aboutsummaryrefslogtreecommitdiffstats
path: root/hw
Commit message (Collapse)AuthorAgeFilesLines
* target-arm: integrate upstream ARM translator.David 'Digit' Turner2011-06-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The new translator has the following benefits: - faster emulation of ARMv5TE code (through improved JIT) - proper support for ARMv7 and NEON - rebuilding the full-eng platform images for ARMv7-A results in additionnal speed increases (a.k.a. Thumb-2 rocks!). Note that, as an interesting side effect, NEON machine code is generally slower than the equivalent C code it is supposed to replace when run inside the emulator. This can be explained by the fact that for now the translator simply translates each NEON instruction into a series of sequential host instructions (and also requires over-head for packing/unpacking/saturation/ etc...). This change has been tested by running the "full-eng" platform image rebuilt for ARMv7-A and Neon and using an appropriate kernel image (prebuilt/android-arm/kernel/kernel-qemu-armv7). The system could boot and seems to work perfectly. Not a single issue has been experienced during testing. On a 2.4 GHz Xeon CPU, the image boots in about 25 seconds (compared to 40 seconds for a vanilla one without this emulator patch). Thanks to Peter Maydell at Linaro and ARM with his hard work to make this happen (first in upstream, and now on Android). This integration is based on the Meego git repository (git://gitorious.org/qemu-maemo/qemu.git) using the following hash: 7e2d65b0c95c865b1fa6d3d4948e8e822b9ac2fd On top of which, the following upstream patch has been applied (with recommendation from Peter): b7fa9214d8d4f57992c9acd0ccb125c54a095f00 (We chose this repository because it was the closest to the previous integrate. We will probably use the Linaro ones for future work on this part of the emulator). Change-Id: I54837e3d2e908b2380d158411d7a9813630e7e4e
* trace.h: rename to trace-android.hDavid 'Digit' Turner2011-06-011-1/+1
| | | | | | | This is done to allow for future integration of the upstream auto-generated "trace.h" header file. Change-Id: I9d1bba3320d9e8acfd7359540d3ed6695773061e
* ramblocks: integrate upstream implementation (sophisticated)David 'Digit' Turner2011-06-014-21/+25
| | | | Change-Id: I49e96e2d5ae571849b0b6fef0a30b41ecdee8d23
* qemu-char-android.c: rename to qemu-char.cDavid 'Digit' Turner2011-06-012-2/+2
| | | | | | + convert to QemuOpts* initialization functions. Change-Id: I5a245ed2de068af49fdc92de8e3e834d0f49aa27
* savevm.c: minor integrateDavid 'Digit' Turner2011-06-011-5/+6
| | | | Change-Id: I16103c65ac7b15b2dc58dcc7dd6b3052004aa31a
* qemu-timer.c: rename qemu_timer_new_scale()David 'Digit' Turner2011-06-019-66/+56
| | | | | | | | | | | Get rid of qemu_timer_new() implementation, and update all callers to use qemu_timer_new_ms() or qemu_timer_new_ns() instead. Rename qemu_new_timer_scale() to qemu_new_timer() to follow upstream conventions. Change-Id: Id2c04f8597ec5026e02f87b3e2c5507920eb688e
* New goldfish_pipe virtual device.David 'Digit' Turner2011-04-284-614/+996
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds a new virtual hardware device named "goldfish_pipe" used to implement a very fast communication channel between the guest system and the emulator. IMPORTANT: This depends on a special kernel driver, see: https://review.source.android.com/#change,22496 Usage from the guest is simply the following: fd = open("/dev/qemu_pipe", O_RDWR); const char* pipename = "pipe:<name>"; ret = write(fd, pipename, strlen(pipename)+1); if (ret < 0) { /* could not connect to service named <name> */ } /* now you can read()/write()/close() as a normal * file descriptor to exchange data with the service. */ In addition, this implements the following pipe services in the emulator: tcp:<port> tcp:<hostname>:<port> unix:<path> opengles The 'tcp:' and 'unix:' services simply redirect to a TCP or Unix socket on the host with minimal The 'opengles' service simply connects to tcp:locahost:22468 for now. We may change this to be more configurable in the future, but that's the port number used by the current experimental OpenGL ES hardware emulation host libraries / programs. Benchmarking with a simple ping-pong program shows that the guest <-> emulator can achieve a roundtrip bandwidth of 192 MB/s (on a 2.7 Ghz Xeon PC). Using the tcp: service to talk to a ping-pong server listening on localhost reaches 102 MB/s on the same machine, using a Unix socket reaches 140 MB/s. By contrast, using standard sockets in the guest reaches only 3.8 MB/s on the same machine (and requires special privileges from the application anyway).
