From: "Michael S. Tsirkin" <mst@redhat.com>
To: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Cc: ehabkost@redhat.com, kvm@vger.kernel.org, gleb@kernel.org,
mtosatti@redhat.com, qemu-devel@nongnu.org, stefanha@redhat.com,
imammedo@redhat.com, pbonzini@redhat.com,
dan.j.williams@intel.com, rth@twiddle.net
Subject: Re: [Qemu-devel] [PATCH v3 08/32] exec: allow memory to be allocated from any kind of path
Date: Mon, 12 Oct 2015 13:08:42 +0300 [thread overview]
Message-ID: <20151012100842.GA1681@redhat.com> (raw)
In-Reply-To: <1444535584-18220-9-git-send-email-guangrong.xiao@linux.intel.com>
On Sun, Oct 11, 2015 at 11:52:40AM +0800, Xiao Guangrong wrote:
> Currently file_ram_alloc() is designed for hugetlbfs, however, the memory
> of nvdimm can come from either raw pmem device eg, /dev/pmem, or the file
> locates at DAX enabled filesystem
>
> So this patch let it work on any kind of path
>
> Signed-off-by: Xiao Guangrong <guangrong.xiao@linux.intel.com>
This conflicts with map alloc rework.
Please rebase this on top of my tree.
> ---
> exec.c | 55 ++++++++++++++-----------------------------------------
> 1 file changed, 14 insertions(+), 41 deletions(-)
>
> diff --git a/exec.c b/exec.c
> index 7d90a52..70cb0ef 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -1154,32 +1154,6 @@ void qemu_mutex_unlock_ramlist(void)
> }
>
> #ifdef __linux__
> -
> -#include <sys/vfs.h>
> -
> -#define HUGETLBFS_MAGIC 0x958458f6
> -
> -static long gethugepagesize(const char *path, Error **errp)
> -{
> - struct statfs fs;
> - int ret;
> -
> - do {
> - ret = statfs(path, &fs);
> - } while (ret != 0 && errno == EINTR);
> -
> - if (ret != 0) {
> - error_setg_errno(errp, errno, "failed to get page size of file %s",
> - path);
> - return 0;
> - }
> -
> - if (fs.f_type != HUGETLBFS_MAGIC)
> - fprintf(stderr, "Warning: path not on HugeTLBFS: %s\n", path);
> -
> - return fs.f_bsize;
What this *actually* is trying to warn against is that
mapping a regular file (as opposed to hugetlbfs)
means transparent huge pages don't work.
So I don't think we should drop this warning completely.
Either let's add the nvdimm magic, or simply check the
page size.
> -}
> -
> static void *file_ram_alloc(RAMBlock *block,
> ram_addr_t memory,
> const char *path,
> @@ -1191,22 +1165,21 @@ static void *file_ram_alloc(RAMBlock *block,
> void *ptr;
> void *area = NULL;
> int fd;
> - uint64_t hpagesize;
> + uint64_t pagesize;
> uint64_t total;
> - Error *local_err = NULL;
> size_t offset;
>
> - hpagesize = gethugepagesize(path, &local_err);
> - if (local_err) {
> - error_propagate(errp, local_err);
> + pagesize = qemu_file_get_page_size(path);
> + if (!pagesize) {
> + error_setg(errp, "can't get page size for %s", path);
> goto error;
> }
> - block->mr->align = hpagesize;
> + block->mr->align = pagesize;
>
> - if (memory < hpagesize) {
> + if (memory < pagesize) {
> error_setg(errp, "memory size 0x" RAM_ADDR_FMT " must be equal to "
> - "or larger than huge page size 0x%" PRIx64,
> - memory, hpagesize);
> + "or larger than page size 0x%" PRIx64,
> + memory, pagesize);
> goto error;
> }
>
> @@ -1230,15 +1203,15 @@ static void *file_ram_alloc(RAMBlock *block,
> fd = mkstemp(filename);
> if (fd < 0) {
> error_setg_errno(errp, errno,
> - "unable to create backing store for hugepages");
> + "unable to create backing store for path %s", path);
> g_free(filename);
> goto error;
> }
> unlink(filename);
> g_free(filename);
>
> - memory = ROUND_UP(memory, hpagesize);
> - total = memory + hpagesize;
> + memory = ROUND_UP(memory, pagesize);
> + total = memory + pagesize;
>
> /*
> * ftruncate is not supported by hugetlbfs in older
