aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2013-11-25qemu: mempath: prefault pages manually (v4)Marcelo Tosatti
v4: s/fail/failed/ (Peter Maydell) Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
2013-11-25kvm: Query KVM for available memory slotsAlex Williamson
KVM reports the number of available memory slots (KVM_CAP_NR_MEMSLOTS) using the extension interface. Both x86 and s390 implement this, ARM and powerpc do not yet enable it. Convert the static slots array to be dynamically allocated, supporting more slots when available. Default to 32 when KVM_CAP_NR_MEMSLOTS is not implemented. The motivation for this change is to support more assigned devices, where memory mapped PCI MMIO BARs typically take one slot each. Signed-off-by: Alex Williamson <alex.williamson@redhat.com> Reviewed-by: Thomas Huth <thuth@linux.vnet.ibm.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2013-11-21Update version for 1.7.0-rc1 releasev1.7.0-rc1Anthony Liguori
Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21vfio-pci: Fix multifunction=onAlex Williamson
When an assigned device is initialized it copies the device config space into the emulated config space. Unfortunately multifunction is setup prior to the device initfn and gets clobbered. We need to restore it just like pci-assign does. Signed-off-by: Alex Williamson <alex.williamson@redhat.com> Reviewed-by: Bandan Das <bsd@redhat.com> Message-id: 20131112185059.7262.33780.stgit@bling.home Cc: qemu-stable@nongnu.org Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21target-i386: Fix addr32 prefix in gen_lea_modrmRichard Henderson
Fix the following run-test-x86_64 testsuite failures: -lea (%%eax) = 0000000000000001 -lea (%%ebx) = 0000000000000002 -lea (%%ecx) = 0000000000000004 -lea (%%edx) = 0000000000000008 -lea (%%esi) = 0000000000000010 -lea (%%edi) = 0000000000000020 +lea (%%eax) = 0000abcc00000001 +lea (%%ebx) = 0000abcf00000002 +lea (%%ecx) = 0000abc900000004 +lea (%%edx) = 0000abc500000008 +lea (%%esi) = 0000abdd00000010 +lea (%%edi) = 0000abed00000020 In addition, reduce ifdeffery and minimize the number of TCG ops produced during address computation. Signed-off-by: Richard Henderson <rth@twiddle.net> Message-id: 1384219016-5170-1-git-send-email-rth@twiddle.net Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21atomic.h: Fix build with clangPeter Maydell
clang defines __ATOMIC_SEQ_CST but its implementation of the __atomic_exchange() builtin differs from that of gcc. Move the __clang__ branch of the ifdef ladder to the top and fix its implementation (there is no such builtin as __sync_exchange), so we can compile with clang again. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 1382435921-18438-1-git-send-email-peter.maydell@linaro.org Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21target-i386: do not override nr_cores for -cpu hostPaolo Bonzini
Commit 787aaf5 (target-i386: forward CPUID cache leaves when -cpu host is used, 2013-09-02) brings bits 31..26 of CPUID leaf 04h out of sync with the APIC IDs that QEMU reserves for each package. This number must come from "-smp" options rather than from the host CPUID. It also turns out that this unsyncing makes Windows Server 2012R2 fail to boot. Tested-by: Peter Lieven <pl@kamp.de> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Benoit Canet <benoit@irqsave.net> Reviewed-by: Eduardo Habkost <ehabkost@redhat.com> Message-id: 1384879786-6721-1-git-send-email-pbonzini@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21mips jazz: do not raise data bus exception when accessing invalid addressesHervé Poussineau
MIPS Jazz chipset doesn't seem to raise data bus exceptions on invalid accesses. However, there is no easy way to prevent them. Creating a big memory region for the whole address space doesn't prevent memory core to directly call unassigned_mem_read/write which in turn call cpu->do_unassigned_access, which (for MIPS CPU) raise an data bus exception. This fixes a MIPS Jazz regression introduced in c658b94f6e8c206c59d02aa6fbac285b86b53d2c. Signed-off-by: Hervé Poussineau <hpoussin@reactos.org> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Hervé Poussineau <hpoussin@reactos.org> Message-id: 1383603977-7003-1-git-send-email-hpoussin@reactos.org Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21target-i386: yield to another VCPU on PAUSEPaolo Bonzini