* Merge "Converted a variable from char * to const char *. Removes compiler ↵David Turner2011-04-111-2/+2
|\ | | | | | | warnings with some compilers."
| * Converted a variable from char * to const char *.Ashish Bhatia2011-04-081-2/+2
| | | | | | | | | | | | Removes compiler warnings with some compilers. Change-Id: Idfb14557181e744d07bef6b5d0da734012d2ff42
* | hw/android_arm.c: Initialize qemud pipes properly.David 'Digit' Turner2011-04-111-0/+4
| | | | | | | | Change-Id: I7f7f46b6e9526d207f53ad6a9e6baa315bd6af7b
* | goldfish_trace: Hook with goldfish_pipe.h functions.David 'Digit' Turner2011-04-113-36/+102
| | | | | | | | | | | | | | | | Note that this is also forces tracing support during machine initialization. Measurements show that this doesn't have any significant impact on emulation performance. Change-Id: Ie32352659611531ce65ba2c95c5896e2421f509d
* | hw/goldfish_pipe.c: New support code for QEMUD fast-pipes.David 'Digit' Turner2011-04-112-0/+838
|/ | | | Change-Id: I338334d53fa9bc52c87e9da18341d0cb94fd4269
* Revert "Revert "Merge 29e114d2 from master.""Xavier Ducrohet2011-03-181-0/+36
| | | | | | | | | This reverts commit 2facea6512c84af9285682909f6ac6edcd4130ab which was not supposed to merge down from tools_R10 to master since the original reverted commit was a cherry-pick from master to tools_r10 Change-Id: Ide87af693a0f3239bacaa7313b9262e6e8c46800
* merge from tools_r10The Android Open Source Project2011-03-181-36/+0
|\ | | | | | | Change-Id: I70d9ccafb5912c85baa0711f7e9010bec223efaa
| * Revert "Merge 29e114d2 from master. do not merge."Xavier Ducrohet2011-03-181-36/+0
| | | | | | | | This reverts commit e36a1cc2561b2ce27d3c0a5b3a7b79d4e8531f0e.
| * Merge 29e114d2 from master. do not merge.David 'Digit' Turner2011-03-181-0/+36
| | | | | | | | | | | | | | | | | | | | | | | | Fix touchscreen emulation to return correct min/max coordinate bounds. This is needed by future changes in the input framework that depend on touchscreen devices to properly return the min/max bounds of their absolute pointer coordinates. Fixes bug 4126574 Change-Id: I1ace56d5433be5d97eb79895cd57b6e442245ca0
| * Merge 59e8e196 into tools_r10. DO NOT MERGE.David 'Digit' Turner2011-02-161-3/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | Add hw.keyboard.lid This is used to control the lid switch virtual device. + Add support for a "version <number>" field to the skin format. This will later be used for skin format #3. + Fix a bug where string hw properties didn't work/compile Change-Id: I27ff3293b439917a51c584a25139264ddb1aff12
* | Fix touchscreen emulation to return correct min/max coordinate bounds.David 'Digit' Turner2011-03-181-0/+36
| | | | | | | | | | | | | | | | | | | | This is needed by future changes in the input framework that depend on touchscreen devices to properly return the min/max bounds of their absolute pointer coordinates. Fixes bug 4126574 Change-Id: I982c0d8e78fc59912e964af958929d2b0718eaa4
* | Fix a crash that happened when the content directory didn't have a cache.img ↵David 'Digit' Turner2011-03-151-2/+3
| | | | | | | | | | | | | | | | | | file. This only happened for fresh newly-created AVDs. Existing ones used for testing the emulator already had a cache.img and booted properly. Change-Id: I87901ea5fb1dc6aa7d2bb3dd191de58d972c386f
* | x86: Enable KVM mode for Android x86 emulator (update/rebase)Jun Nakajima2011-03-043-1/+45
| | | | | | | | | | | | | | | | | | | | This patch enables KVM mode on x86 Linux to boost performance of x86 emulaiton if the hardware-based virtualization feature is present on the host machine. Change-Id: I4b24474b3ec115a3b9a7bf017801f4f610253b09 Signed-off-by: Xiaohui Xin <xiaohui.xin@intel.com> Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com> Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
* | Move data partition initialization to core.David 'Digit' Turner2011-03-011-1/+1
| | | | | | | | Change-Id: I3c2b4668593391026da028194503fc87246e44ba
