| Message ID | eb2e1de12dad4c11454b009075f916c63de8a2a6.1512774279.git.dvhart@infradead.org (mailing list archive) |
|---|---|
| State | Accepted, archived |
| Delegated to: | Darren Hart |
| Headers | show |
On Saturday, December 9, 2017 12:07:08 AM CET Darren Hart (VMware) wrote: > The new notify_handler logic determining if autorelease should be used or > not is a bit awkward, and can result in more than one call to > sparse_keymap_report_event for the same event (scancode). The nesting > and long lines also made it difficult to read. > > Simplify the logic by eliminating a level of nesting with a goto and > always calculate autorelease and val so we can make a single call to > sparse_keymap_report_event. > > Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org> > Cc: Stefan Brüns <stefan.bruens@rwth-aachen.de> > Cc: AceLan Kao <acelan.kao@canonical.com> > --- > Note: This is based on top of Stefan's v2 patch series for intel-vbtn, > currently in my review-dvhart branch. Hi Darren, is this tree publically available? Thanks, Stefan
On Sat, Dec 09, 2017 at 12:28:18AM +0100, Stefan Brüns wrote: > On Saturday, December 9, 2017 12:07:08 AM CET Darren Hart (VMware) wrote: > > The new notify_handler logic determining if autorelease should be used or > > not is a bit awkward, and can result in more than one call to > > sparse_keymap_report_event for the same event (scancode). The nesting > > and long lines also made it difficult to read. > > > > Simplify the logic by eliminating a level of nesting with a goto and > > always calculate autorelease and val so we can make a single call to > > sparse_keymap_report_event. > > > > Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org> > > Cc: Stefan Brüns <stefan.bruens@rwth-aachen.de> > > Cc: AceLan Kao <acelan.kao@canonical.com> > > --- > > Note: This is based on top of Stefan's v2 patch series for intel-vbtn, > > currently in my review-dvhart branch. > > Hi Darren, > > is this tree publically available? Yes, in the tree listed in MAINTAINERS: git://git.infradead.org/users/dvhart/linux-platform-drivers-x86.git which is actually an alias I should now update to: git://git.infradead.org/linux-platform-drivers-x86.git Note that the review-* branches exist for CI/CD tooling and will rebase often. testing, fixes, and for-next are the only branches considered to be "published".
On Saturday, December 9, 2017 12:07:08 AM CET Darren Hart (VMware) wrote: > The new notify_handler logic determining if autorelease should be used or > not is a bit awkward, and can result in more than one call to > sparse_keymap_report_event for the same event (scancode). The nesting > and long lines also made it difficult to read. > > Simplify the logic by eliminating a level of nesting with a goto and > always calculate autorelease and val so we can make a single call to > sparse_keymap_report_event. > > Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org> > Cc: Stefan Brüns <stefan.bruens@rwth-aachen.de> > Cc: AceLan Kao <acelan.kao@canonical.com> > --- > Note: This is based on top of Stefan's v2 patch series for intel-vbtn, > currently in my review-dvhart branch. > > drivers/platform/x86/intel-vbtn.c | 25 +++++++++++++------------ > 1 file changed, 13 insertions(+), 12 deletions(-) Reviewed-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Tested-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Regards, Stefan
On Mon, Dec 11, 2017 at 02:48:23AM +0100, Stefan Brüns wrote: > On Saturday, December 9, 2017 12:07:08 AM CET Darren Hart (VMware) wrote: > > The new notify_handler logic determining if autorelease should be used or > > not is a bit awkward, and can result in more than one call to > > sparse_keymap_report_event for the same event (scancode). The nesting > > and long lines also made it difficult to read. > > > > Simplify the logic by eliminating a level of nesting with a goto and > > always calculate autorelease and val so we can make a single call to > > sparse_keymap_report_event. > > > > Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org> > > Cc: Stefan Brüns <stefan.bruens@rwth-aachen.de> > > Cc: AceLan Kao <acelan.kao@canonical.com> > > --- > > Note: This is based on top of Stefan's v2 patch series for intel-vbtn, > > currently in my review-dvhart branch. > > > > drivers/platform/x86/intel-vbtn.c | 25 +++++++++++++------------ > > 1 file changed, 13 insertions(+), 12 deletions(-) > > Reviewed-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> > Tested-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Appreciate you taking the time to review and test, thank you!
diff --git a/drivers/platform/x86/intel-vbtn.c b/drivers/platform/x86/intel-vbtn.c index 0861efe..297e1ac 100644 --- a/drivers/platform/x86/intel-vbtn.c +++ b/drivers/platform/x86/intel-vbtn.c @@ -80,6 +80,7 @@ static void notify_handler(acpi_handle handle, u32 event, void *context) { struct platform_device *device = context; struct intel_vbtn_priv *priv = dev_get_drvdata(&device->dev); + unsigned int val = !(event & 1); /* Even=press, Odd=release */ const struct key_entry *ke_rel; bool autorelease; @@ -88,20 +89,20 @@ static void notify_handler(acpi_handle handle, u32 event, void *context) pm_wakeup_hard_event(&device->dev); return; } - } else { - /* Use the fact press/release come in even/odd pairs */ - if ((event & 1) && sparse_keymap_report_event(priv->input_dev, - event, 0, false)) - return; + goto out_unknown; + } - ke_rel = sparse_keymap_entry_from_scancode(priv->input_dev, - event | 1); - autorelease = !ke_rel || ke_rel->type == KE_IGNORE; + /* + * Event press events are autorelease if there is no corresponding odd + * release event, or if the odd event is KE_IGNORE. + */ + ke_rel = sparse_keymap_entry_from_scancode(priv->input_dev, event | 1); + autorelease = val && (!ke_rel || ke_rel->type == KE_IGNORE); - if (sparse_keymap_report_event(priv->input_dev, event, 1, - autorelease)) - return; - } + if (sparse_keymap_report_event(priv->input_dev, event, val, autorelease)) + return; + +out_unknown: dev_dbg(&device->dev, "unknown event index 0x%x\n", event); }
The new notify_handler logic determining if autorelease should be used or not is a bit awkward, and can result in more than one call to sparse_keymap_report_event for the same event (scancode). The nesting and long lines also made it difficult to read. Simplify the logic by eliminating a level of nesting with a goto and always calculate autorelease and val so we can make a single call to sparse_keymap_report_event. Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org> Cc: Stefan Brüns <stefan.bruens@rwth-aachen.de> Cc: AceLan Kao <acelan.kao@canonical.com> --- Note: This is based on top of Stefan's v2 patch series for intel-vbtn, currently in my review-dvhart branch. drivers/platform/x86/intel-vbtn.c | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-)