After commit b1bbfe7 (aio / timers: On timer modification, qemu_notify or aio_notify, 2013-08-21) FreeBSD guests report a huge slowdown. The problem shows up as soon as FreeBSD turns out its periodic (~1 ms) tick, but the timers are only the trigger for a pre-existing problem. Before the offending patch, setting a timer did a timer_settime system call. After, setting the timer exits the event loop (which uses poll) and reenters it with a new deadline. This does not cause any slowdown; the difference is between one system call (timer_settime and a signal delivery (SIGALRM) before the patch, and two system calls afterwards (write to a pipe or eventfd + calling poll again when re-entering the event loop). Unfortunately, the exit/enter causes the main loop to grab the iothread lock, which in turns kicks the VCPU thread out of execution. This causes TCG to execute the next VCPU in its round-robin scheduling of VCPUS. When the second VCPU is mostly unused, FreeBSD runs a "pause" instruction in its idle loop which only burns cycles without any progress. As soon as the timer tick expires, the first VCPU runs the interrupt handler but very soon it sets it again---and QEMU then goes back doing nothing in the second VCPU. The fix is to make the pause instruction do "cpu_loop_exit". Cc: Richard Henderson <rth@twiddle.net> Reported-by: Luigi Rizzo <rizzo@iet.unipi.it> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Richard Henderson <rth@twiddle.net> Message-id: 1384948442-24217-1-git-send-email-pbonzini@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21rng-egd: offset the point when repeatedly read from the bufferAmos Kong
The buffer content might be read out more than once, currently we just repeatedly read the first data block, buffer offset is missing. Cc: qemu-stable@nongnu.org Signed-off-by: Amos Kong <akong@redhat.com> Message-id: 1385023371-8198-3-git-send-email-akong@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21rng-egd: remove redundant freeAmos Kong
We didn't set default chr_name, the free is redundant. Signed-off-by: Amos Kong <akong@redhat.com> Message-id: 1385023371-8198-2-git-send-email-akong@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21virtio-rng: add check of periodAmos Kong
If period is assigned to 0, limit timer will expire immediately. It causes a qemu warning: "main-loop: WARNING: I/O thread spun for 1000 iterations" This limit is meaningless. This patch forbids to assign 0 to period. Reviewed-by: Amit Shah <amit.shah@redhat.com> Signed-off-by: Amos Kong <akong@redhat.com> Message-id: 1385031203-23790-1-git-send-email-akong@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21s390x: fix flat file load on 32 bit systemsMichael S. Tsirkin
pc-bios/s390-zipl.rom is a flat image so it's expected that loading it as elf will fail. It should fall back on loading a flat file, but doesn't on 32 bit systems, instead it fails printing: qemu: hardware error: could not load bootloader 's390-zipl.rom' The result is boot failure. The reason is that a 64 bit unsigned interger which is set to -1 on error is compared to -1UL which on a 32 bit system with gcc is a 32 bit unsigned interger. Since both are unsigned, no sign extension takes place and comparison evaluates to non-equal. There's no reason to do clever tricks: all functions we call actually return int so just use int. And then we can use == -1 everywhere, consistently. Reviewed-by: Alexander Graf <agraf@suse.de> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Message-id: 20131121133426.GA30827@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21acpi-build: fix build on glib < 2.14Michael S. Tsirkin
