OSDN Git Service

android-x86/external-mksh.git
6 years agoMerge tag 'android-8.1.0_r1' into oreo-x86 oreo-x86 android-x86-8.1-r1 android-x86-8.1-r2 android-x86-8.1-r3 android-x86-8.1-r4 android-x86-8.1-r5 android-x86-8.1-r6
Chih-Wei Huang [Fri, 8 Dec 2017 08:37:08 +0000 (16:37 +0800)]
Merge tag 'android-8.1.0_r1' into oreo-x86

Android 8.1.0 Release 1 (OPM1.171019.011)

6 years agomkshrc: add back the aliases I like multiwindow-oreo
Chih-Wei Huang [Wed, 1 Jul 2015 16:07:20 +0000 (00:07 +0800)]
mkshrc: add back the aliases I like

6 years agorelease-request-1b0d6769-ebfa-4e2d-b128-039ac0dcd5c9-for-git_oc-mr1-release-4140700...
android-build-team Robot [Wed, 28 Jun 2017 07:31:17 +0000 (07:31 +0000)]
release-request-1b0d6769-ebfa-4e2d-b128-039ac0dcd5c9-for-git_oc-mr1-release-4140700 snap-temp-L81200000078275553

Change-Id: Ic9f8964a03622afd1b0387ea41a09e9e52fbd10f

6 years agoMerge "Build vendor shell unconditionally." am: bba8f4a328 am: 8547d5a6a9 am: 83c29a0b24
padarshr [Tue, 27 Jun 2017 19:26:00 +0000 (19:26 +0000)]
Merge "Build vendor shell unconditionally." am: bba8f4a328 am: 8547d5a6a9 am: 83c29a0b24
am: 0a685140c5

Change-Id: I0960b248ffc238f017a7fce637866b0f08805dfb

6 years agoMerge "Build vendor shell unconditionally." am: bba8f4a328 am: 8547d5a6a9
padarshr [Tue, 27 Jun 2017 19:03:15 +0000 (19:03 +0000)]
Merge "Build vendor shell unconditionally." am: bba8f4a328 am: 8547d5a6a9
am: 83c29a0b24

Change-Id: Ia67e60e4389d9a21efb9c4a0d508ad71751a021e

6 years agoMerge "Build vendor shell unconditionally." am: bba8f4a328
padarshr [Tue, 27 Jun 2017 19:00:16 +0000 (19:00 +0000)]
Merge "Build vendor shell unconditionally." am: bba8f4a328
am: 8547d5a6a9

Change-Id: I338f80a7e8bb5132988c9030ad70a4cb4c05ee75

6 years agoMerge "Build vendor shell unconditionally."
padarshr [Tue, 27 Jun 2017 18:57:15 +0000 (18:57 +0000)]
Merge "Build vendor shell unconditionally."
am: bba8f4a328

Change-Id: Icf13f5feb51106fd9b373517ae082018dfe7bab9

6 years agoMerge "Build vendor shell unconditionally."
Treehugger Robot [Tue, 27 Jun 2017 18:47:30 +0000 (18:47 +0000)]
Merge "Build vendor shell unconditionally."

6 years agorelease-request-7bfcab52-d1c0-4256-9d6b-5b5092bc78ca-for-git_oc-mr1-release-4133428...
android-build-team Robot [Sun, 25 Jun 2017 07:45:17 +0000 (07:45 +0000)]
release-request-7bfcab52-d1c0-4256-9d6b-5b5092bc78ca-for-git_oc-mr1-release-4133428 snap-temp-L95800000077479875

Change-Id: I57569c684f9cc36d06050223f2a99202a8aa4ad6

6 years agoBuild vendor shell unconditionally.
padarshr [Fri, 23 Jun 2017 17:48:38 +0000 (23:18 +0530)]
Build vendor shell unconditionally.

Always build vendor shell irrespective of whether
PRODUCT_FULL_TREBLE is enabled or not, as even on Non-Full
treble builds this can be useful to start vendor scripts.

Change-Id: I93e514760cfcfd4f7f4ce401eae50d839638c51c

6 years agosh_vendor link to libc as a shared library. am: d79560268f -s ours
Yifan Hong [Thu, 22 Jun 2017 22:56:58 +0000 (22:56 +0000)]
sh_vendor link to libc as a shared library. am: d79560268f  -s ours
am: a0f2ab0af0  -s ours

Change-Id: I7873684ea7e736a3cf53afdac1bedb4107917be5

6 years agosh_vendor link to libc as a shared library.
Yifan Hong [Thu, 22 Jun 2017 22:44:26 +0000 (22:44 +0000)]
sh_vendor link to libc as a shared library.
am: d79560268f  -s ours

Change-Id: Ifdcb08d50c43d406291a007ab687ec9207dc579f

6 years agoMerge "sh_vendor link to libc as a shared library." am: 02aa26d9ab am: 17d120feef...
Yifan Hong [Thu, 22 Jun 2017 19:29:42 +0000 (19:29 +0000)]
Merge "sh_vendor link to libc as a shared library." am: 02aa26d9ab am: 17d120feef am: 140506d9d2
am: cc5acf1455

Change-Id: I762a63f820b4d2a566b64921ce6029e3a6ee39f5

6 years agoMerge "sh_vendor link to libc as a shared library." am: 02aa26d9ab am: 17d120feef
Yifan Hong [Thu, 22 Jun 2017 19:26:12 +0000 (19:26 +0000)]
Merge "sh_vendor link to libc as a shared library." am: 02aa26d9ab am: 17d120feef
am: 140506d9d2

Change-Id: I7ea4311f215fe3837119e7c06e5fc6e6c1659266

6 years agoMerge "sh_vendor link to libc as a shared library." am: 02aa26d9ab
Yifan Hong [Thu, 22 Jun 2017 19:22:47 +0000 (19:22 +0000)]
Merge "sh_vendor link to libc as a shared library." am: 02aa26d9ab
am: 17d120feef

Change-Id: I690fa32795488aa9678dcce1bc4fb0df6f92ff08

6 years agoMerge "sh_vendor link to libc as a shared library."
Yifan Hong [Thu, 22 Jun 2017 19:18:11 +0000 (19:18 +0000)]
Merge "sh_vendor link to libc as a shared library."
am: 02aa26d9ab

Change-Id: If97c273883223e2a516bd788df2d58f25681666c

6 years agoMerge "sh_vendor link to libc as a shared library."
Treehugger Robot [Thu, 22 Jun 2017 19:14:22 +0000 (19:14 +0000)]
Merge "sh_vendor link to libc as a shared library."

