From: Peter Lieven <pl@kamp.de>
To: Paolo Bonzini <pbonzini@redhat.com>, qemu-devel@nongnu.org
Cc: dgilbert@redhat.com, quintela@redhat.com
Subject: Re: [Qemu-devel] [PATCH] migration: cache memory region ram ptr
Date: Sat, 10 May 2014 18:32:21 +0200 [thread overview]
Message-ID: <536E5495.6080401@kamp.de> (raw)
In-Reply-To: <536E46C3.1050502@redhat.com>
Am 10.05.2014 17:33, schrieb Paolo Bonzini:
> Il 10/05/2014 12:51, Peter Lieven ha scritto:
>> we currently look up the ram ptr for each single page. Cache
>> the pointer while we operate on the same block.
>
> Why don't you instead cache the result in the MemoryRegion, so that memory_region_get_ram_ptr becomes a simple, inline field access?
This seems to work. Wondering if it has other side implications. Basic tests like booting vServers and migration work. What about XEN?
diff --git a/include/exec/memory.h b/include/exec/memory.h
index 1d55ad9..3003875 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -161,6 +161,7 @@ struct MemoryRegion {
unsigned ioeventfd_nb;
MemoryRegionIoeventfd *ioeventfds;
NotifierList iommu_notify;
+ void *ram_ptr;
};
/**
diff --git a/memory.c b/memory.c
index 3f1df23..78d4032 100644
--- a/memory.c
+++ b/memory.c
@@ -862,6 +862,7 @@ void memory_region_init(MemoryRegion *mr,
mr->ioeventfd_nb = 0;
mr->ioeventfds = NULL;
mr->flush_coalesced_mmio = false;
+ mr->ram_ptr = NULL;
}
static uint64_t unassigned_mem_read(void *opaque, hwaddr addr,
@@ -1249,7 +1250,12 @@ void *memory_region_get_ram_ptr(MemoryRegion *mr)
assert(mr->terminates);
- return qemu_get_ram_ptr(mr->ram_addr & TARGET_PAGE_MASK);
+ if (mr->ram_ptr) {
+ return mr->ram_ptr;
+ }
+
+ mr->ram_ptr = qemu_get_ram_ptr(mr->ram_addr & TARGET_PAGE_MASK);
+ return mr->ram_ptr;
}
static void memory_region_update_coalesced_range_as(MemoryRegion *mr, AddressSpace *as)
Peter
>
> Paolo
>
>> Signed-off-by: Peter Lieven <pl@kamp.de>
>> ---
>> arch_init.c | 23 ++++++++++++++++-------
>> 1 file changed, 16 insertions(+), 7 deletions(-)
>>
>> diff --git a/arch_init.c b/arch_init.c
>> index 582b716..ce338aa 100644
>> --- a/arch_init.c
>> +++ b/arch_init.c
>> @@ -594,13 +594,19 @@ static int ram_save_block(QEMUFile *f, bool last_stage)
>> ram_bulk_stage = false;
>> }
>> } else {
>> + static uint8_t *ram_ptr;
>> int ret;
>> uint8_t *p;
>> bool send_async = true;
>> - int cont = (block == last_sent_block) ?
>> - RAM_SAVE_FLAG_CONTINUE : 0;
>> + int cont = 0;
>>
>> - p = memory_region_get_ram_ptr(mr) + offset;
>> + if (block != last_sent_block) {
>> + ram_ptr = memory_region_get_ram_ptr(mr);
>> + } else {
>> + cont = RAM_SAVE_FLAG_CONTINUE;
>> + }
>> +
>> + p = ram_ptr + offset;
>>
>> /* In doubt sent page as normal */
>> bytes_sent = -1;
>> @@ -990,16 +996,17 @@ static inline void *host_from_stream_offset(QEMUFile *f,
>> int flags)
>> {
>> static RAMBlock *block = NULL;
>> + static uint8_t *ram_ptr;
>> char id[256];
>> uint8_t len;
>>
>> if (flags & RAM_SAVE_FLAG_CONTINUE) {
>> - if (!block) {
>> + if (!block || !ram_ptr) {
>> fprintf(stderr, "Ack, bad migration stream!\n");
>> return NULL;
>> }
>>
>> - return memory_region_get_ram_ptr(block->mr) + offset;
>> + return ram_ptr + offset;
>> }
>>
>> len = qemu_get_byte(f);
>> @@ -1007,8 +1014,10 @@ static inline void *host_from_stream_offset(QEMUFile *f,
>> id[len] = 0;
>>
>> QTAILQ_FOREACH(block, &ram_list.blocks, next) {
>> - if (!strncmp(id, block->idstr, sizeof(id)))
>> - return memory_region_get_ram_ptr(block->mr) + offset;
>> + if (!strncmp(id, block->idstr, sizeof(id))) {
>> + ram_ptr = memory_region_get_ram_ptr(block->mr);
>> + return ram_ptr + offset;
>> + }
>> }
>>
>> fprintf(stderr, "Can't find block %s!\n", id);
>>
>
next prev parent reply other threads:[~2014-05-10 16:32 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-10 10:51 [Qemu-devel] [PATCH] migration: cache memory region ram ptr Peter Lieven
2014-05-10 11:38 ` 陈梁
2014-05-10 13:13 ` Peter Lieven
2014-05-10 15:33 ` Paolo Bonzini
2014-05-10 16:07 ` Peter Lieven
2014-05-10 16:32 ` Peter Lieven [this message]
2014-05-10 16:44 ` Peter Lieven
2014-05-12 6:09 ` Paolo Bonzini
2014-05-12 8:16 ` Peter Lieven
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=536E5495.6080401@kamp.de \
--to=pl@kamp.de \
--cc=dgilbert@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).