g_array_get_element_size was only added in glib 2.14. Fortunately we don't use it for any arrays where element size is > 1, so just add an assert. Reported-by: Richard Henderson <rth@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 1385036128-8753-2-git-send-email-mst@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21acpi-build: fix build on glib < 2.22Michael S. Tsirkin
g_string_vprintf was only introduced in 2.24 so switch to vsnprintf instead. A bit uglier but name size is fixed at 4 bytes here so it's easy. Reported-by: Richard Henderson <rth@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 1385036128-8753-1-git-send-email-mst@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21Merge remote-tracking branch 'qemu-kvm/uq/master' into stagingAnthony Liguori
# By Jan Kiszka (1) and others # Via Gleb Natapov * qemu-kvm/uq/master: kvm: Fix uninitialized cpuid_data pci-assign: Remove dead code for direct I/O region access from userspace KVM: x86: fix typo in KVM_GET_XCRS Message-id: cover.1385040432.git.gleb@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21Merge remote-tracking branch 'jliu/or32' into stagingAnthony Liguori
# By Sebastian Macke # Via Jia Liu * jliu/or32: target-openrisc: Correct carry flag check of l.addc and l.addic test cases target-openrisc: Correct memory bounds checking for the tlb buffers openrisc-timer: Reduce overhead, Separate clock update functions target-openrisc: Correct wrong epcr register in interrupt handler target-openrisc: Remove executable flag for every page target-openrisc: Remove unnecessary code generated by jump instructions target-openrisc: Speed up move instruction Message-id: 1384958318-9145-1-git-send-email-proljc@gmail.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-21Merge remote-tracking branch 'quintela/migration.next' into stagingAnthony Liguori
# By Wangting (Kathy) # Via Juan Quintela * quintela/migration.next: The calculation of bytes_xfer in qemu_put_buffer() is wrong Message-id: 1384946787-8190-1-git-send-email-quintela@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-20target-openrisc: Correct carry flag check of l.addc and l.addic test casesSebastian Macke
The test cases did not correctly test for the carry flag. Signed-off-by: Sebastian Macke <sebastian@macke.de> Reviewed-by: Jia Liu <proljc@gmail.com> Signed-off-by: Jia Liu <proljc@gmail.com>
2013-11-20target-openrisc: Correct memory bounds checking for the tlb buffersSebastian Macke
The mtspr and mfspr routines didn't check for the correct memory boundaries. This fixes a segmentation fault while booting Linux. Signed-off-by: Sebastian Macke <sebastian@macke.de> Reviewed-by: Jia Liu <proljc@gmail.com> Signed-off-by: Jia Liu <proljc@gmail.com>
2013-11-20openrisc-timer: Reduce overhead, Separate clock update functionsSebastian Macke
The clock value is only evaluated when really necessary reducing the overhead of the timer handling. This also solves a problem in the way the Linux kernel handles the timer and the expected accuracy. The old version could lead to inaccurate timings. Signed-off-by: Sebastian Macke <sebastian@macke.de> Reviewed-by: Jia Liu <proljc@gmail.com> Signed-off-by: Jia Liu <proljc@gmail.com>
2013-11-20target-openrisc: Correct wrong epcr register in interrupt handlerSebastian Macke
This patch corrects several misbehaviors during an interrupt process. Most of the time the pc is already correct and therefore no special treatment of the exceptions is necessary. Tested by checking crashing programs which otherwise work in or1ksim. Signed-off-by: Sebastian Macke <sebastian@macke.de> Reviewed-by: Jia Liu <proljc@gmail.com> Signed-off-by: Jia Liu <proljc@gmail.com>
2013-11-20target-openrisc: Remove executable flag for every pageSebastian Macke
Pages should be flagged executable only if the tlb executable flag is set or the mmu is off. Signed-off-by: Sebastian Macke <sebastian@macke.de> Reviewed-by: Jia Liu <proljc@gmail.com> Signed-off-by: Jia Liu <proljc@gmail.com>
