69669 2011-10-07 14:27:09 -0700 WebKit2/Shared/APIClient.h complains that array subscript is below array bounds 2014-02-05 10:51:12 -0800 1 1 1 Unclassified WebKit WebKit2 528+ (Nightly build) PC Linux NEW P2 Normal --- 1 rafael.lobo webkit-unassigned ap cmarcelo kling menard oldest_to_newest 480314 0 rafael.lobo 2011-10-07 14:27:09 -0700 I'm getting the following warning trying to build WebKit2 (compiling Source/WebKit2/UIProcess/WebContext.cpp): cc1plus: warnings being treated as errors ../../../Source/WebKit2/Shared/APIClient.h: In member function 'void WebKit::WebContext::initializeInjectedBundleClient(const WKContextInjectedBundleClient*)': ../../../Source/WebKit2/Shared/APIClient.h:52:13: error: array subscript is below array bounds I'm using openSUSE 11.4 (x86_64) and g++ (SUSE Linux) 4.5.1 20101208 [gcc-4_5-branch revision 167585]. I wonder if someone else have the same problem. 480345 1 110225 rafael.lobo 2011-10-07 15:22:01 -0700 Created attachment 110225 Just ignored the warning on APIClient 480371 2 110225 darin 2011-10-07 16:26:07 -0700 Comment on attachment 110225 Just ignored the warning on APIClient View in context: https://bugs.webkit.org/attachment.cgi?id=110225&action=review Could you be more specific about what the warning is? Maybe it’s fixable instead of something we need to turn off. > Source/WebKit2/Shared/APIClient.h:40 > +#pragma GCC diagnostic ignored "-Warray-bounds" I don’t think this is the best practice for using a compiler-specific pragma. In the past we have used #if around them. > Source/WebKit2/Shared/APIClient.h:55 > +#pragma GCC diagnostic warning "-Warray-bounds" It’s not good to unconditionally turn a warning on even if it wasn’t on before; instead you should use push/pop. 482365 3 110674 rafael.lobo 2011-10-12 05:45:22 -0700 Created attachment 110674 Added a check for non-negative value for array's index I've googled a bit about this warning and I've found out that sometimes the compiler make some optimizations on the code and the access on arrays may be negative values after these changes (sources: http://www.mail-archive.com/gcc@gcc.gnu.org/msg46397.html and http://gcc.gnu.org/ml/gcc/2009-09/msg00270.html). So I've just added a check for non-negative values for that index and it has stopped complaining. :) 483318 4 110674 menard 2011-10-13 11:03:42 -0700 Comment on attachment 110674 Added a check for non-negative value for array's index View in context: https://bugs.webkit.org/attachment.cgi?id=110674&action=review > Source/WebKit2/Shared/APIClient.h:51 > + if (client && client->version < currentVersion && client->version >= 0) Can currentVersion be a unsigned? 