* | Move cache partition initialization to core.David 'Digit' Turner2011-03-011-0/+2
| | | | | | | | Change-Id: I1f887e6f8fc38e43b1fff3f7bab3814b52542762
* | x86: Fixing audio.Jun Nakajima2011-02-241-0/+6
| | | | | | | | | | | | Change-Id: I0614450cd512d52bf24020dda1c933f0b22cb09b Signed-off-by: Xiaohui Xin <xiaohui.xin@intel.com> Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
* | Minor cleanups for core/ui Makefile separation.David 'Digit' Turner2011-02-231-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch contains minor cleanups that will be needed by a few future patches that will improve the way core and ui sources are separated and built. - remove a few compiler warnings - formatting - remove the memcheck CONFIG_MEMCHECK checks (they later get in the way) - refine hw/hw.h inclusion's of cpu.h (this comes from upstream) - add missing definitions for PRUd64 and PRUx64 (used later) - remove CONFIG_SHAPER test, replace with CONFIG_ANDROID instead. - add missing strdup() calls. Change-Id: Ic7d6681a51af718c298f0ee4bd884b1d8750f28e
* | Add hw.keyboard.lidDavid 'Digit' Turner2011-02-101-3/+2
| | | | | | | | | | | | | | | | | | | | | | This is used to control the lid switch virtual device. + Add support for a "version <number>" field to the skin format. This will later be used for skin format #3. + Fix a bug where string hw properties didn't work/compile Change-Id: I678a0988d2cd24169dd0c2ece96e7ab0b2822104
* | Revert "Build arm and x86 binaries at the same time."Xavier Ducrohet2011-02-091-3/+0
| | | | | | | | | | | | This fails to build on MacOS X This reverts commit a39b10bd2574825a815d6ad854499dd127cfa9cb.
* | Build arm and x86 binaries at the same time.David 'Digit' Turner2011-02-101-0/+3
|/ | | | Change-Id: I105c5a097c988cb964b47b40b71c7a08af0d9210
* Fixing aborts caused by compute_fb_update_rect_linear() when the framebuffersJun Nakajima2011-02-081-1/+10
| | | | | | | | are large. Also, added a boot parameter for the goldfish kernel to use PIT, rather than TSC. Change-Id: I5824c7248e0c1a4c487aeb52da112d870846ae62 Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
* events: Fix no-Dpad emulation when hw.keyboard=yesDavid 'Digit' Turner2011-02-061-0/+23
| | | | | | | | | | This fixes a bug where Dpad emulation was forced when hw.keyboard was set to 'yes' in hardware.ini, even if hw.dpad was set to no. The problem was that the events device set the Dpad EV_KEY bits unconditionally since they are in the 0..0xff range. Change-Id: Ice2488341170f43d3fe230d816a089da97c4d5ac
* x86: Add x86 support. Rebase the change (20906/1) due to a minor conflict.Jun Nakajima2011-02-023-4/+21
| | | | | | | Change-Id: Ic73cca0fc6c6e5cf74f63daa6080d00aa7c392bb Signed-off-by: Xiaohui Xin <xiaohui.xin@intel.com> Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com> Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
* New files are from the upstream QEMU (0.10.5), and the code is based onJun Nakajima2011-01-2915-0/+7212
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | the following commit (dated 2009-06-19, see CHANGES.TXT): d2e9fd8f703203c2eeeed120b1ef6c3a6574e0ab new file: hw/apic.c new file: hw/fdc.h new file: hw/fw_cfg.c new file: hw/fw_cfg.h new file: hw/i8254.c new file: hw/i8259.c new file: hw/ioapic.c new file: hw/mc146818rtc.c new file: hw/ne2000.c new file: hw/pc.c new file: hw/pckbd.c new file: hw/piix_pci.c new file: hw/ps2.c new file: hw/ps2.h new file: hw/smbios.c new file: target-i386/TODO new file: target-i386/cpu.h new file: target-i386/exec.h new file: target-i386/helper.c new file: target-i386/helper.h new file: target-i386/helper_template.h new file: target-i386/kvm.c new file: target-i386/machine.c new file: target-i386/op_helper.c new file: target-i386/ops_sse.h new file: target-i386/ops_sse_header.h new file: target-i386/svm.h new file: target-i386/translate.c Change-Id: I55c62ed7516f002c882705545e7c21997ece9927
* Security fix from Tavis Ormandyrich cannings2011-01-251-3/+5
