Skip to content

Q
qemu

openeuler / qemu

通知 10
Star 0
Fork 0
Q
qemu
切换分支/标签
  1. 30 1月, 2014 1 次提交
  3. 13 1月, 2014 13 次提交
  5. 31 12月, 2013 1 次提交
  7. 20 12月, 2013 1 次提交
    • A
      roms: Flush icache when writing roms to guest memory · 582b55a9
      由 Alexander Graf 提交于 
      We use the rom infrastructure to write firmware and/or initial kernel
blobs into guest address space. So we're basically emulating the cache
off phase on very early system bootup.

That phase is usually responsible for clearing the instruction cache for
anything it writes into cachable memory, to ensure that after reboot we
don't happen to execute stale bits from the instruction cache.

So we need to invalidate the icache every time we write a rom into guest
address space. We do not need to do this for every DMA since the guest
expects it has to flush the icache manually in that case.

This fixes random reboot issues on e5500 (booke ppc) for me.
Signed-off-by: NAlexander Graf <agraf@suse.de>
      582b55a9
  9. 13 12月, 2013 1 次提交
    • A
      roms: Flush icache when writing roms to guest memory · a94b36dd
      由 Alexander Graf 提交于 
      We use the rom infrastructure to write firmware and/or initial kernel
blobs into guest address space. So we're basically emulating the cache
off phase on very early system bootup.

That phase is usually responsible for clearing the instruction cache for
anything it writes into cachable memory, to ensure that after reboot we
don't happen to execute stale bits from the instruction cache.

So we need to invalidate the icache every time we write a rom into guest
address space. We do not need to do this for every DMA since the guest
expects it has to flush the icache manually in that case.

This fixes random reboot issues on e5500 (booke ppc) for me.
Signed-off-by: NAlexander Graf <agraf@suse.de>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
      a94b36dd
  11. 12 12月, 2013 1 次提交
  13. 10 12月, 2013 7 次提交
    • M
      exec: reduce L2_PAGE_SIZE · 026736ce
      由 Michael S. Tsirkin 提交于 
      With the single exception of ppc with 16M pages,
we get the same number of levels
with L2_PAGE_SIZE = 10 as with L2_PAGE_SIZE = 9.

by doing this we reduce memory footprint of a single level
in the node memory map by 2x without runtime overhead.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
      026736ce
    • P
      exec: make address spaces 64-bit wide · 57271d63
      由 Paolo Bonzini 提交于 
      As an alternative to commit 818f86b8 (exec: limit system memory
size, 2013-11-04) let's just make all address spaces 64-bit wide.
This eliminates problems with phys_page_find ignoring bits above
TARGET_PHYS_ADDR_SPACE_BITS and address_space_translate_internal
consequently messing up the computations.

In Luiz's reported crash, at startup gdb attempts to read from address
0xffffffffffffffe6 to 0xffffffffffffffff inclusive.  The region it gets
is the newly introduced master abort region, which is as big as the PCI
address space (see pci_bus_init).  Due to a typo that's only 2^63-1,
not 2^64.  But we get it anyway because phys_page_find ignores the upper
bits of the physical address.  In address_space_translate_internal then

    diff = int128_sub(section->mr->size, int128_make64(addr));
    *plen = int128_get64(int128_min(diff, int128_make64(*plen)));

diff becomes negative, and int128_get64 booms.

The size of the PCI address space region should be fixed anyway.
Reported-by: NLuiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
      57271d63
    • M
      exec: memory radix tree page level compression · b35ba30f
      由 Michael S. Tsirkin 提交于 
      At the moment, memory radix tree is already variable width, but it can
only skip the low bits of address.

This is efficient if we have huge memory regions but inefficient if we
are only using a tiny portion of the address space.

After we have built up the map, detect
configurations where a single L2 entry is valid.