> @@ -1254,12 +1227,12 @@ static void *file_ram_alloc(RAMBlock *block,
> -1, 0);
> if (ptr == MAP_FAILED) {
> error_setg_errno(errp, errno,
> - "unable to allocate memory range for hugepages");
> + "unable to allocate memory range for path %s", path);
> close(fd);
> goto error;
> }
>
> - offset = QEMU_ALIGN_UP((uintptr_t)ptr, hpagesize) - (uintptr_t)ptr;
> + offset = QEMU_ALIGN_UP((uintptr_t)ptr, pagesize) - (uintptr_t)ptr;
>
> area = mmap(ptr + offset, memory, PROT_READ | PROT_WRITE,
> (block->flags & RAM_SHARED ? MAP_SHARED : MAP_PRIVATE) |
> @@ -1267,7 +1240,7 @@ static void *file_ram_alloc(RAMBlock *block,
> fd, 0);
> if (area == MAP_FAILED) {
> error_setg_errno(errp, errno,
> - "unable to map backing store for hugepages");
> + "unable to map backing store for path %s", path);
> munmap(ptr, total);
> close(fd);
> goto error;
> --
> 1.8.3.1
next prev parent reply other threads:[~2015-10-12 10:08 UTC|newest]
Thread overview: 102+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-11 3:52 [Qemu-devel] [PATCH v3 00/32] implement vNVDIMM Xiao Guangrong
2015-10-10 21:17 ` Dan Williams
2015-10-12 4:33 ` Xiao Guangrong
2015-10-12 16:36 ` Dan Williams
2015-10-13 3:14 ` Xiao Guangrong
2015-10-13 3:38 ` Dan Williams
2015-10-13 5:49 ` Xiao Guangrong
2015-10-13 6:36 ` Dan Williams
2015-10-14 4:03 ` Xiao Guangrong
2015-10-14 19:20 ` Dan Williams
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 01/32] acpi: add aml_derefof Xiao Guangrong
2015-10-13 12:30 ` Igor Mammedov
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 02/32] acpi: add aml_sizeof Xiao Guangrong
2015-10-13 12:33 ` Igor Mammedov
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 03/32] acpi: add aml_create_field Xiao Guangrong
2015-10-13 12:38 ` Igor Mammedov
2015-10-13 16:36 ` Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 04/32] acpi: add aml_mutex, aml_acquire, aml_release Xiao Guangrong
2015-10-13 13:34 ` Igor Mammedov
2015-10-13 16:44 ` Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 05/32] acpi: add aml_concatenate Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 06/32] acpi: add aml_object_type Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 07/32] util: introduce qemu_file_get_page_size() Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 08/32] exec: allow memory to be allocated from any kind of path Xiao Guangrong
2015-10-12 10:08 ` Michael S. Tsirkin [this message]
2015-10-13 3:31 ` Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 09/32] exec: allow file_ram_alloc to work on file Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 10/32] hostmem-file: clean up memory allocation Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 11/32] hostmem-file: use whole file size if possible Xiao Guangrong
2015-10-13 11:50 ` Vladimir Sementsov-Ogievskiy
2015-10-13 16:53 ` Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 12/32] pc-dimm: remove DEFAULT_PC_DIMMSIZE Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 13/32] pc-dimm: make pc_existing_dimms_capacity static and rename it Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 14/32] pc-dimm: drop the prefix of pc-dimm Xiao Guangrong
2015-10-12 16:43 ` Eric Blake
2015-10-13 3:32 ` Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 15/32] stubs: rename qmp_pc_dimm_device_list.c Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 16/32] pc-dimm: rename pc-dimm.c and pc-dimm.h Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 17/32] dimm: abstract dimm device from pc-dimm Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 18/32] dimm: get mapped memory region from DIMMDeviceClass->get_memory_region Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 19/32] dimm: keep the state of the whole backend memory Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 20/32] dimm: introduce realize callback Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 21/32] nvdimm: implement NVDIMM device abstract Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 22/32] nvdimm: init the address region used by NVDIMM ACPI Xiao Guangrong
