87440 2012-05-24 16:47:24 -0700 DOM APIs don't deflate device dimensions by the device scale factor 2012-06-29 12:41:25 -0700 1 1 1 Unclassified WebKit New Bugs 528+ (Nightly build) Unspecified Unspecified RESOLVED INVALID P2 Normal --- 66687 1 abarth abarth aelias bdakin danakj fsamuel jamesr johnme kenneth kpiascik simon.fraser oldest_to_newest 633369 0 abarth 2012-05-24 16:47:24 -0700 Figure out what to do with devicePixelRatioForDeviceDimensions 633373 1 143928 abarth 2012-05-24 16:49:05 -0700 Created attachment 143928 work in progress 633376 2 abarth 2012-05-24 16:51:47 -0700 This patch is from the chromium-android branch. I'm trying to sort out whether this is a valuable change that we should make upstream or whether it's something we need to remove from the branch. Fady seemed to indicate that this was a valuable change, but that there were some complexities in creating a correct implementation. 633377 3 143928 simon.fraser 2012-05-24 16:52:33 -0700 Comment on attachment 143928 work in progress devicePixelRatioForDeviceDimensions just doesn't make sense. devicePixelRatio is a fixed number (1 or 2). 633380 4 fsamuel 2012-05-24 16:54:06 -0700 (In reply to comment #2) > This patch is from the chromium-android branch. I'm trying to sort out whether this is a valuable change that we should make upstream or whether it's something we need to remove from the branch. Fady seemed to indicate that this was a valuable change, but that there were some complexities in creating a correct implementation. We wanted to do the same thing but with defaultDeviceScaleFactor(), but opted for making WebCore mostly oblivious to device scaling due to bugs in fixed layout and time constraints. 633390 5 abarth 2012-05-24 17:09:11 -0700 > devicePixelRatioForDeviceDimensions just doesn't make sense. devicePixelRatio is a fixed number (1 or 2). Yes, I agree that the patch, as written, doesn't make much sense. I had first posted the reverse of this patch to the chromium-android patch to revert the change, but Fady suggested that we discuss whether it makes sense to do something like this upstream. IMHO, it's better to have those sorts of discussions in the WebKit bug tracker. ;) 633395 6 aelias 2012-05-24 17:22:53 -0700 I'm having some difficulty understanding this value and why it's defined as: result.devicePixelRatioForDeviceDimensions = isAutoDPI ? (deviceDPI / 160.0f) : result.devicePixelRatio; In the autoDPI case which is 99.9% of websites (any website without a viewport tag explicitly requesting DPI), this value is just the same as defaultDeviceScaleFactor (the physical DPI of the device). In the rare explicitly-requested-DPI case, it's devicePixelRatio (= deviceScaleFactor). I don't fully understand the need to introduce this third value, could we just use defaultDeviceScaleFactor instead? John, what do you think? 633409 7 143935 abarth 2012-05-24 17:40:53 -0700 Created attachment 143935 Alternate version that doesn't introduce a new variable 633414 8 aelias 2012-05-24 17:47:41 -0700 Hmm, looking at this further, I think we actually do need a third value because of ChromeOS's different deviceScale decisions. Clank's viewport is specified in physical pixels whereas ChromeOS's is in CSS pixels. So actually, ChromeOS does not need this diff at all, and we could rename the new value to viewportScaleFactor and set it to: viewportScaleFactor = defaultDeviceScaleFactor / WebViewImpl::m_deviceScaleInCompositor; This would give a value of 1 on ChromeOS and of defaultDeviceScaleFactor on Android. 633421 9 abarth 2012-05-24 17:59:28 -0700 It is wise for Android and ChromeOS to make different decisions in this regard? That seems like a recipe for sadness in the future. 