6 years agosh_vendor link to libc as a shared library.
Yifan Hong [Thu, 22 Jun 2017 00:05:15 +0000 (17:05 -0700)]
sh_vendor link to libc as a shared library.

Test: BOARD_VNDK_VERSION m -j -k
Test: m -j -k and boots

Bug: 33241851
Change-Id: Ica98c551892ff83b7e9e661d2ddfa4bac7d2c6f4
Merged-In: Ica98c551892ff83b7e9e661d2ddfa4bac7d2c6f4

6 years agosh_vendor link to libc as a shared library.
Yifan Hong [Thu, 22 Jun 2017 00:05:15 +0000 (17:05 -0700)]
sh_vendor link to libc as a shared library.

Test: BOARD_VNDK_VERSION m -j -k
Test: m -j -k and boots

Bug: 33241851
Change-Id: Ica98c551892ff83b7e9e661d2ddfa4bac7d2c6f4
Merged-In: Ica98c551892ff83b7e9e661d2ddfa4bac7d2c6f4

6 years agoMerge "install copy of shell in vendor partition" am: ee224ea092 am: 5a6144a098 ...
Sandeep Patil [Wed, 21 Jun 2017 23:10:42 +0000 (23:10 +0000)]
Merge "install copy of shell in vendor partition" am: ee224ea092 am: 5a6144a098  -s ours am: 632282bc91  -s ours
am: 1942fd0335  -s ours

Change-Id: I90596692e63df5f814d3da3f1ed38af4764bc85f

6 years agoMerge "install copy of shell in vendor partition" am: ee224ea092 am: 5a6144a098 ...
Sandeep Patil [Wed, 21 Jun 2017 23:07:41 +0000 (23:07 +0000)]
Merge "install copy of shell in vendor partition" am: ee224ea092 am: 5a6144a098  -s ours
am: 632282bc91  -s ours

Change-Id: Id61c8fccafce31b17a9c90d506db05bb64b9c7c2

6 years agoMerge "install copy of shell in vendor partition" am: ee224ea092
Sandeep Patil [Wed, 21 Jun 2017 23:04:59 +0000 (23:04 +0000)]
Merge "install copy of shell in vendor partition" am: ee224ea092
am: 5a6144a098  -s ours

Change-Id: I05b9d8ad63a5b69fec0c7a4dc768d1fcade3a8a6

6 years agoMerge "install copy of shell in vendor partition"
Sandeep Patil [Wed, 21 Jun 2017 23:01:42 +0000 (23:01 +0000)]
Merge "install copy of shell in vendor partition"
am: ee224ea092

Change-Id: Id919730d78760daefef01bba45b64556658cf8a5

6 years agoMerge "install copy of shell in vendor partition"
Sandeep Patil [Wed, 21 Jun 2017 22:58:32 +0000 (22:58 +0000)]
Merge "install copy of shell in vendor partition"

6 years agoinstall copy of shell in vendor partition
Sandeep Patil [Thu, 13 Apr 2017 01:02:50 +0000 (18:02 -0700)]
install copy of shell in vendor partition

This is to be used by vendor processes. Particularly, vendor shell
scripts that get launched as init services. The shell is restricted from
running as interactive mode in SEPolicy. Unless it is launched from rooted
system shell. The change also adds the ability to override the default
$PATH regardless of the environment.

Bug: 36463595
Test: Boot sailfish with new shell and ensure all init services exit
      with status 0

Merged-In: I1eaf2c8a8f6905caf00d53e067add67eaea00e09
Change-Id: I1eaf2c8a8f6905caf00d53e067add67eaea00e09
Signed-off-by: Sandeep Patil <sspatil@google.com>
(cherry picked from commit c2036115102d80c026b70846812a42204522f9ac)

7 years agomerge in oc-release history after reset to oc-dev
gitbuildkicker [Sun, 16 Apr 2017 07:08:17 +0000 (00:08 -0700)]
merge in oc-release history after reset to oc-dev

7 years agoinstall copy of shell in vendor partition am: c203611510
Sandeep Patil [Fri, 14 Apr 2017 21:03:40 +0000 (21:03 +0000)]
install copy of shell in vendor partition am: c203611510
am: 5bfe1ea22e

Change-Id: I7040d3e62b6709013fdb3798ebff318c4693d40c

7 years agoinstall copy of shell in vendor partition
Sandeep Patil [Fri, 14 Apr 2017 21:01:07 +0000 (21:01 +0000)]
install copy of shell in vendor partition
am: c203611510

Change-Id: I6a1da672c65522768647550df8768f9cd3285b23

7 years agoMerge "Upgrade to mksh R55." am: 693c2ea1e4 am: e0d2df7f07 am: 0d2db5fba1
Elliott Hughes [Fri, 14 Apr 2017 06:59:17 +0000 (06:59 +0000)]
Merge "Upgrade to mksh R55." am: 693c2ea1e4 am: e0d2df7f07 am: 0d2db5fba1
am: 651a7bb8b8

Change-Id: I1137b9d86533f909f850ba2ec4b6b0a98a4c29a0

7 years agoMerge "Upgrade to mksh R55." am: 693c2ea1e4 am: e0d2df7f07
Elliott Hughes [Fri, 14 Apr 2017 06:07:26 +0000 (06:07 +0000)]
Merge "Upgrade to mksh R55." am: 693c2ea1e4 am: e0d2df7f07
am: 0d2db5fba1

Change-Id: I14e66c15ea92f0ae7b282245bb07f911dfccf935

7 years agoMerge "Upgrade to mksh R55." am: 693c2ea1e4
Elliott Hughes [Fri, 14 Apr 2017 06:00:07 +0000 (06:00 +0000)]
Merge "Upgrade to mksh R55." am: 693c2ea1e4
am: e0d2df7f07

Change-Id: I2e1ef7d4911a211b3dc58e3cab8e6c6e0ad2778a

7 years agoMerge "Upgrade to mksh R55."
Elliott Hughes [Fri, 14 Apr 2017 05:53:42 +0000 (05:53 +0000)]
Merge "Upgrade to mksh R55."
am: 693c2ea1e4

Change-Id: If722a24bb137ba653d1374814397c705f3b10af2

7 years agoMerge "Upgrade to mksh R55."
Elliott Hughes [Fri, 14 Apr 2017 05:43:06 +0000 (05:43 +0000)]
Merge "Upgrade to mksh R55."