483331 5 menard 2011-10-13 11:14:26 -0700 (In reply to comment #4) > (From update of attachment 110674 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=110674&action=review > > > Source/WebKit2/Shared/APIClient.h:51 > > + if (client && client->version < currentVersion && client->version >= 0) > > Can currentVersion be a unsigned? And client->version obviously. Not sure if that's possible but it looks like it could be a better fix. 483349 6 rafael.lobo 2011-10-13 11:44:45 -0700 (In reply to comment #4) > (From update of attachment 110674 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=110674&action=review > > > Source/WebKit2/Shared/APIClient.h:51 > > + if (client && client->version < currentVersion && client->version >= 0) > > Can currentVersion be a unsigned? I've tried to change it to be unsigned, but then the compiler started to complain about comparison of unsigned expression < 0 being always false. The expression it is talking about is "client->version < currentVersion" when currentVersion is 0 which is possible. So I would just get into another warning. 976822 7 110674 andersca 2014-02-05 10:51:12 -0800 Comment on attachment 110674 Added a check for non-negative value for array's index Clearing review flag on patches from before 2014. If this patch is still relevant, please reset the r? flag. 110225 2011-10-07 15:22:01 -0700 2011-10-12 05:45:22 -0700 Just ignored the warning on APIClient warning.diff text/plain 1491 rafael.lobo ZGlmZiAtLWdpdCBhL1NvdXJjZS9XZWJLaXQyL0NoYW5nZUxvZyBiL1NvdXJjZS9XZWJLaXQyL0No YW5nZUxvZwppbmRleCAzNGRiNmZiLi5lMTlhMWY0IDEwMDY0NAotLS0gYS9Tb3VyY2UvV2ViS2l0 Mi9DaGFuZ2VMb2cKKysrIGIvU291cmNlL1dlYktpdDIvQ2hhbmdlTG9nCkBAIC0xLDMgKzEsMTUg QEAKKzIwMTEtMTAtMDcgIFJhZmFlbCBCcmFuZGFvICA8cmFmYWVsLmxvYm9Ab3BlbmJvc3NhLm9y Zz4KKworICAgICAgICBEaXNhYmxlIHdhcm5pbmcgKGFycmF5IGJvdW5kcykgb24gV2ViS2l0MiBB UElDbGllbnQKKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lk PTY5NjY5CisKKyAgICAgICAgQWRkZWQgYW4gZXhjZXB0aW9uIHRvIEFQSUNsaWVudCwgc28gaXQg d2lsbCBpZ25vcmUgYXJyYXkKKyAgICAgICAgYm91bmRzIHdhcm5pbmcgYXJvdW5kIGl0cyBmdW5j dGlvbiBjYWxsZWQgJ2luaXRpYWxpemUnLgorCisgICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAo T09QUyEpLgorCisgICAgICAgICogU2hhcmVkL0FQSUNsaWVudC5oOgorCiAyMDExLTEwLTA3ICBT aGVyaWZmIEJvdCAgPHdlYmtpdC5yZXZpZXcuYm90QGdtYWlsLmNvbT4KIAogICAgICAgICBVbnJl dmlld2VkLCByb2xsaW5nIG91dCByOTY5MzguCmRpZmYgLS1naXQgYS9Tb3VyY2UvV2ViS2l0Mi9T aGFyZWQvQVBJQ2xpZW50LmggYi9Tb3VyY2UvV2ViS2l0Mi9TaGFyZWQvQVBJQ2xpZW50LmgKaW5k ZXggYmU2ZWYzYS4uOWZiMjkyNSAxMDA2NDQKLS0tIGEvU291cmNlL1dlYktpdDIvU2hhcmVkL0FQ SUNsaWVudC5oCisrKyBiL1NvdXJjZS9XZWJLaXQyL1NoYXJlZC9BUElDbGllbnQuaApAQCAtMzcs