633426 10 aelias 2012-05-24 18:04:23 -0700 Not very, no, and we had a debate about this. Each approach has its own set of bugs which impact one platform more severely than the other, so we ended up being forced to diverge for now. 633552 11 abarth 2012-05-24 22:02:26 -0700 Was that discusion captured in a bug I can read? I'm not sure how it affects this patch. 634622 12 kenneth 2012-05-26 01:50:28 -0700 (In reply to comment #3) > (From update of attachment 143928 [details]) > devicePixelRatioForDeviceDimensions just doesn't make sense. devicePixelRatio is a fixed number (1 or 2). Actually some Nokia devices and Android devices use the value of 1.5, so basically it can be 1, 1.5 and 2 634623 13 143935 kenneth 2012-05-26 01:52:43 -0700 Comment on attachment 143935 Alternate version that doesn't introduce a new variable View in context: https://bugs.webkit.org/attachment.cgi?id=143935&action=review > Source/WebCore/page/DOMWindow.cpp:1098 > - return static_cast<int>(page->chrome()->windowRect().height()); > + return static_cast<int>(page->chrome()->windowRect().height() / page->settings()->defaultDeviceScaleFactor()); hmm I don't get this. For Qt, the size of the windowRect is in device pixels and not upscaled 636682 14 aelias 2012-05-29 18:27:58 -0700 (In reply to comment #11) > Was that discusion captured in a bug I can read? I'm not sure how it affects this patch. The different ChromeOS behavior was introduced in bug 86051, not sure how much it would clarify things to read all of it though. 636701 15 danakj 2012-05-29 18:37:33 -0700 For CROS we did it in such a way that the frame's size is equal to the number of CSS pixels, so it's not inflated to physical pixels as would be done with fixed-layout-mode. So the devicePixelRatioForDeviceDimensions might be 1 in that case.. and the variable is more a ratio between CSS/layout pixels and the frame's size? 636722 16 abarth 2012-05-29 18:57:37 -0700 I'm hesitant to introduce yet another scale factor into the project. It seems that folks cannot even keep the ones we have currently straight. 636725 17 aelias 2012-05-29 19:00:04 -0700 One remaining question I have is what size viewport does the Mac port pass in from the embedder? Physical pixels or device-independent pixels? 636726 18 abarth 2012-05-29 19:02:00 -0700 (In reply to comment #17) > One remaining question I have is what size viewport does the Mac port pass in from the embedder? Physical pixels or device-independent pixels? I suspect bdakin might be able to answer your question. 636727 19 abarth 2012-05-29 19:03:49 -0700 IMHO, we should use the same approach here across ports, and Android and Chrome OS should also use that same approach. Using a different architecture on Android and Chrome OS might be easier in the short term, but it's a recipe for sadness and not in our long-term interests. 636732 20 danakj 2012-05-29 19:12:00 -0700 (In reply to comment #19) > IMHO, we should use the same approach here across ports, and Android and Chrome OS should also use that same approach. Using a different architecture on Android and Chrome OS might be easier in the short term, but it's a recipe for sadness and not in our long-term interests. AIUI Chrome OS plans to use the same approach as Android, and so these two should converge, but it was not feasible for the next Chromium milestone. The difference right now is essentially that Chrome OS applies the device scale during composite instead of as a CSS transform on the root layer. Doing otherwise at the moment broke some important things, like Google Docs and CSS fixed position elements. 636735 21 danakj 2012-05-29 19:14:56 -0700 (In reply to comment #16) > I'm hesitant to introduce yet another scale factor into the project. It seems that folks cannot even keep the ones we have currently straight. Oh yes.. I don't mean to do that. I think we have too many probably, and it's confusing because they are not the right ones currently. IMO, we should have the most basic factors possible and then combine and use them as appropriate in code. It just seems like this one's name is misleading if the device scale is not part of the page scale (and thus transforming the root layer)? 