7 years agoinstall copy of shell in vendor partition
Sandeep Patil [Thu, 13 Apr 2017 01:02:50 +0000 (18:02 -0700)]
install copy of shell in vendor partition

This is to be used by vendor processes. Particularly, vendor shell
scripts that get launched as init services. The shell is restricted from
running as interactive mode in SEPolicy. Unless it is launched from rooted
system shell. The change also adds the ability to override the default
$PATH regardless of the environment.

Bug: 36463595
Test: Boot sailfish with new shell and ensure all init services exit
      with status 0

Change-Id: I1eaf2c8a8f6905caf00d53e067add67eaea00e09
Signed-off-by: Sandeep Patil <sspatil@google.com>
7 years agoUpgrade to mksh R55.
Elliott Hughes [Wed, 12 Apr 2017 23:52:30 +0000 (16:52 -0700)]
Upgrade to mksh R55.

R55 is mostly a feature release with summary bugfixes:

[komh] Fix OS/2 search_access() and UNC path logic
[tg] Undocument printf(1) to avoid user confusion
[Jean Delvare, tg] Fix printf builtin -R option
[tg] Make ${var@x}, unknown x, fail (thanks izabera)
[tg] ${var=x} must evaluate x in scalar context (10x Martijn Dekker)
[tg] Fixup relation between lksh and mksh, reduce delta
[tg] Improve manpage display; add OS/2 $PATH FAQ
[Jean Delvare] Fix bugs in manpage
[tg] Review tilde expansion, removing “odd use of KEEPASN” and introduce POSIX “declaration utility” concept; wait isn’t one
[tg] Add \builtin utility, declaration utility forwarder
[tg] Make $'\xz' expand to xz, not \0
[tg] Use fixed string pooling (requires the above change in host mksh)
[tg] POSIX declaration commands can have varassign and redirections
[Martijn Dekker] Add typeset -g, replacing homegrown “global”
[Harvey-OS] Disable NOPROSPECTOFWORK, APEX is reportedly fixed now
[tg] Display ulimit -a output with flags; improve Haiku
[tg] Drop old let] hack, use \builtin internally
[tg] Fix padding in Lb64encode in dot.mkshrc
[tg] Move FAQ content to a separate, new FAQ section in the manpage
[tg] Add new standard variable PATHSEP (‘:’, ‘;’ on OS/2)
[Martijn Dekker] Fix LINENO in eval and alias
[komh] Fix “\builtin” on OS/2
[tg] Improve (internal) character classes code for speed
[tg] Fix: the underscore is no drive letter
[tg] No longer hard-disable persistent history support in lksh
[tg] Introduce build flag -T for enabling “textmode” on OS/2 (supporting CR+LF line endings, but incompatible with mksh proper)
[tg] Merge mksh-os2
[tg] Permit changing $OS2_SHELL during a running shell
[tg] Fix multibyte handling in ^R (Emacs search-history)
[tg] Allow “typeset -p arrname[2]” to work
[tg] Make some error messages more consistent
[tg, komh] Disable UTF-8 detection code for OS/2 as unrealistic
[tg, sdaoden] Limit alias name chars to POSIX plus non-leading ‘-’
[tg, Martijn Dekker] Expand aliases at COMSUB parse time
[tg] Make “typeset -f” output alias-resistent
[tg, Martijn Dekker] Permit “eval break” and “eval continue”
[tg] Make -masm=intel safe on i386
[tg] Disambiguate $((…)) vs. $((…)…) in “typeset -f” output
[Jean Delvare] Clarify the effect of exit and return in a subshell
[tg] Simplify compile-time asserts and make them actually compile-time
[tg] Fix ^O in Emacs mode if the line was modified (LP#1675842)
[tg] Address Coverity Scan… stuff… now that it builds again
[Martijn Dekker, tg] Add test -v
[tg] Document set -o posix/sh completely

Bug: N/A
Test: manual
Change-Id: Ifce1d879933a5773e98b4f34f4a9bb86a6bdff3b

7 years agoMerge "Restore the "print" builtin." am: ad7a062c12 am: be32749ec0 am: 57198a7ef7
Elliott Hughes [Thu, 22 Dec 2016 01:49:59 +0000 (01:49 +0000)]
Merge "Restore the "print" builtin." am: ad7a062c12 am: be32749ec0 am: 57198a7ef7
am: 2ad3638506

Change-Id: I56028e1c1930f472e39a6d2b6097d8bff89df144

7 years agoMerge "Restore the "print" builtin." am: ad7a062c12 am: be32749ec0
Elliott Hughes [Thu, 22 Dec 2016 01:44:58 +0000 (01:44 +0000)]
Merge "Restore the "print" builtin." am: ad7a062c12 am: be32749ec0
am: 57198a7ef7

Change-Id: Ia1ce7e6b20ffe477b92220e515d6922c8cd7eb9e

7 years agoMerge "Restore the "print" builtin." am: ad7a062c12
Elliott Hughes [Thu, 22 Dec 2016 01:39:28 +0000 (01:39 +0000)]
Merge "Restore the "print" builtin." am: ad7a062c12
am: be32749ec0

Change-Id: I2632c14de0896ff3d43b8af4b8fa6c5d978d8215

7 years agoRestore the "print" builtin.
Elliott Hughes [Wed, 21 Dec 2016 18:59:39 +0000 (10:59 -0800)]
Restore the "print" builtin.

otapreopt_script is currently using this. We should probably move it onto
the more portable echo or printf, but for now I'm assuming there's a reason
why print is being used --- the use of -u implies isn't not just a typo for
printf.

