Age | Commit message (Collapse) | Author |
|
Change-Id: Ie4dde5aa412dae05d2629ce5675356db70b2529b
|
|
Change-Id: I9a685fd0bf2e1285299d99ea39c29501bf8fadcd
|
|
Change-Id: I27110c810b1d4d3dd3541625eff861be3e058ef6
|
|
Change-Id: Ibbf92a0156cbb086901a59c6ed0fa950601570f5
|
|
Change-Id: Ic1320988e7b47cfc2b0d589390341a56b1ff8aa7
|
|
Change-Id: Id4151d24fbb12688c7916dbcdac21ca4c6df0b42
|
|
Change-Id: I74931a6dfe08d443125956d0a93f48f7d2886601
|
|
Change-Id: I41d59d386b8f1fd5dc1fb9744a649085c66143f9
|
|
Change-Id: I28e4d96fb48420a19e51d52b89895625e7f9ba93
|
|
which had been hidden by now because of an semi-optimized code path which got
removed in the latest hyphen version
Change-Id: Ica4b0cfa16b19469fea1f630bda09eaff82e8857
|
|
...so that it is not added to a block of lines from the original commit message
that happen to start with letters followed by a colon but then not followed by a
space, as happened for 9e73f972b0e9a3e7b59c8e770b14c3d539600b06.
Change-Id: I388c194645a1577e5446891f508b31f761d1bbbf
|
|
Change-Id: Ife65c0d9417c31e80d11e2e97ffd06765d90fb1e
|
|
Change-Id: I6e242de572595fdf39553d76932bc0e953cd3a34
|
|
Even with the proposed fix of commit id d999ae10 it still seems
to be broken with reference to i#112978 . Maybe a better way
of fixing it. Hence reverting this patch partially to fix fdo#45260
|
|
Change-Id: I96cae21e79fe0b6f6cd8f3aaae2560cd1f76d65e
|
|
Change-Id: Ie4b7325eb238064540681d5e19e5d2e7666f094d
|
|
my mips supports hash-style=both, but if you actually use it ld
barfs with "gnu.hash is incompatible with the mips abi"
everywhere else we just do -Wl,--hash-style=$(WITH_LINKER_HASH_STYLE)
Change-Id: Ie310b7edad0d3068e5c4fcfed22aaea96698a06b
|
|
instead of taking it from a localized resource. This way
we have more control over localizations, no need to ping
each l10n team.
Change-Id: I4f8904b4601cd623399146ce060b9697d7b52671
|
|
Change-Id: Ia1130d241f6c01079449e670e26054bbc1a798ea
|
|
Change-Id: Ia33f9ef485fef1610b91df9895e6ac805cde0360
|
|
Change-Id: I5700a051e64171f1d2d1ec19c63fbf9778e4bc30
|
|
Change-Id: I0be64324cb92fb7411e3b341d3c1e949e8edbe09
|
|
Change-Id: Iebeb98ed2959bd4424f3f271f01a5809a03fdd70
|
|
When we do glyph fallback we check to see if the glyph is too large to fit the
space available, and if it is we try and rescale it smaller.
This worked wonderfully under Linux, but under Windows it seems that the
SalLayout impl keeps an old unscaled gdi font handle which would give visually
broken glyphs for me when the base font was smaller than the replacement font.
So, now instead rather than just changing the font for the fallback level, we
throw away the SalLayout using the old font, set the font for the fallback
level, and recreate a new SalLayout using the new font.
Hopefully this is the true reason behind the crashes of fdo#45355 although
that remains unreproducible for me.
Change-Id: I415b6e7a77777d4aa5e62079c52c6c18473c0ebd
|
|
Change-Id: Ib453048fdb0e858936903c8bf2a24c74932b1b85
|
|
Change-Id: Ifa1fa8d9e1f18163ba64edb8f26e1fcf7b21153e
|
|
Clearly we don't want to expose std::vector<Value>::insert here, and
neither e.g. push_back.
Change-Id: I89917a23d6d9f36f56474cdc361ba4d513516122
|
|
Change-Id: Ia6e4f2b4d1cfb1282dafd376ee56f5fa2e493b26
|
|
We use this kind of container a lot, so creating a single
implementation makes sense.
Change-Id: I67ead58becd7d2a287812145c11d93ab1c593c0f
|
|
Change-Id: I31783eecc28cdc6f4d8c40841302d5338a2cd7be
|
|
Change-Id: I94b2cccff386ccf331ca670df17d7a738e026879
|
|
the MSDffImportRecords now leaks memory, so use a boost::ptr_set
instead, that also doesn't seem to need a special comparator class.
Change-Id: I52935205a57cfb03f439d83401b80b1bf69ede47
|
|
Change-Id: I92d280f4e95a3d29851e73a955935b44966f5d03
|
|
Change-Id: I2ccd12b8e9516f59a39db7c9a123b2b76a0b6c28
|
|
Change-Id: I08959c0de113264aaa7b1ca8c4f3124b2177d30f
|
|
Change-Id: If9faa49b3d3fc36f36db4a7cc6ab82f8af54935a
|
|
Change-Id: Ib83ea7e5b4647f5335f7efe7bebf819fa6810864
|
|
Change-Id: I96f030fe55c40861faf249b1617606f0977f6295
|
|
Change-Id: I46f970fa23a0afd54485efb58b7f4f757c7d7752
|
|
Change-Id: I0e94f44c27bbcb9d1ab12ae4f3f00e89d5388d69
|
|
1) The values were not properly exclusive.
E.g. when compiling under OpenJDK 1.6, iscjava12 was set (!)
2) On Java 1.6, should not enable the parts that are (only) for Java7!
3) When non-JAVA7, need to preprocess the files that have JAVA7-only code, so that it gets removed
Change-Id: I99b6bd9e781fee1007a04a4dc3d3f952c1b2eb00
|
|
|
|
Change-Id: Ie07c6ac0471ecdf6236d194f1ec3389dd47d4b31
|
|
Change-Id: I65f681d9fd5f0e7471da77608279c2dade38d7cb
|
|
Change-Id: I15d8be58924918cde827ee1d12308cf491cff35b
|
|
Change-Id: I44633089c40a5a02655d4abb08fedd1acb014d15
|
|
Change-Id: I5a4f65da17062db1334bb0140b7b4545494adfa6
|
|
Change-Id: I83b18d86c61e38ebe344d0b4153513ad65b20720
|
|
Change-Id: I9bdf184f5a8f414144d9dea814d6179ec2134209
|
|
Change-Id: I967ce4765fdb85263eea4cea5a8728470b32ad82
|