122234 - nsProtocolProxyService::ExamineForProxy should check for both success and non empty scheme

Status: RESOLVED INCOMPLETE

(Core :: Networking, defect)

Description

[timeless]

#5  0x28948d4e in nsProtocolProxyService::ExamineForProxy (this=0x80fae80, aURI=0x8996604, aResult=0xbfbfd24c)
    at /home/timeless/mozilla/netwerk/base/src/nsProtocolProxyService.cpp:362
362         rv = ios->GetProtocolHandler(scheme.get(), getter_AddRefs(handler));
(gdb) l
357         nsXPIDLCString scheme;
358         rv = aURI->GetScheme(getter_Copies(scheme));
359         if (NS_FAILED(rv)) return rv;
360
361         nsCOMPtr<nsIProtocolHandler> handler;
362         rv = ios->GetProtocolHandler(scheme.get(), getter_AddRefs(handler));
363         if (NS_FAILED(rv)) return rv;
364
365         PRUint32 flags;
366         rv = handler->GetProtocolFlags(&flags);

#3  0x282d38d4 in nsDebug::WarnIfFalse (aStr=0x28a03cc7 "NS_ENSURE_TRUE(scheme) failed", aExpr=0x28a03cc0 "scheme",
    aFile=0x28a03b00 "/home/timeless/mozilla/netwerk/base/src/nsIOService.cpp", aLine=409) at /home/timeless/mozilla/xpcom/base/nsDebug.cpp:396
#4  0x2893fc27 in nsIOService::GetProtocolHandler (this=0x80fa780, scheme=0x0, result=0xbfbfd180)
    at /home/timeless/mozilla/netwerk/base/src/nsIOService.cpp:409
#5  0x28948d4e in nsProtocolProxyService::ExamineForProxy (this=0x80fae80, aURI=0x8996604, aResult=0xbfbfd24c)
    at /home/timeless/mozilla/netwerk/base/src/nsProtocolProxyService.cpp:362
#6  0x2894171b in nsIOService::NewChannelFromURI (this=0x80fa780, aURI=0x8996604, result=0xbfbfd4f4)
    at /home/timeless/mozilla/netwerk/base/src/nsIOService.cpp:775
#7  0x2964d67c in imgLoader::LoadImage (this=0x84fbf20, aURI=0x8996604, aLoadGroup=0x8842c40, aObserver=0x9022c10, aCX=0x8c76400, aLoadFlags=0,
    cacheKey=0x0, aRequest=0x8ac7040, _retval=0xbfbfd5f0) at /home/timeless/mozilla/modules/libpr0n/src/imgLoader.cpp:289
#8  0x29b48cc0 in nsImageFrame::RealLoadImage (this=0x900f1f8, aSpec=@0xbfbfd6a8, aPresContext=0x8c76400, aRequest=0x8ac7040)
    at /home/timeless/mozilla/layout/html/base/src/nsImageFrame.cpp:1904
#9  0x29b487b9 in nsImageFrame::LoadImage (this=0x900f1f8, aSpec=@0xbfbfd6a8, aPresContext=0x8c76400, aRequest=0x8ac7040)
    at /home/timeless/mozilla/layout/html/base/src/nsImageFrame.cpp:1846

I'll find a browser that can attach patches and attach one :)

Comment 1

[timeless]

fwiw i might be running with one unregistered patch: for QueryAttributeValue. [i am definitely running with my drop large images patch, but that's irrelevant here]

Comment 2

[timeless]

(gdb) up
#10 0x29b4273e in nsImageFrame::Init (this=0x900f1f8, aPresContext=0x8c76400, aContent=0x8ac7400, aParent=0x900f064, aContext=0x900f1a4,
    aPrevInFlow=0x0) at /home/timeless/mozilla/layout/html/base/src/nsImageFrame.cpp:319
319       rv = LoadImage(src, aPresContext, mLoads[0].mRequest); // if the image was found in the cache,
(gdb) l
314       mCanSendLoadEvent = PR_TRUE;
315
316       LoadIcons(aPresContext);
317
318       mLoads[0].mRequest = do_CreateInstance("@mozilla.org/image/request;1");
319       rv = LoadImage(src, aPresContext, mLoads[0].mRequest); // if the image was found in the cache,
320                                                              // it is possible that LoadImage will result in
321                                                              // a call to OnStartContainer()
322
323       return rv;
(gdb) p src
$8 = {<nsString> = {<nsAFlatString> = {<nsASingleFragmentString> = {<nsAString> = {
          _vptr$ = 0x28388580}, <No data fields>}, <No data fields>}, <nsStr> = {mLength = 10, mCapacity = 63, {mStr = 0xbfbfd6bc "l",
        mUStr = 0xbfbfd6bc}, mCharSize = 1 '\001', mOwnsBuffer = 0 '\000'}, <No data fields>},
  mBuffer = "l\000a\000u\000n\000c\000h\000.\000g\000i\000f\000\000\000¿¿êv¿)\214¿¿\002\000\000\000\020¿¿Ér¿)\004«î)oÄ\000\t I9)\221\217)\004«î)`ò\
000\tò\000\toñ\000\t±i\006\b`ò\000\t@¿¿A\220)lò\000\t\000\000\000\000\000\000\000\000\215\220)\004«î)oñ\000\t°Vì)X\e·) Aÿ\b"}

mBuffer appears to be 'launch.gif' which i think is from an attachment.
this is definitely a mail problem but i don't see mail on stack...

Comment 3

[Darin Fisher]

timeless: i'm having a hard time understanding what bug you are describing here.
 can you please explain what problem you are seeing?  is mozilla crashing?  or,
what?  also, please attach a patch if you can to explain what solution you are
talking about.  thx!

Comment 4

[Bradley Baetz (:bbaetz)]

Your uri doesn't have a scheme, by the look of it. Thats the real bug....

Comment 5

[Darin Fisher]

-> future

Comment 6

[Darin Fisher]

-> default owner

Comment 7

[benc]

Timeless, any thoughts on what Darin and Bradley wrote?

Comment 8

[Patrick McManus [:mcmanus]]

open question (comment 7) and this code doesn't even exist like this anymore