636783 22 bdakin 2012-05-29 20:37:23 -0700 (In reply to comment #17) > One remaining question I have is what size viewport does the Mac port pass in from the embedder? Physical pixels or device-independent pixels? I must admit that I don't fully understand the question. What you mean by the embedder? But generally we use CSS aka device-independent pixels for just about everything. 637628 23 johnme 2012-05-30 13:32:41 -0700 (In reply to comment #20) > The difference right now is essentially that Chrome OS applies the device scale during composite instead of as a CSS transform on the root layer. I'm confused - why is either port transforming things by the deviceScaleFactor? Do you mean the pageScaleFactor? 637661 24 danakj 2012-05-30 13:51:48 -0700 (In reply to comment #23) > (In reply to comment #20) > > The difference right now is essentially that Chrome OS applies the device scale during composite instead of as a CSS transform on the root layer. > > I'm confused - why is either port transforming things by the deviceScaleFactor? Do you mean the pageScaleFactor? In Android deviceScaleFactor is a component of pageScaleFactor, so I do. In ChromeOS right now, pageScaleFactor is always 1, and everything is scaled during draw by the deviceScaleFactor. This is to avoid scaling the Frame as pageScaleFactor would have done, and to match the Mac port more closely. 650718 25 abarth 2012-06-15 21:05:02 -0700 defaultDeviceScaleFactor doesn't exist anymore. 660536 26 johnme 2012-06-29 08:55:16 -0700 (In reply to comment #25) > defaultDeviceScaleFactor doesn't exist anymore. That seems orthogonal to "DOM APIs don't deflate device dimensions by the device scale factor". We still need to divide device dimensions by deviceScaleFactor on Android and other ports that (unfortunately) size their screen/windowRect in physical screen pixels (as mentioned in comment #8, Chrome OS doesn't need this since they apparently already have it in CSS pixels). 660538 27 jamesr 2012-06-29 08:58:10 -0700 (In reply to comment #26) > (In reply to comment #25) > > defaultDeviceScaleFactor doesn't exist anymore. > > That seems orthogonal to "DOM APIs don't deflate device dimensions by the device scale factor". We still need to divide device dimensions by deviceScaleFactor on Android and other ports that (unfortunately) size their screen/windowRect in physical screen pixels (as mentioned in comment #8, Chrome OS doesn't need this since they apparently already have it in CSS pixels). I think Android should just do the same thing as ChromeOS / Mac/ etc. 660714 28 johnme 2012-06-29 12:41:25 -0700 (In reply to comment #27) > (In reply to comment #26) > > (In reply to comment #25) > > > defaultDeviceScaleFactor doesn't exist anymore. > > > > That seems orthogonal to "DOM APIs don't deflate device dimensions by the device scale factor". We still need to divide device dimensions by deviceScaleFactor on Android and other ports that (unfortunately) size their screen/windowRect in physical screen pixels (as mentioned in comment #8, Chrome OS doesn't need this since they apparently already have it in CSS pixels). > > I think Android should just do the same thing as ChromeOS / Mac/ etc. Sounds reasonable. Alex, what's our status here? 143928 2012-05-24 16:49:05 -0700 2012-05-24 16:52:33 -0700 work in progress bug-87440-20120524164903.patch text/plain 9670 abarth SW5kZXg6IFNvdXJjZS9XZWJDb3JlL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2Vi Q29yZS9DaGFuZ2VMb2cJKHJldmlzaW9uIDExODQ0MykKKysrIFNvdXJjZS9XZWJDb3JlL0NoYW5n ZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDM0IEBACisyMDEyLTA1LTI0ICBBZGFtIEJh