aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeLines
...
* Merge branch 'benchmark'Daniel Eklöf2020-05-31-1/+99
|\
| * doc: benchmark: right-align columnsDaniel Eklöf2020-05-31-1/+3
| |
| * doc: benchmark: fix typoDaniel Eklöf2020-05-31-1/+1
| |
| * doc: benchmark: newlinesDaniel Eklöf2020-05-31-2/+4
| |
| * doc: benchmark: newlines (?)Daniel Eklöf2020-05-31-2/+2
| |
| * doc: benchmarks: shell highlighting of vtebench commandsDaniel Eklöf2020-05-31-0/+2
| |
| * readme: fix benchmark linkDaniel Eklöf2020-05-31-1/+1
| |
| * benchmarks: initial resultsDaniel Eklöf2020-05-31-1/+42
| |
| * scripts: benchmarking utility scriptDaniel Eklöf2020-05-31-0/+51
|/
* vt: utf8: use mbtowc() instead of mbrtowc()Daniel Eklöf2020-05-31-8/+2
| | | | | | | | | This is slightly faster, since we don't need to initialize an mbstate_t struct (using mbrtowc() with a NULL-pointer for 'ps' also works). Also, avoid a branch by setting wc=0 and then ignoring the result/error code from mbtowc().
* vt: utf8: don't need one entry action for each UTF8 variantDaniel Eklöf2020-05-31-20/+6
|
* readme: cleanupDaniel Eklöf2020-05-30-8/+2
|
* readme: GCC requires -Wno-missing-profile in a PGO buildDaniel Eklöf2020-05-30-3/+6
|
* readme: add link to "running the new build" at the end of the release build sectionDaniel Eklöf2020-05-30-0/+1
|
* readme: describe what should expect to see when running the benchmarkDaniel Eklöf2020-05-30-0/+3
|
* reaadme: add step to remove the temp fileDaniel Eklöf2020-05-30-0/+1
|
* readme: add missing ')'Daniel Eklöf2020-05-30-1/+1
|
* readme: fix URLDaniel Eklöf2020-05-30-1/+1
|
* readme: describe how to do a PGO buildDaniel Eklöf2020-05-30-10/+85
|
* render: resize: don't try to signal TIOCSWINSZ on a closed ptmx FDDaniel Eklöf2020-05-27-1/+1
|
* Merge branch 'hold-option-in-footclient'Daniel Eklöf2020-05-27-0/+22
|\ | | | | | | | | | | | | This adds --hold to footclient. It was an oversight on my part to not add it from the beginning, when it was added to foot. Closes #17
| * changelog: mention addition of '--hold' to footclientDaniel Eklöf2020-05-26-0/+1
| |
| * doc: footclient: add --holdDaniel Eklöf2020-05-26-0/+3
| |
| * completions: zsh: footclient: add --holdDaniel Eklöf2020-05-26-0/+1
| |
| * client/server: implement '--hold' in footclientDaniel Eklöf2020-05-26-0/+17
|/
* slave: set IUTF8Daniel Eklöf2020-05-26-0/+15
|
* readme: mention that -O3 is default in release buildsDaniel Eklöf2020-05-25-1/+3
|
* readme: document work around for GCC 10.1 O2+PGO regressionDaniel Eklöf2020-05-25-2/+5
|
* meson: don't set -Wno-missing-profileDaniel Eklöf2020-05-25-2/+1
| | | | We don't do PGO by default
* PKGBUILD: always set -fno-plt and -Wno-missing-profileDaniel Eklöf2020-05-25-1/+5
|
* PKGBUILD: append -O3 to makepkg provided CFLAGSDaniel Eklöf2020-05-25-0/+1
| | | | This works around a GCC 10.1 O2+PGO regression.
* readme: compared with -> compared toDaniel Eklöf2020-05-25-1/+1
|
* meson: -fprofile-correction is automatically added by mesonDaniel Eklöf2020-05-25-1/+0
|
* readme: spelling: gpo -> pgoDaniel Eklöf2020-05-25-1/+1
|
* terminfo: put 'cvvis' on a separate lineDaniel Eklöf2020-05-25-1/+2
|
* readme: mention GCC 10.1 performance regressionDaniel Eklöf2020-05-25-0/+3
|
* changelog: is -> areDaniel Eklöf2020-05-21-1/+1
|
* reaper: disable debug loggingDaniel Eklöf2020-05-21-1/+1
|
* slave: restore signal maskDaniel Eklöf2020-05-21-1/+5
|
* term: don't double fork new terminal windowsDaniel Eklöf2020-05-21-57/+59
| | | | | Instead, register their PIDs with the new reaper module and let it handle them.