(Even if we can switch to echo or printf, I'm tempted to just leave print be
anyway, since unlike the other disabled mksh builtins, it's not shadowing a
better implementation, so it's not causing confusion about why the "same"
command in different contexts behaves differently. The worst it will do is
cause portability problems if we move away from mksh, and in the meantime
print is documented on the mksh man page for users sophisticated enough to
know that we're using mksh, so it's arguably more surprising for it to be
missing.)

Bug: http://b/33761687
Test: manual
Change-Id: I440d759a3c4ef6fa24fd496ea832093e18b7060e
(cherry picked from commit 3d9915284b3d2b49b32be3261991f19f77345f2e)

7 years agoMerge "Restore the "print" builtin."
Elliott Hughes [Thu, 22 Dec 2016 01:33:27 +0000 (01:33 +0000)]
Merge "Restore the "print" builtin."
am: ad7a062c12

Change-Id: Ie3efc027233d111d15546e70da7d94c9798b752d

7 years agoMerge "Restore the "print" builtin."
Treehugger Robot [Thu, 22 Dec 2016 01:25:53 +0000 (01:25 +0000)]
Merge "Restore the "print" builtin."

7 years agoRestore the "print" builtin.
Elliott Hughes [Wed, 21 Dec 2016 18:59:39 +0000 (10:59 -0800)]
Restore the "print" builtin.

otapreopt_script is currently using this. We should probably move it onto
the more portable echo or printf, but for now I'm assuming there's a reason
why print is being used --- the use of -u implies isn't not just a typo for
printf.

(Even if we can switch to echo or printf, I'm tempted to just leave print be
anyway, since unlike the other disabled mksh builtins, it's not shadowing a
better implementation, so it's not causing confusion about why the "same"
command in different contexts behaves differently. The worst it will do is
cause portability problems if we move away from mksh, and in the meantime
print is documented on the mksh man page for users sophisticated enough to
know that we're using mksh, so it's arguably more surprising for it to be
missing.)

Bug: http://b/33761687
Test: manual
Change-Id: I440d759a3c4ef6fa24fd496ea832093e18b7060e

7 years agoMerge "Upgrade to mksh R40." am: ceb0b57de2 am: caf111e450 am: 7fba070015
Elliott Hughes [Fri, 9 Dec 2016 02:25:18 +0000 (02:25 +0000)]
Merge "Upgrade to mksh R40." am: ceb0b57de2 am: caf111e450 am: 7fba070015
am: 6e9f7fbc73

Change-Id: Ic9e97307012402b0c0f4b5565d1849400dbccc56

7 years agoMerge "Upgrade to mksh R40." am: ceb0b57de2 am: caf111e450
Elliott Hughes [Fri, 9 Dec 2016 02:20:36 +0000 (02:20 +0000)]
Merge "Upgrade to mksh R40." am: ceb0b57de2 am: caf111e450
am: 7fba070015

Change-Id: I880307b7a52a252c9d7f56b8a6be684da6f6b471

7 years agoMerge "Upgrade to mksh R40." am: ceb0b57de2
Elliott Hughes [Fri, 9 Dec 2016 02:15:36 +0000 (02:15 +0000)]
Merge "Upgrade to mksh R40." am: ceb0b57de2
am: caf111e450

Change-Id: I1c004b01a61d85022defd2c461a12c5db8778307

7 years agoMerge "Upgrade to mksh R40."
Elliott Hughes [Fri, 9 Dec 2016 02:10:35 +0000 (02:10 +0000)]
Merge "Upgrade to mksh R40."
am: ceb0b57de2

Change-Id: Ib164451b87d9653f15c276c8ecd441a05ff6253e

7 years agoMerge "Upgrade to mksh R40."
Treehugger Robot [Fri, 9 Dec 2016 02:06:25 +0000 (02:06 +0000)]
Merge "Upgrade to mksh R40."

7 years agoUpgrade to mksh R40.
Elliott Hughes [Thu, 8 Dec 2016 23:56:04 +0000 (15:56 -0800)]
Upgrade to mksh R40.

From the release notes...

R54 is a bugfix release with moderate new features:

[tg] Simplify and improve code and manual page
[tg] Try GCC 5’s new -malign-data=abi
[tg] Allow interrupting builtin cat even on fast devices (LP#1616692)
[tg] Update to Unicode 9.0.0
[Andreas Buschka] Correct English spelling
[tg] Handle set -e-related error propagation in || and && constructs correctly
[tg] Initialise memory for RNG even when not targeting Valgrind
[tg] Shrink binary size
[Brian Callahan] Improve support for the contemporary pcc compiler
[tg] Fix side effects with lazy evaluation; spotted by ormaaj
[tg] New flags -c (columnise), -l, -N for the print builtin
[Larry Hynes] Fix English, spelling mistakes, typos in the manpage
[tg, ormaah] Return 128+SIGALRM if read -t times out, like GNU bash
[Martijn Dekker] Install both manpages from Build.sh
[Martijn Dekker] Document case changes are ASCII-only
[Ronald G. Minnich, Elbing Miss, Álvaro Jurado, tg] Begin porting to Harvey-OS and APEX (similar to Plan 9 and APE)
[KO Myung-Hun] More infrastructure for the OS/2 (EMX, KLIBC) port

I've continued to disable the cat, realpath, and sleep builtins,
and added print to the list (now I've realized it's unrelated to
the print on the desktop).

