summaryrefslogtreecommitdiff
path: root/fs/jbd2/commit.c
diff options
context:
space:
mode:
authorIngrid Gallardo <ingridg@codeaurora.org>2017-01-12 17:31:15 -0800
committerIngrid Gallardo <ingridg@codeaurora.org>2017-01-16 14:55:16 -0800
commite51702e4ddecdfcabfbfcfcb9e9b40938b4f0644 (patch)
tree49d878516ab06830e0ebeb10204865e1f388816e /fs/jbd2/commit.c
parent6998daf8c5fbfa58f8fc9e33407cafeb39de714e (diff)
msm: mdss: fix for the mmss gdsc sequence
Current Display driver sets to true the ram status variable shared with the RPM to prevent MMSS GDSC to go OFF only when we are about to disable the MDSS GDSC regulator for the Idle Screen scenario; The potential problem with this is that there is a 4-vsyncs period of time between the end of the display last transaction (ping pong done) and the release of the MDSS GDSC regulator (for Idle Screen) where the Display related clocks are already released and MDSS GDSC regulator is still on (so rpm variable is still false), causing that Xo shutdown (and therefore MMSS GDSC off) can happen while the MDSS GDSC is still active during this period of time. This change makes sure that any time that the MDSS GDSC regulator is enabled or when the MDSS GDSC regulator is disabled but the Display is still active but in Idle Screen, Display sets the flag to ensure that the MMSS GDSC regulator is not disabled by the RPM, flag is released once our MDSS regulator is disabled for Suspend. Also, re-arrange the sequence to make sure the change of the flag happens after we disabled the MDSS GDSC regulator, so we prevent that by any reason the MDSS GDSC could go off before the MMSS GDSC regulator. Change-Id: I141cb893c4e2a54f99e65383e69b4ce88ab4c761 Signed-off-by: Ingrid Gallardo <ingridg@codeaurora.org>
Diffstat (limited to 'fs/jbd2/commit.c')
0 files changed, 0 insertions, 0 deletions