cnRoICA8YWJhcnRoQHdlYmtpdC5vcmc+CisKKyAgICAgICAgRmlndXJlIG91dCB3aGF0IHRvIGRv IHdpdGggZGV2aWNlUGl4ZWxSYXRpb0ZvckRldmljZURpbWVuc2lvbnMKKyAgICAgICAgaHR0cHM6 Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTg3NDQwCisKKyAgICAgICAgUmV2aWV3 ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgTm8gbmV3IHRlc3RzLiAoT09QUyEpCisK KyAgICAgICAgKiBjc3MvTWVkaWFRdWVyeUV2YWx1YXRvci5jcHA6CisgICAgICAgIChXZWJDb3Jl OjpkZXZpY2VfaGVpZ2h0TWVkaWFGZWF0dXJlRXZhbCk6CisgICAgICAgIChXZWJDb3JlOjpkZXZp Y2Vfd2lkdGhNZWRpYUZlYXR1cmVFdmFsKToKKyAgICAgICAgKiBkb20vVmlld3BvcnRBcmd1bWVu dHMuY3BwOgorICAgICAgICAoV2ViQ29yZTo6Y29tcHV0ZVZpZXdwb3J0QXR0cmlidXRlcyk6Cisg ICAgICAgICogZG9tL1ZpZXdwb3J0QXJndW1lbnRzLmg6CisgICAgICAgIChWaWV3cG9ydEF0dHJp YnV0ZXMpOgorICAgICAgICAqIHBhZ2UvRE9NV2luZG93LmNwcDoKKyAgICAgICAgKFdlYkNvcmU6 OkRPTVdpbmRvdzo6b3V0ZXJIZWlnaHQpOgorICAgICAgICAoV2ViQ29yZTo6RE9NV2luZG93Ojpv dXRlcldpZHRoKToKKyAgICAgICAgKiBwYWdlL1BhZ2UuY3BwOgorICAgICAgICAoV2ViQ29yZTo6 UGFnZTo6UGFnZSk6CisgICAgICAgICogcGFnZS9QYWdlLmg6CisgICAgICAgIChXZWJDb3JlOjpQ YWdlOjpkZXZpY2VQaXhlbFJhdGlvRm9yRGV2aWNlRGltZW5zaW9ucyk6CisgICAgICAgIChXZWJD b3JlOjpQYWdlOjpzZXREZXZpY2VQaXhlbFJhdGlvRm9yRGV2aWNlRGltZW5zaW9ucyk6CisgICAg ICAgIChQYWdlKToKKyAgICAgICAgKiBwYWdlL1NjcmVlbi5jcHA6CisgICAgICAgIChXZWJDb3Jl OjpTY3JlZW46OmhlaWdodCk6CisgICAgICAgIChXZWJDb3JlOjpTY3JlZW46OndpZHRoKToKKyAg ICAgICAgKFdlYkNvcmU6OlNjcmVlbjo6YXZhaWxIZWlnaHQpOgorICAgICAgICAoV2ViQ29yZTo6 U2NyZWVuOjphdmFpbFdpZHRoKToKKwogMjAxMi0wNS0yNCAgUGFibG8gRmxvdXJldCAgPHBhYmxv ZkBtb3Rvcm9sYS5jb20+CiAKICAgICAgICAgU3VibWl0IGJ1dHRvbiBkb2Vzbid0IHN1Ym1pdCB0 aGUgZm9ybSBpZiB0aGUgZm9ybSBpcyB3cmFwcGVkIGJ5IGFuIGFuY2hvciB0YWcKSW5kZXg6IFNv dXJjZS9XZWJDb3JlL2Nzcy9NZWRpYVF1ZXJ5RXZhbHVhdG9yLmNwcAo9PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBT b3VyY2UvV2ViQ29yZS9jc3MvTWVkaWFRdWVyeUV2YWx1YXRvci5jcHAJKHJldmlzaW9uIDExODM5 MCkKKysrIFNvdXJjZS9XZWJDb3JlL2Nzcy9NZWRpYVF1ZXJ5RXZhbHVhdG9yLmNwcAkod29ya2lu ZyBjb3B5KQpAQCAtMzMyLDcgKzMzMiw3IEBAIHN0YXRpYyBib29sIGRldmljZV9oZWlnaHRNZWRp YUZlYXR1cmVFdmEKICAgICAgICAgRmxvYXRSZWN0IHNnID0gc2NyZWVuUmVjdChmcmFtZS0+cGFn ZSgpLT5tYWluRnJhbWUoKS0+dmlldygpKTsKICAgICAgICAgUmVuZGVyU3R5bGUqIHJvb3RTdHls ZSA9IGZyYW1lLT5kb2N1bWVudCgpLT5kb2N1bWVudEVsZW1lbnQoKS0+cmVuZGVyU3R5bGUoKTsK ICAgICAgICAgaW50IGxlbmd0aDsKLSAgICAgICAgbG9uZyBoZWlnaHQgPSBzZy5oZWlnaHQoKTsK KyAgICAgICAgbG9uZyBoZWlnaHQgPSBzdGF0aWNfY2FzdDxsb25nPihzZy5oZWlnaHQoKSAvIGZy YW1lLT5wYWdlKCktPmRldmljZVBpeGVsUmF0aW9Gb3JEZXZpY2VEaW1lbnNpb25zKCkpOwogICAg ICAgICBJbnNwZWN0b3JJbnN0cnVtZW50YXRpb246OmFwcGx5U2NyZWVuSGVpZ2h0T3ZlcnJpZGUo ZnJhbWUsICZoZWlnaHQpOwogICAgICAgICByZXR1cm4gY29tcHV0ZUxlbmd0aCh2YWx1ZSwgIWZy YW1lLT5kb2N1bWVudCgpLT5pblF1aXJrc01vZGUoKSwgc3R5bGUsIHJvb3RTdHlsZSwgbGVuZ3Ro KSAmJiBjb21wYXJlVmFsdWUoc3RhdGljX2Nhc3Q8aW50PihoZWlnaHQpLCBsZW5ndGgsIG9wKTsK ICAgICB9CkBAIC0zNDcsNyArMzQ3LDcgQEAgc3RhdGljIGJvb2wgZGV2aWNlX3dpZHRoTWVkaWFG ZWF0dXJlRXZhbAogICAgICAgICBGbG9hdFJlY3Qgc2cgPSBzY3JlZW5SZWN0KGZyYW1lLT5wYWdl KCktPm1haW5GcmFtZSgpLT52aWV3KCkpOwogICAgICAgICBSZW5kZXJTdHlsZSogcm9vdFN0eWxl ID0gZnJhbWUtPmRvY3VtZW50KCktPmRvY3VtZW50RWxlbWVudCgpLT5yZW5kZXJTdHlsZSgpOwog ICAgICAgICBpbnQgbGVuZ3RoOwotICAgICAgICBsb25nIHdpZHRoID0gc2cud2lkdGgoKTsKKyAg ICAgICAgbG9uZyB3aWR0aCA9IHN0YXRpY19jYXN0PGxvbmc+KHNnLndpZHRoKCkgLyBmcmFtZS0+ cGFnZSgpLT5kZXZpY2VQaXhlbFJhdGlvRm9yRGV2aWNlRGltZW5zaW9ucygpKTsKICAgICAgICAg SW5zcGVjdG9ySW5zdHJ1bWVudGF0aW9uOjphcHBseVNjcmVlbldpZHRoT3ZlcnJpZGUoZnJhbWUs ICZ3aWR0aCk7CiAgICAgICAgIHJldHVybiBjb21wdXRlTGVuZ3RoKHZhbHVlLCAhZnJhbWUtPmRv Y3VtZW50KCktPmluUXVpcmtzTW9kZSgpLCBzdHlsZSwgcm9vdFN0eWxlLCBsZW5ndGgpICYmIGNv bXBhcmVWYWx1ZShzdGF0aWNfY2FzdDxpbnQ+KHdpZHRoKSwgbGVuZ3RoLCBvcCk7CiAgICAgfQpJ bmRleDogU291cmNlL1dlYkNvcmUvZG9tL1ZpZXdwb3J0QXJndW1lbnRzLmNwcAo9PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 Ci0tLSBTb3VyY2UvV2ViQ29yZS9kb20vVmlld3BvcnRBcmd1bWVudHMuY3BwCShyZXZpc2lvbiAx MTgzOTApCisrKyBTb3VyY2UvV2ViQ29yZS9kb20vVmlld3BvcnRBcmd1bWVudHMuY3BwCSh3b3Jr