Bug: N/A
Test: manual
Change-Id: Iead4c1b572c6f288b2ab2b884125dcb606f1e356

7 years agoMerge "Restore some of the mksh builtins." am: cd79a34d28 am: b48d5773d8 am: 10787a6eee
Elliott Hughes [Sat, 27 Aug 2016 03:20:39 +0000 (03:20 +0000)]
Merge "Restore some of the mksh builtins." am: cd79a34d28 am: b48d5773d8 am: 10787a6eee
am: ff0f07eb15

Change-Id: Ie3e81fdba2724ab3a1551819c2a34f1414267b87

7 years agoMerge "Restore some of the mksh builtins." am: cd79a34d28 am: b48d5773d8
Elliott Hughes [Sat, 27 Aug 2016 02:50:44 +0000 (02:50 +0000)]
Merge "Restore some of the mksh builtins." am: cd79a34d28 am: b48d5773d8
am: 10787a6eee

Change-Id: I49b1c553ffc8749c9c3e15e73cf036b08a3d18dd

7 years agoMerge "Restore some of the mksh builtins." am: cd79a34d28
Elliott Hughes [Sat, 27 Aug 2016 02:47:14 +0000 (02:47 +0000)]
Merge "Restore some of the mksh builtins." am: cd79a34d28
am: b48d5773d8

Change-Id: I8c6a2f2b317fda8699b33990eb33af61d8e48211

7 years agoMerge "Restore some of the mksh builtins."
Elliott Hughes [Sat, 27 Aug 2016 02:44:16 +0000 (02:44 +0000)]
Merge "Restore some of the mksh builtins."
am: cd79a34d28

Change-Id: I4916f190dc300a629f7266d0409d34a8bbfd3bf4

7 years agoMerge "Restore some of the mksh builtins."
Treehugger Robot [Sat, 27 Aug 2016 02:36:56 +0000 (02:36 +0000)]
Merge "Restore some of the mksh builtins."

7 years agoRestore some of the mksh builtins.
Elliott Hughes [Sat, 27 Aug 2016 01:10:18 +0000 (18:10 -0700)]
Restore some of the mksh builtins.

This makes us match bash in terms of what is/isn't a builtin.

Bug: http://b/20720875
Change-Id: Id90aeb115cf5e5217084bc1c5f382e14033b7d49

7 years agoMerge "Disable various mksh builtins." am: bf92a24781 am: 28c041f5a3 am: b894dff617
Elliott Hughes [Thu, 25 Aug 2016 20:16:55 +0000 (20:16 +0000)]
Merge "Disable various mksh builtins." am: bf92a24781 am: 28c041f5a3 am: b894dff617
am: 93d7921fa0

Change-Id: I6a1e5bb6c469d96a31a42f16969e264ecf4a012c

7 years agoMerge "Disable various mksh builtins." am: bf92a24781 am: 28c041f5a3
Elliott Hughes [Thu, 25 Aug 2016 19:56:30 +0000 (19:56 +0000)]
Merge "Disable various mksh builtins." am: bf92a24781 am: 28c041f5a3
am: b894dff617

Change-Id: Ia9862beb35b52782b4e77d48bac80d038e500ad7

7 years agoMerge "Disable various mksh builtins." am: bf92a24781
Elliott Hughes [Thu, 25 Aug 2016 19:53:29 +0000 (19:53 +0000)]
Merge "Disable various mksh builtins." am: bf92a24781
am: 28c041f5a3

Change-Id: I4ba944ff459b419d316f0a97f901fa1290428036

7 years agoMerge "Disable various mksh builtins."
Elliott Hughes [Thu, 25 Aug 2016 19:50:31 +0000 (19:50 +0000)]
Merge "Disable various mksh builtins."
am: bf92a24781

Change-Id: I8de5f26b5b9c7fe338c9402bc5b39eb00bee90a7

7 years agoMerge "Disable various mksh builtins."
Treehugger Robot [Thu, 25 Aug 2016 19:41:32 +0000 (19:41 +0000)]
Merge "Disable various mksh builtins."

7 years agoDisable various mksh builtins.
Elliott Hughes [Thu, 25 Aug 2016 00:18:10 +0000 (17:18 -0700)]
Disable various mksh builtins.

They're buggy and/or lacking in functionality. Disable them and use toybox
instead.

Bug: http://b/27064311
Bug: http://b/20720875
Change-Id: Ie57efffd7b98f172eafe9d4bd5ddf723811b6f29

7 years agoMerge "Upgrade to mksh R53a." am: c19d20531c am: 8b302a0123 am: c8ed698e56
Elliott Hughes [Sat, 13 Aug 2016 18:37:04 +0000 (18:37 +0000)]
Merge "Upgrade to mksh R53a." am: c19d20531c am: 8b302a0123 am: c8ed698e56
am: 0ff4e75810

Change-Id: I6dfd2d1ae1cde2c5d9d556ba9d18904bf9794dc0

7 years agoMerge "Upgrade to mksh R53a." am: c19d20531c am: 8b302a0123
Elliott Hughes [Sat, 13 Aug 2016 18:32:27 +0000 (18:32 +0000)]
Merge "Upgrade to mksh R53a." am: c19d20531c am: 8b302a0123
am: c8ed698e56

Change-Id: Iabb58f30fb722bf021f8f164780a8422403299de

7 years agoMerge "Upgrade to mksh R53a." am: c19d20531c
Elliott Hughes [Sat, 13 Aug 2016 18:28:26 +0000 (18:28 +0000)]
Merge "Upgrade to mksh R53a." am: c19d20531c
am: 8b302a0123

Change-Id: I7316be74d3aeb694790612601aef692d3663545e

7 years agoMerge "Upgrade to mksh R53a."
Elliott Hughes [Sat, 13 Aug 2016 18:25:00 +0000 (18:25 +0000)]
Merge "Upgrade to mksh R53a."
am: c19d20531c

Change-Id: I999ef9dd8ef426549886c70ea8757ff3b89c28c5

7 years agoMerge "Upgrade to mksh R53a."
Elliott Hughes [Sat, 13 Aug 2016 18:19:55 +0000 (18:19 +0000)]
Merge "Upgrade to mksh R53a."

7 years agoUpgrade to mksh R53a.
Elliott Hughes [Fri, 12 Aug 2016 22:06:53 +0000 (15:06 -0700)]
Upgrade to mksh R53a.

Note that we skipped R52c because it wouldn't build with clang (thanks to
the "Shave 200 bytes off .text by revisiting string pooling" change).