* reaper: new module, uses a signalfd to wait() on child processesDaniel Eklöf2020-05-21-0/+161
| | | | | | | | Use a signalfd to listen for SIGCHLD signals. When we receive a SIGCHLD over the signalfd, reap all dead children by looping over all registered child PIDs and call waitpid(WNOHANG) on them.
* readme: mention that gcc produces a much faster binary than clangDaniel Eklöf2020-05-21-0/+4
|
* meson: add -fprofile-correctionDaniel Eklöf2020-05-21-0/+1
|
* scrolling: fix crash when offset was exactly at the wrap-aroundDaniel Eklöf2020-05-19-2/+5
| | | | | | | When making sure we don't scroll past the scrollback history, and the current offset was exactly at the wrap around, the new view port was set to offset+1, which in this particular case meant outside the valid row numbers.
* selection: performance: check for < 0 or >= 0 instead of == -1Daniel Eklöf2020-05-19-9/+9
|
* selection: handle viewport wrap around correctlyDaniel Eklöf2020-05-19-49/+64
| | | | | When the viewport wraps around, the selection points may be "incorrect".
* term: scrolling: calling function must clamp scroll amountDaniel Eklöf2020-05-19-8/+16
|
* readme: add suggested cflags for a release buildDaniel Eklöf2020-05-18-3/+10
|
* term: erase_cell_range: set row->dirty before calling memset()Daniel Eklöf2020-05-17-2/+2
|
* term: scrolling: mark selection top if-statement as 'unlikely'Daniel Eklöf2020-05-17-4/+4
|
* term: scrolling: hopefully fix all selection/scrolling related crashesDaniel Eklöf2020-05-17-17/+89
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When scrolling, there are a couple of cases where an existing selection must be canceled because we cannot meaningfully represent it after scrolling. These are when the selection is (partly) inside: * The top scrolling region * The bottom scrolling region * The new lines scrolled in. I.e. re-used lines For the scrolling regions, the real problem is when the selection crosses the scrolling region boundary; a selection that is completely inside a scrolling regions _might_ be possible to keep, but we would need to translate the selection coordinates to the new scrolling region lines. For simplicity, we cancel the selection if it touches the scrolling region. Period. The last item, newly scrolled in lines is when the selection covers very old lines and we're now wrapping around the scrollback history. Then there's a fourth problem case: when the user has started a selection, but hasn't yet moved the cursor. In this case, we have no end point. What's more problematic is that when the user (after scrolling) moves the cursor, we try to create a huge selection that covers mostly empty (NULL) rows, causing us to crash. This can happen e.g. when reverse scrolling in such a way that we wrap around the scrollback history. The actual viewport in this case is something like `-n - m`. But the selection we'll end up trying to create will be `m - (rows - n)`. This range may very well contain NULL rows. To deal with this, we simply cancel the selection.
* selection: on_rows_in_view: fix range checkDaniel Eklöf2020-05-17-7/+5
|
* term: scrolling: sixel: simplify now that row count is clampedDaniel Eklöf2020-05-17-2/+2
|
* term: scrolling: in debug builds, verify all rows are allocatedDaniel Eklöf2020-05-16-0/+10
|
* term: scrolling: simplify, now that scroll row count is clampedDaniel Eklöf2020-05-16-11/+5
|
* term: scrolling: clamp number of rows to number of rows in scrolling regionDaniel Eklöf2020-05-16-13/+6
|
* grid: swap_row: remove unused parameter 'initialize'Daniel Eklöf2020-05-16-6/+6
|
* selection: copy: insert line break if either cell is emptyDaniel Eklöf2020-05-16-1/+5
|
* selection/render: sync cells' 'selected' bit before renderingDaniel Eklöf2020-05-16-0/+31
| | | | | | | | | | | | | | | | | | | | | | | For performance reasons, we track whether a cell is selected or not using a bit in a cell's attributes. This makes it easy for the renderer to determine if the cells should be rendered as selected or not - it just have to look at the 'selected' bit instead of doing a complex range check against the current selection. This works nicely in most cases. But, if the cell is updated, the 'selected' bit is cleared. This results in the renderer rendering the cell normally, i.e. _not_ selected. Checking for this, and re-setting the 'selected' bit when the cell is updated (printed to) is way too expensive as it is in the hot path. Instead, sync the 'selected' bits just before rendering. This isn't so bad as it may sound; if there is no selection this is a no-op. Even if there is a selection, only those cells whose 'selected' bit have been cleared are dirtied (and thus re-rendered) - these cells would have been re-rendered anyway.