aW5nIGNvcHkpCkBAIC02MSw2ICs2MSw4IEBAIFZpZXdwb3J0QXR0cmlidXRlcyBjb21wdXRlVmll d3BvcnRBdHRyaWIKICAgICAgICAgYnJlYWs7CiAgICAgfQogCisgICAgYm9vbCBpc0F1dG9EUEkg PSBhcmdzLnRhcmdldERlbnNpdHlEcGkgPT0gVmlld3BvcnRBcmd1bWVudHM6OlZhbHVlQXV0bzsK KwogICAgIHN3aXRjaCAoaW50KGFyZ3MudGFyZ2V0RGVuc2l0eURwaSkpIHsKICAgICBjYXNlIFZp ZXdwb3J0QXJndW1lbnRzOjpWYWx1ZURldmljZURQSToKICAgICAgICAgYXJncy50YXJnZXREZW5z aXR5RHBpID0gZGV2aWNlRFBJOwpAQCAtODksNiArOTEsOCBAQCBWaWV3cG9ydEF0dHJpYnV0ZXMg Y29tcHV0ZVZpZXdwb3J0QXR0cmliCiAgICAgICAgIGRldmljZUhlaWdodCAvPSByZXN1bHQuZGV2 aWNlUGl4ZWxSYXRpbzsKICAgICB9CiAKKyAgICByZXN1bHQuZGV2aWNlUGl4ZWxSYXRpb0ZvckRl dmljZURpbWVuc2lvbnMgPSBpc0F1dG9EUEkgPyAoZGV2aWNlRFBJIC8gMTYwLjBmKSA6IHJlc3Vs dC5kZXZpY2VQaXhlbFJhdGlvOworCiAgICAgc3dpdGNoIChpbnQoYXJncy53aWR0aCkpIHsKICAg ICBjYXNlIFZpZXdwb3J0QXJndW1lbnRzOjpWYWx1ZURlc2t0b3BXaWR0aDoKICAgICAgICAgYXJn cy53aWR0aCA9IGRlc2t0b3BXaWR0aDsKSW5kZXg6IFNvdXJjZS9XZWJDb3JlL2RvbS9WaWV3cG9y dEFyZ3VtZW50cy5oCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFNvdXJjZS9XZWJDb3JlL2RvbS9WaWV3cG9ydEFy Z3VtZW50cy5oCShyZXZpc2lvbiAxMTgzOTApCisrKyBTb3VyY2UvV2ViQ29yZS9kb20vVmlld3Bv cnRBcmd1bWVudHMuaAkod29ya2luZyBjb3B5KQpAQCAtNDYsNiArNDYsNyBAQCBzdHJ1Y3QgVmll d3BvcnRBdHRyaWJ1dGVzIHsKICAgICBGbG9hdFNpemUgbGF5b3V0U2l6ZTsKIAogICAgIGZsb2F0 IGRldmljZVBpeGVsUmF0aW87CisgICAgZmxvYXQgZGV2aWNlUGl4ZWxSYXRpb0ZvckRldmljZURp bWVuc2lvbnM7CiAKICAgICBmbG9hdCBpbml0aWFsU2NhbGU7CiAgICAgZmxvYXQgbWluaW11bVNj YWxlOwpJbmRleDogU291cmNlL1dlYkNvcmUvcGFnZS9ET01XaW5kb3cuY3BwCj09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0K LS0tIFNvdXJjZS9XZWJDb3JlL3BhZ2UvRE9NV2luZG93LmNwcAkocmV2aXNpb24gMTE4MzkwKQor KysgU291cmNlL1dlYkNvcmUvcGFnZS9ET01XaW5kb3cuY3BwCSh3b3JraW5nIGNvcHkpCkBAIC0x MDk1LDcgKzEwOTUsNyBAQCBpbnQgRE9NV2luZG93OjpvdXRlckhlaWdodCgpIGNvbnN0CiAgICAg aWYgKCFwYWdlKQogICAgICAgICByZXR1cm4gMDsKIAotICAgIHJldHVybiBzdGF0aWNfY2FzdDxp bnQ+KHBhZ2UtPmNocm9tZSgpLT53aW5kb3dSZWN0KCkuaGVpZ2h0KCkpOworICAgIHJldHVybiBz dGF0aWNfY2FzdDxpbnQ+KHBhZ2UtPmNocm9tZSgpLT53aW5kb3dSZWN0KCkuaGVpZ2h0KCkgLyBw YWdlLT5kZXZpY2VQaXhlbFJhdGlvRm9yRGV2aWNlRGltZW5zaW9ucygpKTsKIH0KIAogaW50IERP TVdpbmRvdzo6b3V0ZXJXaWR0aCgpIGNvbnN0CkBAIC0xMTA3LDcgKzExMDcsNyBAQCBpbnQgRE9N V2luZG93OjpvdXRlcldpZHRoKCkgY29uc3QKICAgICBpZiAoIXBhZ2UpCiAgICAgICAgIHJldHVy biAwOwogCi0gICAgcmV0dXJuIHN0YXRpY19jYXN0PGludD4ocGFnZS0+Y2hyb21lKCktPndpbmRv d1JlY3QoKS53aWR0aCgpKTsKKyAgICByZXR1cm4gc3RhdGljX2Nhc3Q8aW50PihwYWdlLT5jaHJv bWUoKS0+d2luZG93UmVjdCgpLndpZHRoKCkgLyBwYWdlLT5kZXZpY2VQaXhlbFJhdGlvRm9yRGV2 aWNlRGltZW5zaW9ucygpKTsKIH0KIAogaW50IERPTVdpbmRvdzo6aW5uZXJIZWlnaHQoKSBjb25z dApJbmRleDogU291cmNlL1dlYkNvcmUvcGFnZS9QYWdlLmNwcAo9PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3Vy Y2UvV2ViQ29yZS9wYWdlL1BhZ2UuY3BwCShyZXZpc2lvbiAxMTgzOTApCisrKyBTb3VyY2UvV2Vi Q29yZS9wYWdlL1BhZ2UuY3BwCSh3b3JraW5nIGNvcHkpCkBAIC0xNDQsNiArMTQ0LDcgQEAgUGFn ZTo6UGFnZShQYWdlQ2xpZW50cyYgcGFnZUNsaWVudHMpCiAgICAgLCBtX21lZGlhVm9sdW1lKDEp CiAgICAgLCBtX3BhZ2VTY2FsZUZhY3RvcigxKQogICAgICwgbV9kZXZpY2VTY2FsZUZhY3Rvcigx KQorICAgICwgbV9kZXZpY2VQaXhlbFJhdGlvRm9yRGV2aWNlRGltZW5zaW9ucygxKQogICAgICwg bV9qYXZhU2NyaXB0VVJMc0FyZUFsbG93ZWQodHJ1ZSkKICAgICAsIG1fZGlkTG9hZFVzZXJTdHls ZVNoZWV0KGZhbHNlKQogICAgICwgbV91c2VyU3R5bGVTaGVldE1vZGlmaWNhdGlvblRpbWUoMCkK SW5kZXg6IFNvdXJjZS9XZWJDb3JlL3BhZ2UvUGFnZS5oCj09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFNvdXJjZS9X ZWJDb3JlL3BhZ2UvUGFnZS5oCShyZXZpc2lvbiAxMTgzOTApCisrKyBTb3VyY2UvV2ViQ29yZS9w YWdlL1BhZ2UuaAkod29ya2luZyBjb3B5KQpAQCAtMjQ5LDYgKzI0OSw5IEBAIG5hbWVzcGFjZSBX ZWJDb3JlIHsKICAgICAgICAgZmxvYXQgZGV2aWNlU2NhbGVGYWN0b3IoKSBjb25zdCB7IHJldHVy biBtX2RldmljZVNjYWxlRmFjdG9yOyB9CiAgICAgICAgIHZvaWQgc2V0RGV2aWNlU2NhbGVGYWN0 b3IoZmxvYXQpOwogCisgICAgICAgIGZsb2F0IGRldmljZVBpeGVsUmF0aW9Gb3JEZXZpY2VEaW1l bnNpb25zKCkgY29uc3QgeyByZXR1cm4gbV9kZXZpY2VQaXhlbFJhdGlvRm9yRGV2aWNlRGltZW5z aW9uczsgfQorICAgICAgICB2b2lkIHNldERldmljZVBpeGVsUmF0aW9Gb3JEZXZpY2VEaW1lbnNp