| | | | | | Perform null check before calling r->transfer_handler. Change-Id: Ide3cd7edc7bde12a0635572bfa72f3a6dd05a926
* Make "qwerty2" the default charmap.David 'Digit' Turner2011-01-191-8/+22
| | | | | | | | | | | | | | | This patch makes "qwerty2" the default charmap for skins that don't specifiy one explicitely. Note that the "qwerty" charmap is here for historical reasons only. All skins provided by all SDK Android platforms explicitely mention "qwerty2", so this patch will mainly impact "magic skins" like those created by an option like "-skin 1024x768". This gets rid of the obsolete "qwerty" charmap (it was never used by any of the skins provided by the SDK, and is incorrect). + Simplify android/charmap.[hc] code. + Lazy-load the charmap name in hw/goldfish_events_device.c
* Sligh speedup for pixel conversion routines with Duff's device.David 'Digit' Turner2011-01-181-20/+33
| | | | Change-Id: If832bc5844945f7a2027b2f8d09393586545d8d5
* hw: framebuffer: add support for 32-bit framebuffers.David 'Digit' Turner2011-01-181-14/+21
| | | | | | | This provides a new read i/o port to return the framebuffer's format to the kernel driver. Change-Id: Ifd1c4d8a4dc75fa73d5f6750e0b501c34468aec5
* hw: goldfish_fb: Minimize the size of the update rectangle.David 'Digit' Turner2011-01-161-92/+323
| | | | | | | | | | | | | | | | This patch modifies the framebuffer's update callback to compute the minimum bounding rectangle of each framebuffer update. Before that, the code only computed the minimum and maximum Y coordinates, assuming xmin=0 and xmax=width-1. This should help reduce un-necessary traffic/churn for our core->UI framebuffer protocol. Moreover, introduce the ability to support framebuffers / display surfaces with different depths. This is a first step towards providing 32-bit framebuffers to the guest system to slightly speed-up emulated rendering on Gingerbread and above. Change-Id: I6194824ce67a3b1d6d95b5d97ddec3ef459e408e
* upstream: integrate block changesDavid 'Digit' Turner2011-01-024-14/+133
| | | | | | | | This large patch upgrades the block support code to the upstream version available in ba5e7f82169f32ab8163c707d97c799ca09f8924 dated 2010-08-08 Change-Id: I8b24df0c287e72f6620650a4d6a62e1bb315453e
* Add -audio-test-out option to the core.David 'Digit' Turner2011-01-021-1/+1
| | | | | | | | | | | | | This option can be used to quickly check audio output (without having to boot a full AVD and launch a sound-playing application). Usage: emulator <other-options> -qemu -audio-test-out qemu-android <other-options> -audio-test-out This simply generates an ugly saw signal, but that's enough for us. Change-Id: I060300b4000b9705d181c6262de8d4d13c749e69
* Replacement of padding with ftruncate.Tim Baverstock2010-11-231-33/+53
| | | | | | | | Also stored some other parameters, speculatively. Guard ftruncate and lseek against EINTR Change-Id: I4237640318564db6c375a90415097c00e4fcafaa
* Add env var ANDROID_TMP to override /tmp/android location.Raphael2010-11-161-1/+1
| | | | | | | | | | | | This is needed if there's already a /tmp/android which the current user can't access on the system. It would prevent the emulator from starting with a cryptic error message: NAND: could not create temp file for system NAND disk image: Permission deniedemulator: User-config was not changed. SDK Bug: 3154431 Change-Id: Iefe7fd6a5139da545aaef5670fb7f42d210d0edc
* Fix linux-x86_64 build.David 'Digit' Turner2010-10-212-3/+3
| | | | | | | | | Fix various 64-bitness issues in the source code to make the --try-64 option work again on Linux. Note that the generated binary is not faster than its 32-bit variant when it comes to benchmarking the boot sequence. Change-Id: Iad248e033757d4cd25524a438a5dbe1cf3aca6cf
* Fix incorrect orientation at boot.David 'Digit' Turner2010-10-201-6/+41
