----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/252/#review567 -----------------------------------------------------------
I guess there is no real reason why there's a space before some semicolons but not others, or is there? Even although I like to look with the space there, I think we should remove those spaces to be consistent with common practice. indra/newview/llviewertexturelist.h <http://codereview.secondlife.com/r/252/#comment495> Any reason for this to be BOOL instead of bool? Also, remove the space between the name and the semicolon. indra/newview/llviewertexturelist.cpp <http://codereview.secondlife.com/r/252/#comment502> Remove space between value and semicolon. indra/newview/llviewertexturelist.cpp <http://codereview.secondlife.com/r/252/#comment507> I guess the intention is that this gets only executed by one single thread. (And probably only once?) So before setting sRenderThreadID to the current thread's ID, we might want to assert that it hasn't been changed before (i.e., is still 0). (If we want to allow to execute this assignment several times but only by one thread, check whether it's 0 or already has the value of LLThread::currentID().) indra/newview/llviewertexturelist.cpp <http://codereview.secondlife.com/r/252/#comment503> Remove space before semicolon. indra/newview/llviewertexturelist.cpp <http://codereview.secondlife.com/r/252/#comment504> Same here. indra/newview/llviewertexturelist.cpp <http://codereview.secondlife.com/r/252/#comment505> And while we're at it, here, too. indra/newview/llviewertexturelist.cpp <http://codereview.secondlife.com/r/252/#comment506> And there's more of these spaces. (I won't comment on all.) - Boroondas On April 8, 2011, 10:26 a.m., Xiaohong Bao wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://codereview.secondlife.com/r/252/ > ----------------------------------------------------------- > > (Updated April 8, 2011, 10:26 a.m.) > > > Review request for Viewer. > > > Summary > ------- > > this is to resubmit the patch for storm-973. > > We are not very clear what causes this. But this fix is targeting three most > possible causes: > 1, a texture is failed to add into mImageList but its flag is set to be > successful; > 2, a texture status is changed not from the main thread, because gTextureList > is not thread-safe; > 3, gTextureList is accessed before it is initialized. > > I regenerated the viewer-development-storm-973 branch based on the latest > viewer-development branch. If you still can not apply the patch directly, I > am afraid you should do the manual merge. Otherwise grant me the permission, > I will do it. > > > This addresses bug storm-973. > http://jira.secondlife.com/browse/storm-973 > > > Diffs > ----- > > indra/newview/lldrawpoolbump.cpp 13670741a0a8 > indra/newview/llviewertexturelist.h 13670741a0a8 > indra/newview/llviewertexturelist.cpp 13670741a0a8 > > Diff: http://codereview.secondlife.com/r/252/diff > > > Testing > ------- > > > Thanks, > > Xiaohong > >
_______________________________________________ Policies and (un)subscribe information available here: http://wiki.secondlife.com/wiki/OpenSource-Dev Please read the policies before posting to keep unmoderated posting privileges