b25zKGZsb2F0IHNjYWxlRmFjdG9yKSB7IG1fZGV2aWNlUGl4ZWxSYXRpb0ZvckRldmljZURpbWVu c2lvbnMgPSBzY2FsZUZhY3RvcjsgfQorCiAgICAgICAgIHN0cnVjdCBQYWdpbmF0aW9uIHsKICAg ICAgICAgICAgIGVudW0gTW9kZSB7IFVucGFnaW5hdGVkLCBIb3Jpem9udGFsbHlQYWdpbmF0ZWQs IFZlcnRpY2FsbHlQYWdpbmF0ZWQgfTsKIApAQCAtNDEwLDYgKzQxMyw3IEBAIG5hbWVzcGFjZSBX ZWJDb3JlIHsKIAogICAgICAgICBmbG9hdCBtX3BhZ2VTY2FsZUZhY3RvcjsKICAgICAgICAgZmxv YXQgbV9kZXZpY2VTY2FsZUZhY3RvcjsKKyAgICAgICAgZmxvYXQgbV9kZXZpY2VQaXhlbFJhdGlv Rm9yRGV2aWNlRGltZW5zaW9uczsKIAogICAgICAgICBQYWdpbmF0aW9uIG1fcGFnaW5hdGlvbjsK IApJbmRleDogU291cmNlL1dlYkNvcmUvcGFnZS9TY3JlZW4uY3BwCj09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFNv dXJjZS9XZWJDb3JlL3BhZ2UvU2NyZWVuLmNwcAkocmV2aXNpb24gMTE4MzkwKQorKysgU291cmNl L1dlYkNvcmUvcGFnZS9TY3JlZW4uY3BwCSh3b3JraW5nIGNvcHkpCkBAIC02Miw3ICs2Miw3IEBA IHVuc2lnbmVkIFNjcmVlbjo6aGVpZ2h0KCkgY29uc3QKIHsKICAgICBpZiAoIW1fZnJhbWUpCiAg ICAgICAgIHJldHVybiAwOwotICAgIGxvbmcgaGVpZ2h0ID0gc3RhdGljX2Nhc3Q8bG9uZz4oc2Ny ZWVuUmVjdChtX2ZyYW1lLT52aWV3KCkpLmhlaWdodCgpKTsKKyAgICBsb25nIGhlaWdodCA9IHN0 YXRpY19jYXN0PGxvbmc+KHNjcmVlblJlY3QobV9mcmFtZS0+dmlldygpKS5oZWlnaHQoKSAvIG1f ZnJhbWUtPnBhZ2UoKS0+ZGV2aWNlUGl4ZWxSYXRpb0ZvckRldmljZURpbWVuc2lvbnMoKSk7CiAg ICAgSW5zcGVjdG9ySW5zdHJ1bWVudGF0aW9uOjphcHBseVNjcmVlbkhlaWdodE92ZXJyaWRlKG1f ZnJhbWUsICZoZWlnaHQpOwogICAgIHJldHVybiBzdGF0aWNfY2FzdDx1bnNpZ25lZD4oaGVpZ2h0 KTsKIH0KQEAgLTcxLDcgKzcxLDcgQEAgdW5zaWduZWQgU2NyZWVuOjp3aWR0aCgpIGNvbnN0CiB7 CiAgICAgaWYgKCFtX2ZyYW1lKQogICAgICAgICByZXR1cm4gMDsKLSAgICBsb25nIHdpZHRoID0g c3RhdGljX2Nhc3Q8bG9uZz4oc2NyZWVuUmVjdChtX2ZyYW1lLT52aWV3KCkpLndpZHRoKCkpOwor ICAgIGxvbmcgd2lkdGggPSBzdGF0aWNfY2FzdDxsb25nPihzY3JlZW5SZWN0KG1fZnJhbWUtPnZp ZXcoKSkud2lkdGgoKSAvIG1fZnJhbWUtPnBhZ2UoKS0+ZGV2aWNlUGl4ZWxSYXRpb0ZvckRldmlj ZURpbWVuc2lvbnMoKSk7CiAgICAgSW5zcGVjdG9ySW5zdHJ1bWVudGF0aW9uOjphcHBseVNjcmVl bldpZHRoT3ZlcnJpZGUobV9mcmFtZSwgJndpZHRoKTsKICAgICByZXR1cm4gc3RhdGljX2Nhc3Q8 dW5zaWduZWQ+KHdpZHRoKTsKIH0KQEAgLTEwOCwxNCArMTA4LDE0IEBAIHVuc2lnbmVkIFNjcmVl bjo6YXZhaWxIZWlnaHQoKSBjb25zdAogewogICAgIGlmICghbV9mcmFtZSkKICAgICAgICAgcmV0 dXJuIDA7Ci0gICAgcmV0dXJuIHN0YXRpY19jYXN0PHVuc2lnbmVkPihzY3JlZW5BdmFpbGFibGVS ZWN0KG1fZnJhbWUtPnZpZXcoKSkuaGVpZ2h0KCkpOworICAgIHJldHVybiBzdGF0aWNfY2FzdDx1 bnNpZ25lZD4oc2NyZWVuQXZhaWxhYmxlUmVjdChtX2ZyYW1lLT52aWV3KCkpLmhlaWdodCgpIC8g bV9mcmFtZS0+cGFnZSgpLT5kZXZpY2VQaXhlbFJhdGlvRm9yRGV2aWNlRGltZW5zaW9ucygpKTsK IH0KIAogdW5zaWduZWQgU2NyZWVuOjphdmFpbFdpZHRoKCkgY29uc3QKIHsKICAgICBpZiAoIW1f ZnJhbWUpCiAgICAgICAgIHJldHVybiAwOwotICAgIHJldHVybiBzdGF0aWNfY2FzdDx1bnNpZ25l ZD4oc2NyZWVuQXZhaWxhYmxlUmVjdChtX2ZyYW1lLT52aWV3KCkpLndpZHRoKCkpOworICAgIHJl dHVybiBzdGF0aWNfY2FzdDx1bnNpZ25lZD4oc2NyZWVuQXZhaWxhYmxlUmVjdChtX2ZyYW1lLT52 aWV3KCkpLndpZHRoKCkgLyBtX2ZyYW1lLT5wYWdlKCktPmRldmljZVBpeGVsUmF0aW9Gb3JEZXZp Y2VEaW1lbnNpb25zKCkpOwogfQogCiB9IC8vIG5hbWVzcGFjZSBXZWJDb3JlCkluZGV4OiBTb3Vy Y2UvV2ViS2l0L2Nocm9taXVtL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2ViS2l0 L2Nocm9taXVtL0NoYW5nZUxvZwkocmV2aXNpb24gMTE4NDQzKQorKysgU291cmNlL1dlYktpdC9j aHJvbWl1bS9DaGFuZ2VMb2cJKHdvcmtpbmcgY29weSkKQEAgLTEsMyArMSwxMyBAQAorMjAxMi0w NS0yNCAgQWRhbSBCYXJ0aCAgPGFiYXJ0aEB3ZWJraXQub3JnPgorCisgICAgICAgIEZpZ3VyZSBv dXQgd2hhdCB0byBkbyB3aXRoIGRldmljZVBpeGVsUmF0aW9Gb3JEZXZpY2VEaW1lbnNpb25zCisg ICAgICAgIGh0dHBzOi8vYnVncy53ZWJraXQub3JnL3Nob3dfYnVnLmNnaT9pZD04NzQ0MAorCisg ICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAoT09QUyEpLgorCisgICAgICAgICogc3JjL0Nocm9t ZUNsaWVudEltcGwuY3BwOgorICAgICAgICAoV2ViS2l0OjpDaHJvbWVDbGllbnRJbXBsOjpkaXNw YXRjaFZpZXdwb3J0UHJvcGVydGllc0RpZENoYW5nZSk6CisKIDIwMTItMDUtMjQgIEFkYW0gQmFy dGggIDxhYmFydGhAd2Via2l0Lm9yZz4KIAogICAgICAgICBGaXggdGhlIGNocm9taXVtLWFuZHJv aWQgYnVpbGQuCkluZGV4OiBTb3VyY2UvV2ViS2l0L2Nocm9taXVtL3NyYy9DaHJvbWVDbGllbnRJ bXBsLmNwcAo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2ViS2l0L2Nocm9taXVtL3NyYy9DaHJvbWVD bGllbnRJbXBsLmNwcAkocmV2aXNpb24gMTE4MzkwKQorKysgU291cmNlL1dlYktpdC9jaHJvbWl1 