From the release notes:

R53a is a snapshot/feature release:

[lintian] Fix spelling
[tg] Unbreak multi-line command history broken by history flush
[tg] Fix redefining POSIX functions that were Korn functions before
[tg, TNF] Fix bounds checks in Vi editing mode
[tg] Handle combining characters at end of string or output correctly
[tg] Fix ${!#} ${!?} ${!-} (POSIX, prompted by izabera)
[tg] Fix shf.c-internal buffer overread on printing digits
[J�rg] Fix a typo in the testsuite
[arekm] Increase default edit line size (unless MKSH_SMALL)
[tg] Improve description of Emacs mode keybindings, especially ^U
[tg, arekm, jilles] Abort read builtin in case of read(2) errors
[tg, izabera, carstenh] Fix most of the ambiguous corner cases related to ${[pfx]var[op[word]]} (${@:-1} still unsupported)
[carstenh] Contribute some more testsuite coverage
[tg] WDS_TPUTS now emits QCHAR newline reentrant-safe
[tg] Fix var=<< implementation (LP#1380389)
[tg, FreeBSD] Make XSI test(1) extensions behave as if they were POSIX
[tg, izabera] Add $(<<<x) and $(<<EOF…) implementation
[tg] Lower minimum screen size accepted as “sane” from the OS to 4×2
[tg, Torsten Sillke] Simplify tilde-expanded parameters
[tg, Torsten Sillke] Fix default PS1 for substring matches
[tg] Apply defer-builtin-with-arguments logic to realpath builtin
[tg] Rework string pooling (own vs. compiler’s) (LP#1580348)
[tg] Feature: print -A, prints arguments as characters
[tg, izabera] Replace <<< and >>> as ROL and ROR operators with their new ^< and ^> spelling as per this proposal
[tg, slagtc] Clear-to-EOL under tmux to work around its anti-feature
[tg, p120ph37] Remove support for using file descriptors with more than a single digit, in preparation for named file descriptors
[tg] Correct, but simplify (at the potential cost of more tty I/O than strictly necessary, though never redundant and (probably) not more than before when it was miscalculated), line clearing and redrawing
[slagtc, tg] Implement new evaluate-region editing command Esc+Ctrl-E
[tg] Prefer external rename utility over the recovery builtin
[tg] Remove redundant full-line redraws
[tg, Natureshadow] Fix errorlevel of ‘.’ (“dot” special builtin) when the sourced script does not run any commands, for POSIX compliance
[tg] Refactor op tokens and edchars to shave off some more bytes
[tg] Fix some bugs in the manpage and some occasional/minor code bugs
[tg, Brian Callahan] Mark tests requiring new perl as !need-pass
[tg, slagtc] Add $KSH_MATCH and, to make it usable, ${foo@/bar/baz}
[tg, Score_Under] Fix bogus patch from OpenBSD: only NULL the global source in unwind when actually reclaiming its Area
[izabera] Mention in the manpage that integer bases go up to 36
[Natureshadow] Fix /= operator broken during refactoring

R52c is a bugfix-only release:

[tg] Shave 200 bytes off .text by revisiting string pooling
[tg, J�rg] Fix manpage for ditroff on Schillix
[tg, wbx] Use sed 1q instead of unportable head(1)
[tg] Implement underrun debugging tool for area-based memory allocator
[tg] Fix history underrun when first interactive command is entered
[tg, bef0rd] Do not misinterpret “${0/}” as “${0//”, fixes segfault
[tg, Stéphane Chazelas] Fix display problems with special parameters
[tg, Stéphane Chazelas] Catch attempt to trim $* and $@ with ?, fixes segfault (Todd Miller did this in 2004 for ${x[*]} already, so just sync)
[Martijn Dekker] Fix “command -p” with -Vv to behave as POSIX requires
[tg, jilles, Oleg Bulatov] Fix recusive parser with active heredocs
[tg] Flush even syntax-failing or interrupted commands to history
[tg, fmunozs] Fix invalid memory access for “'\0'” in arithmetics
[tg] Explicitly reserve SIGEXIT and SIGERR for ksh
[tg, izabera] Catch missing here documents at EOF even under “set -n”
[kre, tg] Document Austin#1015 handling (not considered a violation)
[tg, fmunozs] Fix buffer overread for empty nameref targets
[tg] Fix warnings pointed out by latest Debian gcc-snapshot
[tg, Martijn Dekker] Document upcoming set +o changes
[Martijn Dekker] Expand testsuite for command/whence

Change-Id: Ife475b25acae5a85277757cfe9f14c6575eac0cd

8 years agoMerge "Upgrade to mksh R52b."
Elliott Hughes [Wed, 3 Feb 2016 01:14:56 +0000 (01:14 +0000)]
Merge "Upgrade to mksh R52b."

8 years agoUpgrade to mksh R52b.
Elliott Hughes [Tue, 2 Feb 2016 23:26:47 +0000 (15:26 -0800)]
Upgrade to mksh R52b.

From the release notes:

R52b is a strongly recommended bugfix-only release:

[tg] Recognise ksh93 compiled scripts and LZIP compressed files as binary (i.e. to not run as mksh plaintext script)
[tg] Document that we will implement locale tracking later
[tg] Add EEXIST to failback strerror(3)
[jilles] Make set -C; :>foo race-free
[tg] Don’t use unset in portable build script
[tg] Plug warning on GNU/kFreeBSD, GNU/Hurd
[tg] Document read -a resets the integer base
[J�rg] Fix manpage: time is not a builtin but a reserved word
[J�rg, tg] Make exit (and return) eat -1
[tg] parse “$( (( … ) … ) … )” correctly (LP#1532621), Jan Palus
[tg] reduce memory footprint by free(3)ing more aggressively
[tg] fix buffer overrun (LP#1533394), bugreport by izabera
[tg] correctly handle nested ADELIM parsing (LP#1453827), Teckids
[tg] permit “read -A/-a arr[idx]” as long as only one element is read; fix corruption of array indicēs with this construct (LP#1533396), izabera
[tg] Sanitise OS-provided signal number in even more places
[tg] As requested by J�rg, be clear manpage advice is for mksh
[tg] Revert (as it was a regression) POSIX bugfix from R52/2005 related to accent gravis-style command substitution until POSIX decides either way
[tg] Handle export et al. after command (Austin#351)
[tg] Catch EPIPE in built-in cat and return as SIGPIPE (LP#1532621)
[tg] Fix errno in print/echo builtin; optimise that and unbksl
[tg] Update documentation, point out POSIX violation (Austin#1015)

