| Message ID | 20211121191129.256713-2-hdegoede@redhat.com (mailing list archive) |
|---|---|
| State | Accepted, archived |
| Headers | show |
| Series | thinkpad_acpi: 1 bugfix + a bunch of cleanups | expand |
diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c index 055930f5d9f5..62daf0109b4a 100644 --- a/drivers/platform/x86/thinkpad_acpi.c +++ b/drivers/platform/x86/thinkpad_acpi.c @@ -10738,8 +10738,8 @@ static int __init ibm_init(struct ibm_init_struct *iibm) if (iibm->init) { ret = iibm->init(iibm); - if (ret > 0) - return 0; /* probe failed */ + if (ret > 0 || ret == -ENODEV) + return 0; /* subdriver functionality not available */ if (ret) return ret;
Commit 79f960e29cfc ("platform/x86: thinkpad_acpi: Convert platform driver to use dev_groups") accidentally modified tpacpi_kbdlang_init() causing it to return -ENODEV instead of 0 on machines without kbdlang support (which are most of them). ibm_init() sees this -ENODEV as an error causing the entire module to not load, not good. Note that technically tpacpi_kbdlang_init() was already buggy before, it should have returned 1 instead of 0 if the feature is not present. Rather then fixing tpacpi_kbdlang_init() though, IMHO it is bettter to just make ibm_init() treat -ENODEV as 1 to fix the issue; and then in a followup commit also change all the existing "return 1"s from ibm_init_struct.init() callbacks to "return -ENODEV" as -ENODEV clearly states what it going on where as a magic return of "1" requires a deep dive into the code to figure out what is going on. This will also allow removing some extra ifs to translate -ENODEV to return 1 in a couple of init() callbacks. Fixes: 79f960e29cfc ("platform/x86: thinkpad_acpi: Convert platform driver to use dev_groups") Cc: Len Baker <len.baker@gmx.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com> --- drivers/platform/x86/thinkpad_acpi.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)