bS9zcmMvQ2hyb21lQ2xpZW50SW1wbC5jcHAJKHdvcmtpbmcgY29weSkKQEAgLTY3MCw2ICs2NzAs NyBAQCB2b2lkIENocm9tZUNsaWVudEltcGw6OmRpc3BhdGNoVmlld3BvcnRQCiAgICAgICAgIG1f d2ViVmlldy0+c2V0RGV2aWNlU2NhbGVGYWN0b3Ioc2V0dGluZ3MtPmRlZmF1bHREZXZpY2VTY2Fs ZUZhY3RvcigpKTsKICAgICBlbHNlCiAgICAgICAgIG1fd2ViVmlldy0+c2V0RGV2aWNlU2NhbGVG YWN0b3IoY29tcHV0ZWQuZGV2aWNlUGl4ZWxSYXRpbyk7CisgICAgbV93ZWJWaWV3LT5wYWdlKCkt PnNldERldmljZVBpeGVsUmF0aW9Gb3JEZXZpY2VEaW1lbnNpb25zKGNvbXB1dGVkLmRldmljZVBp eGVsUmF0aW9Gb3JEZXZpY2VEaW1lbnNpb25zKTsKICAgICBtX3dlYlZpZXctPnNldFBhZ2VTY2Fs ZUZhY3RvckxpbWl0cyhjb21wdXRlZC5taW5pbXVtU2NhbGUsIGNvbXB1dGVkLm1heGltdW1TY2Fs ZSk7CiAgICAgaWYgKG5lZWRJbml0aWFsaXplUGFnZVNjYWxlKQogICAgICAgICBtX3dlYlZpZXct PnNldFBhZ2VTY2FsZUZhY3RvclByZXNlcnZpbmdTY3JvbGxPZmZzZXQoY29tcHV0ZWQuaW5pdGlh bFNjYWxlICogY29tcHV0ZWQuZGV2aWNlUGl4ZWxSYXRpbyk7Cg== 143935 2012-05-24 17:40:53 -0700 2012-05-26 01:52:43 -0700 Alternate version that doesn't introduce a new variable bug-87440-20120524174051.patch text/plain 5090 abarth SW5kZXg6IFNvdXJjZS9XZWJDb3JlL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2Vi Q29yZS9DaGFuZ2VMb2cJKHJldmlzaW9uIDExODQ0MykKKysrIFNvdXJjZS9XZWJDb3JlL0NoYW5n ZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDI0IEBACisyMDEyLTA1LTI0ICBBZGFtIEJh cnRoICA8YWJhcnRoQHdlYmtpdC5vcmc+CisKKyAgICAgICAgRE9NIEFQSXMgZG9uJ3QgZGVmbGF0 ZSBkZXZpY2UgZGltZW5zaW9ucyBieSB0aGUgZGV2aWNlIHNjYWxlIGZhY3RvcgorICAgICAgICBo dHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93X2J1Zy5jZ2k/aWQ9ODc0NDAKKworICAgICAgICBS ZXZpZXdlZCBieSBOT0JPRFkgKE9PUFMhKS4KKworICAgICAgICBObyBuZXcgdGVzdHMuIChPT1BT ISkKKworICAgICAgICAqIGNzcy9NZWRpYVF1ZXJ5RXZhbHVhdG9yLmNwcDoKKyAgICAgICAgKFdl YkNvcmU6OmRldmljZV9oZWlnaHRNZWRpYUZlYXR1cmVFdmFsKToKKyAgICAgICAgKFdlYkNvcmU6 OmRldmljZV93aWR0aE1lZGlhRmVhdHVyZUV2YWwpOgorICAgICAgICAqIHBhZ2UvRE9NV2luZG93 LmNwcDoKKyAgICAgICAgKFdlYkNvcmU6OkRPTVdpbmRvdzo6b3V0ZXJIZWlnaHQpOgorICAgICAg ICAoV2ViQ29yZTo6RE9NV2luZG93OjpvdXRlcldpZHRoKToKKyAgICAgICAgKiBwYWdlL1NjcmVl bi5jcHA6CisgICAgICAgIChXZWJDb3JlOjpTY3JlZW46OmhlaWdodCk6CisgICAgICAgIChXZWJD b3JlOjpTY3JlZW46OndpZHRoKToKKyAgICAgICAgKFdlYkNvcmU6OlNjcmVlbjo6YXZhaWxIZWln aHQpOgorICAgICAgICAoV2ViQ29yZTo6U2NyZWVuOjphdmFpbFdpZHRoKToKKwogMjAxMi0wNS0y NCAgUGFibG8gRmxvdXJldCAgPHBhYmxvZkBtb3Rvcm9sYS5jb20+CiAKICAgICAgICAgU3VibWl0 IGJ1dHRvbiBkb2Vzbid0IHN1Ym1pdCB0aGUgZm9ybSBpZiB0aGUgZm9ybSBpcyB3cmFwcGVkIGJ5 IGFuIGFuY2hvciB0YWcKSW5kZXg6IFNvdXJjZS9XZWJDb3JlL2Nzcy9NZWRpYVF1ZXJ5RXZhbHVh dG9yLmNwcAo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2ViQ29yZS9jc3MvTWVkaWFRdWVyeUV2YWx1 YXRvci5jcHAJKHJldmlzaW9uIDExODM5MCkKKysrIFNvdXJjZS9XZWJDb3JlL2Nzcy9NZWRpYVF1 ZXJ5RXZhbHVhdG9yLmNwcAkod29ya2luZyBjb3B5KQpAQCAtMzMyLDcgKzMzMiw3IEBAIHN0YXRp YyBib29sIGRldmljZV9oZWlnaHRNZWRpYUZlYXR1cmVFdmEKICAgICAgICAgRmxvYXRSZWN0IHNn ID0gc2NyZWVuUmVjdChmcmFtZS0+cGFnZSgpLT5tYWluRnJhbWUoKS0+dmlldygpKTsKICAgICAg ICAgUmVuZGVyU3R5bGUqIHJvb3RTdHlsZSA9IGZyYW1lLT5kb2N1bWVudCgpLT5kb2N1bWVudEVs ZW1lbnQoKS0+cmVuZGVyU3R5bGUoKTsKICAgICAgICAgaW50IGxlbmd0aDsKLSAgICAgICAgbG9u ZyBoZWlnaHQgPSBzZy5oZWlnaHQoKTsKKyAgICAgICAgbG9uZyBoZWlnaHQgPSBzdGF0aWNfY2Fz dDxsb25nPihzZy5oZWlnaHQoKSAvIGZyYW1lLT5wYWdlKCktPnNldHRpbmdzKCktPmRlZmF1bHRE ZXZpY2VTY2FsZUZhY3RvcigpKTsKICAgICAgICAgSW5zcGVjdG9ySW5zdHJ1bWVudGF0aW9uOjph cHBseVNjcmVlbkhlaWdodE92ZXJyaWRlKGZyYW1lLCAmaGVpZ2h0KTsKICAgICAgICAgcmV0dXJu IGNvbXB1dGVMZW5ndGgodmFsdWUsICFmcmFtZS0+ZG9jdW1lbnQoKS0+aW5RdWlya3NNb2RlKCks IHN0eWxlLCByb290U3R5bGUsIGxlbmd0aCkgJiYgY29tcGFyZVZhbHVlKHN0YXRpY19jYXN0PGlu dD4oaGVpZ2h0KSwgbGVuZ3RoLCBvcCk7CiAgICAgfQpAQCAtMzQ3LDcgKzM0Nyw3IEBAIHN0YXRp YyBib29sIGRldmljZV93aWR0aE1lZGlhRmVhdHVyZUV2YWwKICAgICAgICAgRmxvYXRSZWN0IHNn ID0gc2NyZWVuUmVjdChmcmFtZS0+cGFnZSgpLT5tYWluRnJhbWUoKS0+dmlldygpKTsKICAgICAg ICAgUmVuZGVyU3R5bGUqIHJvb3RTdHlsZSA9IGZyYW1lLT5kb2N1bWVudCgpLT5kb2N1bWVudEVs ZW1lbnQoKS0+cmVuZGVyU3R5bGUoKTsKICAgICAgICAgaW50IGxlbmd0aDsKLSAgICAgICAgbG9u