[GTK] Fix compile warnings in CryptoKeyECGCrypt.cpp

[Hyungwook Lee]

../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp: In function ‘size_t WebCore::curveSize(WebCore::CryptoKeyEC::NamedCurve)’:
../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp:48:1: warning: control reaches end of non-void function [-Wreturn-type]
 }
 ^
../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp: In function ‘const char* WebCore::curveName(WebCore::CryptoKeyEC::NamedCurve)’:
../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp:58:1: warning: control reaches end of non-void function [-Wreturn-type]
 }
 ^
../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp: In function ‘unsigned int WebCore::uncompressedPointSizeForCurve(WebCore::CryptoKeyEC::NamedCurve)’:
../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp:68:1: warning: control reaches end of non-void function [-Wreturn-type]
 }
 ^
../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp: In function ‘unsigned int WebCore::uncompressedFieldElementSizeForCurve(WebCore::CryptoKeyEC::NamedCurve)’:
../../Source/WebCore/crypto/gcrypt/CryptoKeyECGCrypt.cpp:78:1: warning: control reaches end of non-void function [-Wreturn-type]

[Hyungwook Lee]

Created attachment 309483 [details]
Patch

[Michael Catanzaro]

Comment on attachment 309483 [details]
Patch

Actually, I'll make a topic about this on webkit-dev, since we should find a standard way of handling this very common cause of warnings. I've been fixing these by adding a RELEASE_ASSERT_NOT_REACHED() in the default case and not having any code below the switch.

[Michael Catanzaro]

See https://lists.webkit.org/pipermail/webkit-dev/2017-May/029023.html

So far, it looks like most people are favoring the approach you took here.

[WebKit Commit Bot]

Comment on attachment 309483 [details]
Patch

Clearing flags on attachment: 309483

Committed r216581: <http://trac.webkit.org/changeset/216581>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.