R52 is a strongly recommended bugfix release:

[_0bitcount] Move moving external link from mksh(1) to the #ksh channel homepage linked therein
[tg] Make setenv “set -u”-safe and fix when invoked with no args
[tg] Make “typeset -f” output reentrant if name is a reserved word
[oksh] Zero-pad seconds in “time” output to align columns
[tg] Check signals and errorlevels from OS to be within bounds
[komh, tg] Quote and document ‘;’ as PATH separator in some places
[oksh, tg] Simplify code to call afree() even if arg is NULL
[tg] Fix tree-printing and reentrancy of multiple here documents
[tg] Work around LP#1030581 by permitting exactly one space after
[tg, oksh] Code quality work, cleanups
[tg] New code for here documents/strings with several bugfixes
[tg] Stop using issetugid(2) for ±p checks, wrong tool for the job
[tg] Reintroduce some -o posix changes lost in 2005, plus fixes
[tg] Make “source” into a built-in command
[tg] Drop “stop” alias, lksh(1) functionality to auto-unalias
[tg] Fix \u0000 ignored in $'…' and print
[tg] Improve portability of Build.sh
[Jilles Tjoelker] Improve portability of testsuite
[tg] Fix tilde expansion for some substitutions (izabera, Chet, Geoff)
[tg] Improve reparsing of ((…) |…) as ( (…) |…)
[Martijn Dekker] Fix test(1) not returning evaluation errors
[tg] Fix ${*:+x} constructs (carstenh)
[tg] Make (( … )) into a compound command (ormaaj)
[tg] Repair a few parameter substitution expansion mistakes

Change-Id: I3d875b5c648c60cc6203c964f11d8cbd42648d9c

8 years agoMerge "adbd sets $TERM now."
Elliott Hughes [Mon, 16 Nov 2015 23:42:01 +0000 (23:42 +0000)]
Merge "adbd sets $TERM now."

8 years agoadbd sets $TERM now.
Elliott Hughes [Mon, 16 Nov 2015 19:09:40 +0000 (11:09 -0800)]
adbd sets $TERM now.

Bug: http://b/25601436
Change-Id: I532233a1cf349ac0117bdfb898ef050c42357178

8 years agoMerge "adbd sets HOME/LOGNAME/SHELL/USER now."
Elliott Hughes [Mon, 2 Nov 2015 22:45:42 +0000 (22:45 +0000)]
Merge "adbd sets HOME/LOGNAME/SHELL/USER now."

8 years agoadbd sets HOME/LOGNAME/SHELL/USER now.
Elliott Hughes [Mon, 2 Nov 2015 21:32:07 +0000 (13:32 -0800)]
adbd sets HOME/LOGNAME/SHELL/USER now.

We also don't need to export variables we inherited --- they're
automatically exported.

Bug: http://b/19635681
Change-Id: Ie8e8480e7487ddc0fb9e984235b8b5ea0c587686

8 years agoMerge "Remove hd."
Elliott Hughes [Fri, 25 Sep 2015 15:15:02 +0000 (15:15 +0000)]
Merge "Remove hd."

8 years agoMerge "Remove setenv."
Elliott Hughes [Fri, 25 Sep 2015 15:14:36 +0000 (15:14 +0000)]
Merge "Remove setenv."

8 years agoMerge "Remove auto addition of ~/bin subdirectories to the path."
Elliott Hughes [Fri, 25 Sep 2015 15:14:14 +0000 (15:14 +0000)]
Merge "Remove auto addition of ~/bin subdirectories to the path."

8 years agoRemove hd.
Elliott Hughes [Fri, 25 Sep 2015 00:22:42 +0000 (17:22 -0700)]
Remove hd.

hd was only usable from interactive shells. xxd is more functional and usable
from interactive and non-interactive shells.

Change-Id: I294acea221e20ca46b6bc7e8e34906bd5d1b6dad

8 years agoRemove setenv.
Elliott Hughes [Fri, 25 Sep 2015 00:19:16 +0000 (17:19 -0700)]
Remove setenv.

No one is expecting the C shell any more.

Change-Id: Iff9422d254a51c988f2a6157379660b4ddaa617f

8 years agoRemove auto addition of ~/bin subdirectories to the path.
Elliott Hughes [Fri, 25 Sep 2015 00:18:28 +0000 (17:18 -0700)]
Remove auto addition of ~/bin subdirectories to the path.

Change-Id: Icde133fb21fa93ba3909b40da62da81654be1bab

8 years agoMerge "Upgrade to mksh 51."
Elliott Hughes [Fri, 17 Jul 2015 20:07:14 +0000 (20:07 +0000)]
Merge "Upgrade to mksh 51."

8 years agoUpgrade to mksh 51.
Elliott Hughes [Fri, 17 Jul 2015 18:39:41 +0000 (11:39 -0700)]
Upgrade to mksh 51.

Bug: http://b/22413333
Change-Id: Id86d5bf463af540b84d03a7b8fe092d23702f414

8 years agoMerge "Upgrade to mksh 50f."
Elliott Hughes [Wed, 22 Apr 2015 19:51:32 +0000 (19:51 +0000)]
Merge "Upgrade to mksh 50f."

9 years agoUpgrade to mksh 50f.
Elliott Hughes [Tue, 21 Apr 2015 20:39:18 +0000 (13:39 -0700)]
Upgrade to mksh 50f.

Bug: http://b/20101805
Change-Id: Ic311851b31217517f6c201b5ec68a0668577eddf

9 years agoMerge "Apparently not all our toolchains have -fstack-protector-strong yet."
Elliott Hughes [Wed, 11 Mar 2015 17:29:33 +0000 (17:29 +0000)]
Merge "Apparently not all our toolchains have -fstack-protector-strong yet."