ZyB3aWR0aCA9IHNnLndpZHRoKCk7CisgICAgICAgIGxvbmcgd2lkdGggPSBzdGF0aWNfY2FzdDxs b25nPihzZy53aWR0aCgpIC8gZnJhbWUtPnBhZ2UoKS0+c2V0dGluZ3MoKS0+ZGVmYXVsdERldmlj ZVNjYWxlRmFjdG9yKCkoKSk7CiAgICAgICAgIEluc3BlY3Rvckluc3RydW1lbnRhdGlvbjo6YXBw bHlTY3JlZW5XaWR0aE92ZXJyaWRlKGZyYW1lLCAmd2lkdGgpOwogICAgICAgICByZXR1cm4gY29t cHV0ZUxlbmd0aCh2YWx1ZSwgIWZyYW1lLT5kb2N1bWVudCgpLT5pblF1aXJrc01vZGUoKSwgc3R5 bGUsIHJvb3RTdHlsZSwgbGVuZ3RoKSAmJiBjb21wYXJlVmFsdWUoc3RhdGljX2Nhc3Q8aW50Pih3 aWR0aCksIGxlbmd0aCwgb3ApOwogICAgIH0KSW5kZXg6IFNvdXJjZS9XZWJDb3JlL3BhZ2UvRE9N V2luZG93LmNwcAo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09Ci0tLSBTb3VyY2UvV2ViQ29yZS9wYWdlL0RPTVdpbmRvdy5j cHAJKHJldmlzaW9uIDExODM5MCkKKysrIFNvdXJjZS9XZWJDb3JlL3BhZ2UvRE9NV2luZG93LmNw cAkod29ya2luZyBjb3B5KQpAQCAtMTA5NSw3ICsxMDk1LDcgQEAgaW50IERPTVdpbmRvdzo6b3V0 ZXJIZWlnaHQoKSBjb25zdAogICAgIGlmICghcGFnZSkKICAgICAgICAgcmV0dXJuIDA7CiAKLSAg ICByZXR1cm4gc3RhdGljX2Nhc3Q8aW50PihwYWdlLT5jaHJvbWUoKS0+d2luZG93UmVjdCgpLmhl aWdodCgpKTsKKyAgICByZXR1cm4gc3RhdGljX2Nhc3Q8aW50PihwYWdlLT5jaHJvbWUoKS0+d2lu ZG93UmVjdCgpLmhlaWdodCgpIC8gcGFnZS0+c2V0dGluZ3MoKS0+ZGVmYXVsdERldmljZVNjYWxl RmFjdG9yKCkpOwogfQogCiBpbnQgRE9NV2luZG93OjpvdXRlcldpZHRoKCkgY29uc3QKQEAgLTEx MDcsNyArMTEwNyw3IEBAIGludCBET01XaW5kb3c6Om91dGVyV2lkdGgoKSBjb25zdAogICAgIGlm ICghcGFnZSkKICAgICAgICAgcmV0dXJuIDA7CiAKLSAgICByZXR1cm4gc3RhdGljX2Nhc3Q8aW50 PihwYWdlLT5jaHJvbWUoKS0+d2luZG93UmVjdCgpLndpZHRoKCkpOworICAgIHJldHVybiBzdGF0 aWNfY2FzdDxpbnQ+KHBhZ2UtPmNocm9tZSgpLT53aW5kb3dSZWN0KCkud2lkdGgoKSAvIHBhZ2Ut PnNldHRpbmdzKCktPmRlZmF1bHREZXZpY2VTY2FsZUZhY3RvcigpKTsKIH0KIAogaW50IERPTVdp bmRvdzo6aW5uZXJIZWlnaHQoKSBjb25zdApJbmRleDogU291cmNlL1dlYkNvcmUvcGFnZS9TY3Jl ZW4uY3BwCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT0KLS0tIFNvdXJjZS9XZWJDb3JlL3BhZ2UvU2NyZWVuLmNwcAkocmV2 aXNpb24gMTE4MzkwKQorKysgU291cmNlL1dlYkNvcmUvcGFnZS9TY3JlZW4uY3BwCSh3b3JraW5n IGNvcHkpCkBAIC02Miw3ICs2Miw3IEBAIHVuc2lnbmVkIFNjcmVlbjo6aGVpZ2h0KCkgY29uc3QK IHsKICAgICBpZiAoIW1fZnJhbWUpCiAgICAgICAgIHJldHVybiAwOwotICAgIGxvbmcgaGVpZ2h0 ID0gc3RhdGljX2Nhc3Q8bG9uZz4oc2NyZWVuUmVjdChtX2ZyYW1lLT52aWV3KCkpLmhlaWdodCgp KTsKKyAgICBsb25nIGhlaWdodCA9IHN0YXRpY19jYXN0PGxvbmc+KHNjcmVlblJlY3QobV9mcmFt ZS0+dmlldygpKS5oZWlnaHQoKSAvIG1fZnJhbWUtPnBhZ2UoKS0+c2V0dGluZ3MoKS0+ZGVmYXVs dERldmljZVNjYWxlRmFjdG9yKCkpOwogICAgIEluc3BlY3Rvckluc3RydW1lbnRhdGlvbjo6YXBw bHlTY3JlZW5IZWlnaHRPdmVycmlkZShtX2ZyYW1lLCAmaGVpZ2h0KTsKICAgICByZXR1cm4gc3Rh dGljX2Nhc3Q8dW5zaWduZWQ+KGhlaWdodCk7CiB9CkBAIC03MSw3ICs3MSw3IEBAIHVuc2lnbmVk IFNjcmVlbjo6d2lkdGgoKSBjb25zdAogewogICAgIGlmICghbV9mcmFtZSkKICAgICAgICAgcmV0 dXJuIDA7Ci0gICAgbG9uZyB3aWR0aCA9IHN0YXRpY19jYXN0PGxvbmc+KHNjcmVlblJlY3QobV9m cmFtZS0+dmlldygpKS53aWR0aCgpKTsKKyAgICBsb25nIHdpZHRoID0gc3RhdGljX2Nhc3Q8bG9u Zz4oc2NyZWVuUmVjdChtX2ZyYW1lLT52aWV3KCkpLndpZHRoKCkgLyBtX2ZyYW1lLT5wYWdlKCkt PnNldHRpbmdzKCktPmRlZmF1bHREZXZpY2VTY2FsZUZhY3RvcigpKTsKICAgICBJbnNwZWN0b3JJ bnN0cnVtZW50YXRpb246OmFwcGx5U2NyZWVuV2lkdGhPdmVycmlkZShtX2ZyYW1lLCAmd2lkdGgp OwogICAgIHJldHVybiBzdGF0aWNfY2FzdDx1bnNpZ25lZD4od2lkdGgpOwogfQpAQCAtMTA4LDE0 ICsxMDgsMTQgQEAgdW5zaWduZWQgU2NyZWVuOjphdmFpbEhlaWdodCgpIGNvbnN0CiB7CiAgICAg aWYgKCFtX2ZyYW1lKQogICAgICAgICByZXR1cm4gMDsKLSAgICByZXR1cm4gc3RhdGljX2Nhc3Q8 dW5zaWduZWQ+KHNjcmVlbkF2YWlsYWJsZVJlY3QobV9mcmFtZS0+dmlldygpKS5oZWlnaHQoKSk7 CisgICAgcmV0dXJuIHN0YXRpY19jYXN0PHVuc2lnbmVkPihzY3JlZW5BdmFpbGFibGVSZWN0KG1f ZnJhbWUtPnZpZXcoKSkuaGVpZ2h0KCkgLyBtX2ZyYW1lLT5wYWdlKCktPnNldHRpbmdzKCktPmRl ZmF1bHREZXZpY2VTY2FsZUZhY3RvcigpKTsKIH0KIAogdW5zaWduZWQgU2NyZWVuOjphdmFpbFdp ZHRoKCkgY29uc3QKIHsKICAgICBpZiAoIW1fZnJhbWUpCiAgICAgICAgIHJldHVybiAwOwotICAg IHJldHVybiBzdGF0aWNfY2FzdDx1bnNpZ25lZD4oc2NyZWVuQXZhaWxhYmxlUmVjdChtX2ZyYW1l LT52aWV3KCkpLndpZHRoKCkpOworICAgIHJldHVybiBzdGF0aWNfY2FzdDx1bnNpZ25lZD4oc2Ny ZWVuQXZhaWxhYmxlUmVjdChtX2ZyYW1lLT52aWV3KCkpLndpZHRoKCkgLyBtX2ZyYW1lLT5wYWdl KCktPnNldHRpbmdzKCktPmRlZmF1bHREZXZpY2VTY2FsZUZhY3RvcigpKTsKIH0KIAogfSAvLyBu YW1lc3BhY2UgV2ViQ29yZQo=