NiArMzcsNyBAQCBwdWJsaWM6CiAgICAgICAgIGluaXRpYWxpemUoMCk7CiAgICAgfQogICAgIAor I3ByYWdtYSBHQ0MgZGlhZ25vc3RpYyBpZ25vcmVkICItV2FycmF5LWJvdW5kcyIKICAgICB2b2lk IGluaXRpYWxpemUoY29uc3QgQ2xpZW50SW50ZXJmYWNlKiBjbGllbnQpCiAgICAgewogICAgICAg ICBDT01QSUxFX0FTU0VSVChzaXplb2YoQVBJQ2xpZW50VHJhaXRzPENsaWVudEludGVyZmFjZT46 OmludGVyZmFjZVNpemVzQnlWZXJzaW9uKSAvIHNpemVvZihzaXplX3QpID09IGN1cnJlbnRWZXJz aW9uICsgMSwgc2l6ZV9vZl9zb21lX2ludGVyZmFjZXNfYXJlX3Vua25vd24pOwpAQCAtNTEsNiAr NTIsNyBAQCBwdWJsaWM6CiAgICAgICAgIGlmIChjbGllbnQgJiYgY2xpZW50LT52ZXJzaW9uIDwg Y3VycmVudFZlcnNpb24pCiAgICAgICAgICAgICBtZW1jcHkoJm1fY2xpZW50LCBjbGllbnQsIEFQ SUNsaWVudFRyYWl0czxDbGllbnRJbnRlcmZhY2U+OjppbnRlcmZhY2VTaXplc0J5VmVyc2lvbltj bGllbnQtPnZlcnNpb25dKTsKICAgICB9CisjcHJhZ21hIEdDQyBkaWFnbm9zdGljIHdhcm5pbmcg Ii1XYXJyYXktYm91bmRzIgogICAgIAogcHJvdGVjdGVkOgogICAgIENsaWVudEludGVyZmFjZSBt X2NsaWVudDsK 110674 2011-10-12 05:45:22 -0700 2014-02-05 10:51:12 -0800 Added a check for non-negative value for array's index fix-warning-below-array-bounds.diff text/plain 1189 rafael.lobo ZGlmZiAtLWdpdCBhL1NvdXJjZS9XZWJLaXQyL0NoYW5nZUxvZyBiL1NvdXJjZS9XZWJLaXQyL0No YW5nZUxvZwppbmRleCAzNGRiNmZiLi40MzJiZGExIDEwMDY0NAotLS0gYS9Tb3VyY2UvV2ViS2l0 Mi9DaGFuZ2VMb2cKKysrIGIvU291cmNlL1dlYktpdDIvQ2hhbmdlTG9nCkBAIC0xLDMgKzEsMTUg QEAKKzIwMTEtMTAtMTIgIFJhZmFlbCBCcmFuZGFvICA8cmFmYWVsLmxvYm9Ab3BlbmJvc3NhLm9y Zz4KKworICAgICAgICBBdm9pZCB3YXJuaW5nIChiZWxvdyBhcnJheSBib3VuZHMpIG9uIFdlYktp dDIgQVBJQ2xpZW50CisgICAgICAgIGh0dHBzOi8vYnVncy53ZWJraXQub3JnL3Nob3dfYnVnLmNn aT9pZD02OTY2OQorCisgICAgICAgIEZvcmNlcyBjb21waWxlciB0byBpZGVudGlmeSB0aGUgYXJy YXkncyBpbmRleCBhcyBhIG5vbi1uZWdhdGl2ZSB2YWx1ZS4KKworICAgICAgICBSZXZpZXdlZCBi eSBOT0JPRFkgKE9PUFMhKS4KKworICAgICAgICAqIFNoYXJlZC9BUElDbGllbnQuaDoKKyAgICAg ICAgKFdlYktpdDo6QVBJQ2xpZW50Ojppbml0aWFsaXplKToKKwogMjAxMS0xMC0wNyAgU2hlcmlm ZiBCb3QgIDx3ZWJraXQucmV2aWV3LmJvdEBnbWFpbC5jb20+CiAKICAgICAgICAgVW5yZXZpZXdl ZCwgcm9sbGluZyBvdXQgcjk2OTM4LgpkaWZmIC0tZ2l0IGEvU291cmNlL1dlYktpdDIvU2hhcmVk L0FQSUNsaWVudC5oIGIvU291cmNlL1dlYktpdDIvU2hhcmVkL0FQSUNsaWVudC5oCmluZGV4IGJl NmVmM2EuLjcxMmEzOTEgMTAwNjQ0Ci0tLSBhL1NvdXJjZS9XZWJLaXQyL1NoYXJlZC9BUElDbGll bnQuaAorKysgYi9Tb3VyY2UvV2ViS2l0Mi9TaGFyZWQvQVBJQ2xpZW50LmgKQEAgLTQ4LDcgKzQ4 LDcgQEAgcHVibGljOgogCiAgICAgICAgIG1lbXNldCgmbV9jbGllbnQsIDAsIHNpemVvZihtX2Ns aWVudCkpOwogCi0gICAgICAgIGlmIChjbGllbnQgJiYgY2xpZW50LT52ZXJzaW9uIDwgY3VycmVu dFZlcnNpb24pCisgICAgICAgIGlmIChjbGllbnQgJiYgY2xpZW50LT52ZXJzaW9uIDwgY3VycmVu dFZlcnNpb24gJiYgY2xpZW50LT52ZXJzaW9uID49IDApCiAgICAgICAgICAgICBtZW1jcHkoJm1f Y2xpZW50LCBjbGllbnQsIEFQSUNsaWVudFRyYWl0czxDbGllbnRJbnRlcmZhY2U+OjppbnRlcmZh Y2VTaXplc0J5VmVyc2lvbltjbGllbnQtPnZlcnNpb25dKTsKICAgICB9CiAgICAgCg==