2013-11-20target-openrisc: Remove unnecessary code generated by jump instructionsSebastian Macke
The sr_f variable is only used for the l.bf and l.bnf instructions. For clarity the code is also rewritten using a switch statement instead of if chaining. Signed-off-by: Sebastian Macke <sebastian@macke.de> Reviewed-by: Jia Liu <proljc@gmail.com> Signed-off-by: Jia Liu <proljc@gmail.com>
2013-11-20target-openrisc: Speed up move instructionSebastian Macke
The OpenRISC architecture does not have its own move register instruction. Instead it uses either "l.addi rd, r0, x" or "l.ori rd, rs, 0" or "l.or rd, rx, r0" The l.ori instruction is automatically optimized but not the l.addi instruction. This patch optimizes for this special case. Signed-off-by: Sebastian Macke <sebastian@macke.de> Reviewed-by: Jia Liu <proljc@gmail.com> Signed-off-by: Jia Liu <proljc@gmail.com>
2013-11-20The calculation of bytes_xfer in qemu_put_buffer() is wrongWangting (Kathy)
In qemu_put_buffer(), bytes_xfer += size is wrong, it will be more than expected, and should be bytes_xfer += l. Signed-off-by: zhangmin <zhangmin6@huawei.com> Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-11-19Merge remote-tracking branch 'quintela/migration.next' into stagingAnthony Liguori
# By Peter Lieven # Via Juan Quintela * quintela/migration.next: migration: drop MADVISE_DONT_NEED for incoming zero pages Message-id: 1384878412-23521-1-git-send-email-quintela@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-19Merge remote-tracking branch 'stefanha/block' into stagingAnthony Liguori
# By Andreas Färber # Via Stefan Hajnoczi * stefanha/block: qtest: Use -display none by default
2013-11-19Merge remote-tracking branch 'afaerber/tags/qom-devices-for-anthony' into ↵Anthony Liguori
staging QOM infrastructure fixes for 1.7 * QOM memory leak fix # gpg: Signature made Tue 19 Nov 2013 01:58:58 AM PST using RSA key ID 3E7E013F # gpg: Can't check signature: public key not found # By Vlad Yasevich # Via Andreas Färber * afaerber/tags/qom-devices-for-anthony: qom: Fix memory leak in object_property_set_link()
2013-11-19Merge remote-tracking branch 'rth/tcg-ia64-17' into stagingAnthony Liguori
# By Richard Henderson # Via Richard Henderson * rth/tcg-ia64-17: tcg-ia64: Introduce tcg_opc_bswap64_i tcg-ia64: Introduce tcg_opc_ext_i tcg-ia64: Introduce tcg_opc_movi_a tcg-ia64: Introduce tcg_opc_mov_a tcg-ia64: Use A3 form of logical operations tcg-ia64: Use SUB_A3 and ADDS_A4 for subtraction tcg-ia64: Use ADDS for small addition tcg-ia64: Avoid unnecessary stop bit in tcg_out_alu tcg-ia64: Move AREG0 to R32 tcg-ia64: Simplify brcond tcg-ia64: Handle constant calls tcg-ia64: Use shortcuts for nop insns tcg-ia64: Use TCGMemOp within qemu_ldst routines Message-id: 1384811395-7097-1-git-send-email-rth@twiddle.net Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-19Merge remote-tracking branch 'mst/tags/for_anthony' into stagingAnthony Liguori
pc last minute fixes for 1.8 This has a patch that drops an unused FW CFG entry. I think it's best to include it before 1.7 to avoid the need to maintain it in compat machine types. There's also a doc bugfix by Amos: I'm guessing doc fixes are still fair game even at this late stage. Signed-off-by: Michael S. Tsirkin <mst@redhat.com> # gpg: Signature made Mon 18 Nov 2013 03:48:14 AM PST using RSA key ID D28D5469 # gpg: Can't check signature: public key not found # By Amos Kong (1) and Igor Mammedov (1) # Via Michael S. Tsirkin * mst/tags/for_anthony: doc: fix hardcoded helper path pc: disable pci-info Message-id: 1384775449-6693-1-git-send-email-mst@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-19Merge remote-tracking branch 'mjt/trivial-patches' into stagingAnthony Liguori
