN_CONT "%s", buffer); > + if (printk_ratelimit()) { > + if (printk_get_level(buffer)) > + printk("%s", buffer); > + else > + printk(KERN_CONT "%s", buffer); > + } > } > #else > if (acpi_debugger_write_log(buffer) < 0) { > - if (printk_get_level(buffer)) > - printk("%s", buffer); > - else > - printk(KERN_CONT "%s", buffer); > + if (printk_ratelimit()) { > + if (printk_get_level(buffer)) > + printk("%s", buffer); > + else > + printk(KERN_CONT "%s", buffer); > + } > } > #endif > } > --
On 18/02/2025 19:13, Rafael J. Wysocki wrote: > +Saket Dumbre > > On Thu, Feb 13, 2025 at 7:16 PM Colin Ian King <colin.i.king@gmail.com> wrote: >> >> In cases where the ACPI AML contains errors there can be quite a large >> amount of ACPICA kernel log spamming. > > Except when someone wants to see them all which also happens. > > And wouldn't this also rate limit debug messages from ACPICA > specifically enabled via the kernel command line? > > If so, I'd rather find a way to tell ACPICA to be less verbose. Yep, makes sense Rafael. The crux of the matter is that there should be some way to reduce repeated identical "ACPI BIOS Error" messages, especially when the AML is triggering this 10+ times a second :-) Colin > >> Reduce this by rate limiting the messages. >> >> Signed-off-by: Colin Ian King <colin.i.king@gmail.com> >> --- >> drivers/acpi/osl.c | 20 ++++++++++++-------- >> 1 file changed, 12 insertions(+), 8 deletions(-) >> >> diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c >> index 5ff343096ece..d4ece68e0fd6 100644 >> --- a/drivers/acpi/osl.c >> +++ b/drivers/acpi/osl.c >> @@ -159,17 +159,21 @@ void __printf(1, 0) acpi_os_vprintf(const char *fmt, va_list args) >> if (acpi_in_debugger) { >> kdb_printf("%s", buffer); >> } else { >> - if (printk_get_level(buffer)) >> - printk("%s", buffer); >> - else >> - printk(KERN_CONT "%s", buffer); >> + if (printk_ratelimit()) { >> + if (printk_get_level(buffer)) >> + printk("%s", buffer); >> + else >> + printk(KERN_CONT "%s", buffer); >> + } >> } >> #else >> if (acpi_debugger_write_log(buffer) < 0) { >> - if (printk_get_level(buffer)) >> - printk("%s", buffer); >> - else >> - printk(KERN_CONT "%s", buffer); >> + if (printk_ratelimit()) { >> + if (printk_get_level(buffer)) >> + printk("%s", buffer); >> + else >> + printk(KERN_CONT "%s", buffer); >> + } >> } >> #endif >> } >> --
diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index 5ff343096ece..d4ece68e0fd6 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -159,17 +159,21 @@ void __printf(1, 0) acpi_os_vprintf(const char *fmt, va_list args) if (acpi_in_debugger) { kdb_printf("%s", buffer); } else { - if (printk_get_level(buffer)) - printk("%s", buffer); - else - printk(KERN_CONT "%s", buffer); + if (printk_ratelimit()) { + if (printk_get_level(buffer)) + printk("%s", buffer); + else + printk(KERN_CONT "%s", buffer); + } } #else if (acpi_debugger_write_log(buffer) < 0) { - if (printk_get_level(buffer)) - printk("%s", buffer); - else - printk(KERN_CONT "%s", buffer); + if (printk_ratelimit()) { + if (printk_get_level(buffer)) + printk("%s", buffer); + else + printk(KERN_CONT "%s", buffer); + } } #endif }
In cases where the ACPI AML contains errors there can be quite a large amount of ACPICA kernel log spamming. Reduce this by rate limiting the messages. Signed-off-by: Colin Ian King <colin.i.king@gmail.com> --- drivers/acpi/osl.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-)