* term: scrolling: clear selection *before* scrollingDaniel Eklöf2020-05-16-12/+15
| | | | | | | | | | | | We normally don't clear the selection when scrolling. The exception is when the selection covers re-used rows. I.e. rows that we scroll in and clear. In this case we cancel the selection (we _could_ modify it and keep as much as possible and only remove the re-used rows...). We must do this *before* scrolling, since scrolling will swap rows (when there's a scrolling region). When this happens, the selection is "corrupted", and canceling it afterwards will not work.
* selection: update: don't assert on 'start' being setDaniel Eklöf2020-05-16-0/+1
|
* selection: don't automatically no-op operations if selection isn't enabledDaniel Eklöf2020-05-16-29/+25
| | | | | | | | | | | | | | | | | | When the client is capturing the mouse, selection can only be done by holding done shift. This is why a lot of selection functions are no-ops if selection isn't currently enabled. However, there are many cases where we actually need to modify the selection. In particular, selection_cancel(). Thus, only check for enabled selection when we're dealing with user input. Bonus: this also fixes a bug where an ongoing selection were finalized as soon as the user released shift, even if he was still holding down the mouse button.
* selection: on_row_in_view() -> on_rows_in_view(), and now takes a rangeDaniel Eklöf2020-05-16-15/+37
| | | | | This allows us to check the selection once when scrolling, instead of once for each scrolled in line.
* render: only apply transparency to the default background colorDaniel Eklöf2020-05-16-2/+9
|
* readme: add 'synchronized updates' to feature listDaniel Eklöf2020-05-16-0/+1
| | | | The support in foot isn't new, but let's make it more visible.
* terminfo: add tmux extension 'Sync' - indicates 'Synchronized Updates' supportDaniel Eklöf2020-05-16-0/+5
|
* changelog: prepare for the next, unreleased versionDaniel Eklöf2020-05-13-0/+11
|
* Merge branch 'releases/1.3'Daniel Eklöf2020-05-13-8/+5
|\
| * meson/PKGBUILD: bump version to 1.3.01.3.0releases/1.3Daniel Eklöf2020-05-13-2/+2
| |
| * changelog 1.3.0Daniel Eklöf2020-05-13-6/+3
|/
* slave: configure pts to be the controlling terminal of the forked processDaniel Eklöf2020-05-13-0/+8
|
* Merge branch 'fcft-precompose'Daniel Eklöf2020-05-13-33920/+37
|\
| * unicode-combine: simplify - remove -Dunicode-precompose optionDaniel Eklöf2020-05-10-44/+7
| | | | | | | | | | | | | | | | | | | | Since the pre-composing functionality is now part of fcft, it makes little sense to have a compile time option - there's no size benefit to be had. Furthermore, virtually all terminal emulators do pre-composing (alacritty being an exception), this really isn't that controversial.
| * readme: unicode-combining: mention when we render the decomposed versionDaniel Eklöf2020-05-09-0/+5
| |
| * meson/PKGBUILD: we now require fcft-2.1.0Daniel Eklöf2020-05-09-2/+2
| |
| * meson: explicitly define _POSIX_C_SOURCEDaniel Eklöf2020-05-09-1/+2
| |
| * unicode-precompose: use fcft's precompose functionalityDaniel Eklöf2020-05-09-33878/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | This allows us more options when determining whether to use a pre-composed character or not: We now only use the pre-composed character if it's from the primary font, or if at least one of the base or combining characters are from a fallback font. I.e. use glyphs from the primary font if possible. But, if one or more of the decomposed glyphs are from a fallback font, use the pre-composed character anyway.
* | wayland: force-trigger a resize on 'configured' eventDaniel Eklöf2020-05-12-1/+2
| | | | | | | | | | | | | | | | | | This ensures we _always_ commit a **new** buffer in response to a configured event. This fixes an issue in Gnome where e.g. tiling the window (on the left/right side) only worked if that caused the windows size to change.