# By Jan Krupa (4) and others # Via Michael Tokarev * mjt/trivial-patches: hw/i386/Makefile.obj: use $(PYTHON) to run .py scripts consistently configure: Use -B switch only for Python versions which support it qga: Fix shutdown command of guest agent to work with SysV console: Remove unused debug code qga: Fix compilation for old versions of MinGW .travis.yml: basic compile and check recipes pci-assign: Fix error_report of pci-stub message qapi: Fix comment for create-type to match code. vl: fix build when configured with no graphic support usb: drop unused USBNetState.inpkt field qemu-char: add missing characters used in keymaps qemu-char: add support for U-prefixed symbols qemu-char: add Czech keymap file qemu-char: add Czech characters to VNC keysyms Message-id: 1384684850-6777-1-git-send-email-mjt@msgid.tls.msk.ru Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-19Merge remote-tracking branch 'kwolf/tags/for-anthony' into stagingAnthony Liguori
Block fixes for 1.7.0 # gpg: Signature made Fri 15 Nov 2013 09:51:25 AM PST using RSA key ID C88F2FD6 # gpg: Can't check signature: public key not found # By Max Reitz (3) and others # Via Kevin Wolf * kwolf/tags/for-anthony: block: Fail if requested driver is not available MAINTAINERS: add block driver sub-maintainers qemu-img: Fix overwriting 'ret' before using qemu-iotests: Test qcow2 count_contiguous_clusters() qcow2: fix possible corruption when reading multiple clusters qmp: access the local QemuOptsLists for drive option MAINTAINERS: add block tree repo URLs qemu-iotests: Extend 041 for unbacked mirroring block/drive-mirror: Check for NULL backing_hd qapi-schema: Update description for NewImageMode block: Print its file name if backing file opening failed Message-id: 1384537999-5972-1-git-send-email-kwolf@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-19Merge remote-tracking branch 'stefanha/net' into stagingAnthony Liguori
# By Amos Kong (1) and Sebastian Huber (1) # Via Stefan Hajnoczi * stefanha/net: virtio-net: fix the memory leak in rxfilter_notify() smc91c111: Fix receive starvation Message-id: 1384532032-19057-1-git-send-email-stefanha@redhat.com Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-19Merge remote-tracking branch 'agraf/tags/signed-ppc-for-upstream-1.7' into ↵Anthony Liguori
staging Patch queue for ppc - 2013-11-08 These are two patches that will hopefully make it into 1.7. The SLOF update fixes -append kernel command line argument passing into the guest kernel. The other patch makes VIO devices appear when using -device '?'. # gpg: Signature made Thu 07 Nov 2013 07:34:54 PM PST using RSA key ID 03FEDC60 # gpg: Can't check signature: public key not found # By Alexey Kardashevskiy # Via Alexander Graf * agraf/tags/signed-ppc-for-upstream-1.7: pseries: Update SLOF firmware image spapr: add vio-bus devices to categories Message-id: 1383881766-13958-1-git-send-email-agraf@suse.de Signed-off-by: Anthony Liguori <aliguori@amazon.com>
2013-11-19migration: drop MADVISE_DONT_NEED for incoming zero pagesPeter Lieven
The madvise for zeroed out pages was introduced when every transferred zero page was memset to zero and thus allocated. Since commit 211ea740 we check for zeroness of a target page before we memset it to zero. Additionally we memmap target memory so it is essentially zero initialized (except for e.g. option roms and bios which are loaded into target memory although they shouldn't). It was reported recently that this madvise causes a performance degradation in some situations. As the madvise should only be called rarely and if it's called it is likely on a busy page (it was non-zero and changed to zero during migration) drop it completely. Reported-By: Zhang Haoyu <haoyu.zhang@huawei.com> Acked-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Peter Lieven <pl@kamp.de> Signed-off-by: Juan Quintela <quintela@redhat.com>
2013-11-19qom: Fix memory leak in object_property_set_link()Vlad Yasevich
Save the result of the call to object_get_canonical_path() so we can free it. Cc: qemu-stable@nongnu.org Signed-off-by: Vlad Yasevich <vyasevic@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2013-11-19qtest: Use -display none by defaultAndreas Färber