We then speed up the lookup by skipping one or more levels.
In case any levels were skipped, we might end up in a valid section
instead of erroring out. We handle this by checking that
the address is in range of the resulting section.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
      b35ba30f
    • M
      exec: pass hw address to phys_page_find · 97115a8d
      由 Michael S. Tsirkin 提交于 
      callers always shift by target page bits so let's just do this
internally.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
      97115a8d
    • M
      exec: extend skip field to 6 bit, page entry to 32 bit · 8b795765
      由 Michael S. Tsirkin 提交于 
      Extend skip to 6 bit. As page entry doesn't fit in 16 bit
any longer anyway, extend it to 32 bit.
This doubles node map memory requirements, but follow-up
patches will save this memory.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
      8b795765
    • M
      exec: replace leaf with skip · 9736e55b
      由 Michael S. Tsirkin 提交于 
      In preparation for dynamic radix tree depth support, rename is_leaf
field to skip, telling us how many bits to skip to next level.
Set to 0 for leaf.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
      9736e55b
    • P
      split definitions for exec.c and translate-all.c radix trees · 03f49957
      由 Paolo Bonzini 提交于 
      The exec.c and translate-all.c radix trees are quite different, and
the exec.c one in particular is not limited to the CPU---it can be
used also by devices that do DMA, and in that case the address space
is not limited to TARGET_PHYS_ADDR_SPACE_BITS bits.

We want to make exec.c's radix trees 64-bit wide.  As a first step,
stop sharing the constants between exec.c and translate-all.c.
exec.c gets P_L2_* constants, translate-all.c gets V_L2_*, for
consistency with the existing V_L1_* symbols.  Though actually
in the softmmu case translate-all.c is also indexed by physical
addresses...

This patch has no semantic change.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
      03f49957
  15. 25 11月, 2013 1 次提交
  17. 10 11月, 2013 1 次提交
  19. 08 11月, 2013 1 次提交
  21. 04 11月, 2013 1 次提交
  23. 29 10月, 2013 1 次提交
    • K
      exec: Fix bounce buffer allocation in address_space_map() · e85d9db5
      由 Kevin Wolf 提交于 
      This fixes a regression introduced by commit e3127ae0, which kept the
allocation size of the bounce buffer limited to one page in order to
avoid unbounded allocations (as explained in the commit message of
6d16c2f8), but broke the reporting of the shortened bounce buffer to
the caller. The caller therefore assumes that the full requested size
was provided and causes memory corruption when writing beyond the end of
the actually allocated buffer.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
      e85d9db5
  25. 17 10月, 2013 1 次提交
  27. 14 10月, 2013 1 次提交
    • S
      exec: Fix prototype of phys_mem_set_alloc and related functions · 575ddeb4
      由 Stefan Weil 提交于 
      phys_mem_alloc and its assigned values qemu_anon_ram_alloc and
legacy_s390_alloc must have identical argument lists.

legacy_s390_alloc uses the size parameter to call mmap, so size_t is
good enough for all of them.

This patch fixes compiler errors on i686 Linux hosts:

  CC    alpha-softmmu/exec.o
exec.c:752:51: error:
 initialization from incompatible pointer type [-Werror]
exec.c: In function 'qemu_ram_alloc_from_ptr':
exec.c:1139:32: error:
 comparison of distinct pointer types lacks a cast [-Werror]
exec.c: In function 'qemu_ram_remap':
exec.c:1283:21: error:
 comparison of distinct pointer types lacks a cast [-Werror]
Signed-off-by: NStefan Weil <sw@weilnetz.de>
Reviewed-by: NMarkus Armbruster <armbru@redhat.com>
Message-id: 1380481005-32399-1-git-send-email-sw@weilnetz.de
Signed-off-by: NAnthony Liguori <aliguori@amazon.com>
      575ddeb4
  29. 07 10月, 2013 1 次提交
  31. 03 10月, 2013 1 次提交
  33. 21 9月, 2013 1 次提交
  35. 20 9月, 2013 1 次提交
    • A
      exec: always use MADV_DONTFORK · 3e469dbf
      由 Andrea Arcangeli 提交于 
      MADV_DONTFORK prevents fork to fail with -ENOMEM if the default
overcommit heuristics decides there's too much anonymous virtual
memory allocated. If the KVM secondary MMU is synchronized with MMU
notifiers or not, doesn't make a difference in that regard.

Secondly it's always more efficient to avoid copying the guest
physical address space in the fork child (so we avoid to mark all the
guest memory readonly in the parent and so we skip the establishment
and teardown of lots of pagetables in the child).

In the common case we can ignore the error if MADV_DONTFORK is not
available. Leave a second invocation that errors out in the KVM path
if MMU notifiers are missing and KVM is enabled, to abort in such
case.
Signed-off-by: NAndrea Arcangeli <aarcange@redhat.com>
Tested-By: NBenoit Canet <benoit@irqsave.net>
Acked-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NGleb Natapov <gleb@redhat.com>
      3e469dbf
  37. 13 9月, 2013 4 次提交