* | readme: use unicode RIGHT ARROW instead of '->'Daniel Eklöf2020-05-10-1/+1
| |
* | readme: short section describing the origins of the name 'foot'Daniel Eklöf2020-05-10-0/+18
| |
* | generate-unicode-precompose: /bin/sh is more standard than /usr/bin/shDaniel Eklöf2020-05-09-2/+2
|/ | | | | | | At the same time, fix `cut` command to pass filename last, as this is also more portable. Fixes build issues on e.g. Alpine Linux
* Revert "term: print: line-wrap multi-column characters instead of cutting them in half"Daniel Eklöf2020-05-09-6/+4
| | | | | | | This reverts commit 8448296bb3c0fa96f5af5cb7ec563bb0c7e99093. Applications need to be multi-column character aware anyway, and they should deal with this.
* term: print: line-wrap multi-column characters instead of cutting them in halfDaniel Eklöf2020-05-09-4/+6
|
* completions: zsh: complete command/shell, and its argumentsDaniel Eklöf2020-05-08-4/+8
|
* main/client: be POSIXLY_CORRECT when parsing command lineDaniel Eklöf2020-05-08-12/+15
| | | | | | | | | | | | | | | | | | | | This means command line parsing stops when it encounters the first nonoption argument. The result is that one no longer need to use '--' to ensure arguments are passed to the shell/command, instead of parsed by foot. That is, instead of foot -- sh -c true one can now do foot sh -c true Arguments to foot *must* go before the command: foot --fullscreen sh -c true
* term: don't register ptmx with FDM if we're shutting downDaniel Eklöf2020-05-04-2/+4
| | | | | This can happen when there's an error instantiating the terminal *after* we've spawned the slave process.
* term: set 'is_shutting_down' before calling term_destroy() on errorDaniel Eklöf2020-05-04-0/+1
|
* csi: ignore invalid parameters in 'CSI Ps <space> q'Daniel Eklöf2020-05-04-4/+10
| | | | | Before this patch, an invalid parameter would not change the cursor style, but could cause the cursor to start (or stop) blinking.
* term: check return value when instantiating thread synchronization primitivesDaniel Eklöf2020-05-04-4/+25
|
* meson: use find_program() to find 'tic'Daniel Eklöf2020-05-03-1/+2
|
* term: thrd_success isn't necessarily 0Daniel Eklöf2020-05-03-3/+3
|
* util: thrd_err_as_string: fix alignmentDaniel Eklöf2020-05-03-1/+1
|
* thrd_create(): handle errors correctlyDaniel Eklöf2020-05-03-13/+31
| | | | | | * Detect thrd_create() failures in all places where we instantiate threads * Log error correctl; thrd_create() does not return an errno value
* util: thrd_err_as_string(): new function, returns a thrd_*() error stringDaniel Eklöf2020-05-03-0/+18
|
* Merge branch 'space-optimize-combining-chars'Daniel Eklöf2020-05-03-115/+137
|\
| * unicode-combining: limit maximum number of allowed composed chainsDaniel Eklöf2020-05-03-5/+13
| |
| * unicode-combining: detect when we've reached the chain limitDaniel Eklöf2020-05-03-0/+14
| | | | | | | | | | | | | | | | | | | | | | We currently store up to 5 combining characters in any given base+combining chain. This adds a check for when that limit is about to be exceeded. When this happens, we log the chain + the new combining character. Since things will break anyway, we simply overwrite the last combining character.
| * changelog: update 'unicode comining' entry - only parts of the feature are optionalDaniel Eklöf2020-05-03-2/+3
| |
| * readme: update 'Unicode combining'Daniel Eklöf2020-05-03-21/+15
| |
| * unicode-combining: store seen combining chains "globally" in the term structDaniel Eklöf2020-05-03-92/+97
|/ | | | | | | | | | | | | | | | | | | | Instead of storing combining data per cell, realize that most combinations are re-occurring and that there's lots of available space left in the unicode range, and store seen base+combining combinations chains in a per-terminal array. When we encounter a combining character, we first try to pre-compose, like before. If that fails, we then search for the current base+combining combo in the list of previously seen combinations. If not found there either, we allocate a new combo and add it to the list. Regardless, the result is an index into this array. We store this index, offsetted by COMB_CHARS_LO=0x40000000ul in the cell. When rendering, we need to check if the cell character is a plain character, or if it's a composed character (identified by checking if the cell character is >= COMB_CHARS_LO). Then we render the grapheme pretty much like before.