2015-10-19 6:56 ` Michael S. Tsirkin
2015-10-19 7:27 ` Xiao Guangrong
2015-10-19 7:39 ` Michael S. Tsirkin
2015-10-19 7:44 ` Xiao Guangrong
2015-10-19 9:17 ` Michael S. Tsirkin
2015-10-19 9:46 ` Igor Mammedov
2015-10-19 10:01 ` Xiao Guangrong
2015-10-19 10:34 ` Michael S. Tsirkin
2015-10-19 10:42 ` Igor Mammedov
2015-10-19 17:56 ` Xiao Guangrong
2015-10-20 2:27 ` Xiao Guangrong
2015-10-19 9:18 ` Igor Mammedov
2015-10-19 10:25 ` Michael S. Tsirkin
2015-10-19 17:54 ` Xiao Guangrong
2015-10-19 21:20 ` Michael S. Tsirkin
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 23/32] nvdimm: build ACPI NFIT table Xiao Guangrong
2015-10-12 11:27 ` Michael S. Tsirkin
2015-10-13 5:13 ` Xiao Guangrong
2015-10-13 5:42 ` Michael S. Tsirkin
2015-10-13 6:06 ` Xiao Guangrong
2015-10-12 16:40 ` Dan Williams
2015-10-13 5:17 ` Xiao Guangrong
2015-10-13 6:07 ` Michael S. Tsirkin
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 24/32] nvdimm: init the address region used by DSM method Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 25/32] nvdimm: build ACPI nvdimm devices Xiao Guangrong
2015-10-13 14:39 ` Igor Mammedov
2015-10-13 17:24 ` Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 26/32] nvdimm: save arg3 for NVDIMM device _DSM method Xiao Guangrong
2015-10-19 6:50 ` Michael S. Tsirkin
2015-10-19 7:14 ` Xiao Guangrong
2015-10-19 7:47 ` Michael S. Tsirkin
2015-10-19 7:51 ` Xiao Guangrong
2015-10-11 3:52 ` [Qemu-devel] [PATCH v3 27/32] nvdimm: support DSM_CMD_IMPLEMENTED function Xiao Guangrong
2015-10-14 9:40 ` Stefan Hajnoczi
2015-10-14 14:50 ` Xiao Guangrong
2015-10-14 17:06 ` Eduardo Habkost
2015-10-15 1:43 ` Xiao Guangrong
2015-10-15 15:07 ` Stefan Hajnoczi
2015-10-16 2:30 ` Xiao Guangrong
2015-10-14 9:41 ` Stefan Hajnoczi
2015-10-14 14:52 ` Xiao Guangrong
2015-10-15 15:01 ` Stefan Hajnoczi
2015-10-16 2:32 ` Xiao Guangrong
2015-10-11 3:53 ` [Qemu-devel] [PATCH v3 28/32] nvdimm: support DSM_CMD_NAMESPACE_LABEL_SIZE function Xiao Guangrong
2015-10-11 3:53 ` [Qemu-devel] [PATCH v3 29/32] nvdimm: support DSM_CMD_GET_NAMESPACE_LABEL_DATA Xiao Guangrong
2015-10-11 3:53 ` [Qemu-devel] [PATCH v3 30/32] nvdimm: support DSM_CMD_SET_NAMESPACE_LABEL_DATA Xiao Guangrong
2015-10-11 3:53 ` [Qemu-devel] [PATCH v3 31/32] nvdimm: allow using whole backend memory as pmem Xiao Guangrong
2015-10-11 3:53 ` [Qemu-devel] [PATCH v3 32/32] nvdimm: add maintain info Xiao Guangrong
2015-10-12 2:59 ` [Qemu-devel] [PATCH v3 00/32] implement vNVDIMM Bharata B Rao
2015-10-12 3:06 ` Xiao Guangrong
2015-10-12 8:20 ` Igor Mammedov
2015-10-12 8:21 ` Xiao Guangrong
2015-10-12 11:55 ` Michael S. Tsirkin
2015-10-13 5:29 ` Xiao Guangrong
2015-10-13 5:57 ` Michael S. Tsirkin
2015-10-13 5:52 ` Xiao Guangrong
2015-10-19 6:57 ` Michael S. Tsirkin
2015-10-19 6:56 ` Michael S. Tsirkin
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=20151012100842.GA1681@redhat.com \
--to=mst@redhat.com \
--cc=dan.j.williams@intel.com \
--cc=ehabkost@redhat.com \
--cc=gleb@kernel.org \
--cc=guangrong.xiao@linux.intel.com \
--cc=imammedo@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=mtosatti@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=rth@twiddle.net \
--cc=stefanha@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).