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The goldfish kernel's event driver is very picky about the state of the virtual device it talks to. Essentially, it can only listen to h/w events if the corresponding IRQ is raised _after_ it has completed some initial setup. If the IRQ is raised before, the driver will refuse to listen to events, and any interaction becomes impossible. We changed the way our UI windows are built and managed previously, and this had the unfortunate effect of sending one h/w event too soon, which, in the old code, would raise the IRQ prematurely. This resulted in an input freeze. To work around this the UI code was modified to not send the initial h/w event, but this resulted in bad orientation of the home screen at the end of the boot sequence. This change allows the virtual device to wait for the kernel driver before raising the IRQ, in the case any h/w event has been buffered. It also removes the hack in the UI code, allowing the send of the initial h/w event at startup. BONUS MINOR CHANGES: EsounD probe: fix error message. console: fix compiler warnings libpng: proper handling of MMX support. android-configure.sh: fix --ignore-audio handling The option didn't do anything if the audio development packages are properly installed on the system. Change-Id: Ib134158873d5cb72290c9676d92d20233226c889
* Added state snapshot support for QemuD.Ot ten Thije2010-10-141-0/+2
| | | | | | | | | | | | | | | | | With this patch, both modem and sensor functionality are correctly restored when a state snapshot is loaded. This was not the case previously because communication with either of these services is done using the qemud daemon, which did not support snapshots. The boot-properties and charpipe services have no specific save/load functionality yet, since the framework itself should be reviewed first. Adding support for bootproperties should not be difficult though, and charpipe may not need it. For a description of the high-level process for saving and loading, consult section IV "State snapshots" in docs/ANDROID-QEMUD.TXT. Change-Id: I5b06d88b911ca096e78060163174904c48a01c66
* Merge "Permanent fix for workaround of timer issue (commit a7f114bc)."Ot ten Thije2010-10-141-2/+1
|\
| * Permanent fix for workaround of timer issue (commit a7f114bc).Ot ten Thije2010-09-241-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | The compiler generated incorrect instructions when loading timers, because qemu_get_be64 was used before it was declared. This caused the compiler to assume that a normal (32 bit) int would be returned, rather than a 64 bit one. Just including the "hw/hw.h" header is sufficient to correct this. The makefile for the project has recently been changed to build with -Wall, which should prevent this from going unnoticed again. Change-Id: I74c98183287cf26b767a36f7a6ff9ebf0b891826
* | Save NAND disk state in snapshots.Ot ten Thije2010-10-131-45/+206
|/ | | | | | | | | | | | | | | | This patch fixes saving and restoring the state of the emulated NAND disks, by copying all contents of each disk into the snapshot file. During restore the disks are then entirely overwritten with the snapshot contents. The main issue with this approach is that the size of the disk snapshots is determined by the maximum capacity of the disk in question, rather than the actual size of the underlying file, which may be considerably smaller. The difference is filled with padding. Unfortunately, obvious approaches to avoid this padding result in incorrect state after restoring. Change-Id: I4abe329ff2cf2f159c3d6dfc206959614cc4ca35
* Remove compiler warnings when building the emulator.David 'Digit' Turner2010-09-226-13/+28
| | | | | | | This forces -Wall during the build. Note that this patch doesn't remove all warnings, but most of the remaining ones are from upstream anyway. Change-Id: I8808d8495e99866e156ce5780d2e3c305eab491f
* upstream: osdep changes + fix mingw build.David Turner2010-09-131-41/+41
|
* upstream: minow hw updates.David Turner2010-09-1312-45/+40
|
* upstream: buffered file minor change.David Turner2010-09-131-1/+3
|
generated by cgit v1.1 at 2025-02-10 16:01:33 +0000