9 years agoApparently not all our toolchains have -fstack-protector-strong yet.
Elliott Hughes [Wed, 11 Mar 2015 17:27:23 +0000 (10:27 -0700)]
Apparently not all our toolchains have -fstack-protector-strong yet.

Change-Id: Ieaf13c9d402a6f1bc13c393de9aa965a4e7a845d

9 years agoMerge "Upgrade to mksh 50e."
Elliott Hughes [Wed, 11 Mar 2015 16:58:17 +0000 (16:58 +0000)]
Merge "Upgrade to mksh 50e."

9 years agoUpgrade to mksh 50e.
Elliott Hughes [Wed, 11 Mar 2015 05:22:24 +0000 (22:22 -0700)]
Upgrade to mksh 50e.

Change-Id: I341673d83c67f83761aa02c8372da6bbc419286f

9 years agoam 948ca12d: Merge "Set to /data/local/tmp, and use `id -un` to get the user\'s...
Elliott Hughes [Tue, 10 Feb 2015 16:52:45 +0000 (16:52 +0000)]
am 948ca12d: Merge "Set  to /data/local/tmp, and use `id -un` to get the user\'s name."

* commit '948ca12d46512a832fe53ee24fd580e7addaf28d':
  Set $TMPDIR to /data/local/tmp, and use `id -un` to get the user's name.

9 years agoMerge "Set $TMPDIR to /data/local/tmp, and use `id -un` to get the user's name."
Elliott Hughes [Mon, 9 Feb 2015 22:02:37 +0000 (22:02 +0000)]
Merge "Set $TMPDIR to /data/local/tmp, and use `id -un` to get the user's name."

9 years agoSet $TMPDIR to /data/local/tmp, and use `id -un` to get the user's name.
Elliott Hughes [Mon, 9 Feb 2015 20:25:41 +0000 (12:25 -0800)]
Set $TMPDIR to /data/local/tmp, and use `id -un` to get the user's name.

Needed to make various command-line tools work out of the box, since we
don't have /tmp, which is what they'd naturally fall back to.

Change-Id: I3091d5ec9235f977de4a698a34f1f147a69e8a33

9 years agoam 508b5747: Merge "Default to xterm, remove some cruft."
Elliott Hughes [Fri, 12 Dec 2014 01:51:27 +0000 (01:51 +0000)]
am 508b5747: Merge "Default to xterm, remove some cruft."

* commit '508b57470e9910e8c3be448ec4428ea02182c473':
  Default to xterm, remove some cruft.

9 years agoMerge "Default to xterm, remove some cruft."
Elliott Hughes [Fri, 12 Dec 2014 01:21:59 +0000 (01:21 +0000)]
Merge "Default to xterm, remove some cruft."

9 years agoDefault to xterm, remove some cruft.
Elliott Hughes [Fri, 12 Dec 2014 01:13:19 +0000 (17:13 -0800)]
Default to xterm, remove some cruft.

gtest won't color its output unless TERM is an xterm, and that seems as
reasonable a default as vt100. More reasonable, in fact. This only fixes
things for login shells, so we still need to fix adb to pass $TERM
through to make yabinc happy...

We can get a better more(1) from toybox, and the aliases seem inappropriate.

Change-Id: Id73035d69ab421b179b9beb575d4e1607fbcce62

9 years agoam a1c1f910: am 319a679f: Merge "Upgrade to mksh 50d."
Elliott Hughes [Sun, 26 Oct 2014 22:48:07 +0000 (22:48 +0000)]
am a1c1f910: am 319a679f: Merge "Upgrade to mksh 50d."

* commit 'a1c1f910f501fd658f7b2bea133bd47cb8d2395d':

9 years agoam fd0e00a2: am 1f487419: Merge "Upgrade to mksh 50c."
Elliott Hughes [Sun, 26 Oct 2014 22:48:07 +0000 (22:48 +0000)]
am fd0e00a2: am 1f487419: Merge "Upgrade to mksh 50c."

* commit 'fd0e00a245aa73ac85e5af7ba5191a999ff7bda3':

9 years agoam a0a78c0a: am 049d9caa: Merge "Make mksh\'s mkmf.sh target-agnostic."
Elliott Hughes [Sun, 26 Oct 2014 22:48:07 +0000 (22:48 +0000)]
am a0a78c0a: am 049d9caa: Merge "Make mksh\'s mkmf.sh target-agnostic."

* commit 'a0a78c0a155ec954a8a67e155348e7424dd63237':

9 years agoam 319a679f: Merge "Upgrade to mksh 50d."
Elliott Hughes [Tue, 7 Oct 2014 23:23:18 +0000 (23:23 +0000)]
am 319a679f: Merge "Upgrade to mksh 50d."

* commit '319a679f02fd6ac5f7cbde03c3e64935b2c9af82':
  Upgrade to mksh 50d.

9 years agoam 319a679f: Merge "Upgrade to mksh 50d."
Elliott Hughes [Tue, 7 Oct 2014 23:20:53 +0000 (23:20 +0000)]
am 319a679f: Merge "Upgrade to mksh 50d."

* commit '319a679f02fd6ac5f7cbde03c3e64935b2c9af82':
  Upgrade to mksh 50d.

9 years agoam ae046317: Upgrade to mksh 50d.
Elliott Hughes [Tue, 7 Oct 2014 22:41:30 +0000 (22:41 +0000)]
am ae046317: Upgrade to mksh 50d.

* commit 'ae04631768fefc45c2d18865ec88872853421a52':
  Upgrade to mksh 50d.

9 years agoUpgrade to mksh 50d.
Elliott Hughes [Tue, 7 Oct 2014 22:04:14 +0000 (15:04 -0700)]
Upgrade to mksh 50d.

Bug: 17814343

(cherry picked from commit f7f795644b6b1016c35b09d88c52f427473b3bae)

Change-Id: Ie8e4251d173b6e5aa7b0c4ef88a29bcccc23715a

9 years agoMerge "Upgrade to mksh 50d."
Elliott Hughes [Tue, 7 Oct 2014 22:09:07 +0000 (22:09 +0000)]
Merge "Upgrade to mksh 50d."