This avoids each test needing to add it to suppress windows popping up. [Commit 7ceeedd016facf8d58e14a0d1417fa7225d71072 ("blockdev-test: add test case for drive_add duplicate IDs") and commit 43cd209803d6cffb1e1a028c9ff2fd0ff4fce954 ("qdev-monitor-test: add device_add leak test cases") added qtest tests without specifying -display none. As a result, "make check" now tries to use graphics (GTK or SDL). Since graphics are not used by the test and inappropriate for headless "make check" runs, add the missing -display none. This fixes "make check" in the QEMU buildbot. -- Stefan] Signed-off-by: Andreas Färber <afaerber@suse.de> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-11-19virtio-net: fix the memory leak in rxfilter_notify()Amos Kong
object_get_canonical_path() returns a gchar*, it should be freed by the caller. Signed-off-by: Amos Kong <akong@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Vlad Yasevich <vyasevic@redhat.com> Reviewed-by: Andreas Färber <afaerber@suse.de> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-11-18doc: fix hardcoded helper pathAmos Kong
The install directory of qemu-bridge-helper is configurable, but we use a fixed path in the documentation. DEFAULT_BRIDGE_HELPER macro isn't available in texi mode, we should always use "/path/to/" prefix for dynamic paths (e.g.: /path/to/image, /path/to/linux, etc). Signed-off-by: Amos Kong <akong@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2013-11-18tcg-ia64: Introduce tcg_opc_bswap64_iRichard Henderson
Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Introduce tcg_opc_ext_iRichard Henderson
Being able to "extend" from 64-bits (with a mov) simplifies a few places where the conditional breaks the train of thought. Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Introduce tcg_opc_movi_aRichard Henderson
Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Introduce tcg_opc_mov_aRichard Henderson
Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Use A3 form of logical operationsRichard Henderson
We can and/or/xor/andcm small constants, saving one cycle. Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Use SUB_A3 and ADDS_A4 for subtractionRichard Henderson
We can subtract from more small constants that just 0 with one insn, and we can add the negative for most small constants. Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Use ADDS for small additionRichard Henderson
Avoids a wasted cycle loading up small constants. Simplify the code assuming the tcg optimizer is going to work and don't expect the first operand of the add to be constant. Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Avoid unnecessary stop bit in tcg_out_aluRichard Henderson
When performing an operation with two input registers, we'd leave the stop bit (and thus an extra cycle) that's only needed when one or the other input is a constant. Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Move AREG0 to R32Richard Henderson
Since the move away from the global areg0, we're no longer globally reserving areg0. Which means our use of R7 clobbers a call-saved register. Shift areg0 into the windowed registers. Indeed, choose the incoming parameter register that it comes to us by. This requires moving the register holding the return address elsewhere. Choose R33 for tidiness. Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>
2013-11-18tcg-ia64: Simplify brcondRichard Henderson
There was a misconception that a stop bit is required between a compare and the branch that uses the predicate set by the compare. This lead to the usage of an extra bundle in which to perform the compare. The extra bundle left room for constants to be loaded for use with the compare insn. If we pack the compare and the branch together in the same bundle, then there's no longer any room for non-zero constants. At which point we can eliminate half the function by not handling them. Acked-by: Aurelien Jarno <aurelien@aurel32.net> Signed-off-by: Richard Henderson <rth@twiddle.net>