Re: [RFC PATCH 1/5] ebpf: Added eBPF initialization by fds and map update.
Hi all, On Thu, Mar 30, 2023 at 10:10 AM Daniel P. Berrangé wrote: > > On Thu, Mar 30, 2023 at 02:53:16PM +0800, Jason Wang wrote: > > On Thu, Mar 30, 2023 at 8:33 AM Andrew Melnychenko > > wrote: > > > > > > Changed eBPF map updates through mmaped array. > > > Mmaped arrays provide direct access to map data. > > > It should omit using bpf_map_update_elem() call, > > > which may require capabilities that are not present. > > > > This requires kernel support, so after this change, eBPF based RSS > > doesn't work on old kernels that only support syscall based map > > updating. > It would not work either - bpf_map_update_elem() requires BPF capabilities. > What kernel version is the cut off ? If I'm not mistaken - something like kernel 5.5<. But in any case - the RSS eBPF program is quite big and would require features like jumpsback. So, on old kernels, it would not load. > > > With regards, > Daniel > -- > |: https://berrange.com -o-https://www.flickr.com/photos/dberrange :| > |: https://libvirt.org -o-https://fstop138.berrange.com :| > |: https://entangle-photo.org-o-https://www.instagram.com/dberrange :| >
Re: [RFC PATCH 1/5] ebpf: Added eBPF initialization by fds and map update.
On Thu, Mar 30, 2023 at 02:53:16PM +0800, Jason Wang wrote: > On Thu, Mar 30, 2023 at 8:33 AM Andrew Melnychenko wrote: > > > > Changed eBPF map updates through mmaped array. > > Mmaped arrays provide direct access to map data. > > It should omit using bpf_map_update_elem() call, > > which may require capabilities that are not present. > > This requires kernel support, so after this change, eBPF based RSS > doesn't work on old kernels that only support syscall based map > updating. What kernel version is the cut off ? With regards, Daniel -- |: https://berrange.com -o-https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o-https://fstop138.berrange.com :| |: https://entangle-photo.org-o-https://www.instagram.com/dberrange :|
Re: [RFC PATCH 1/5] ebpf: Added eBPF initialization by fds and map update.
On Thu, Mar 30, 2023 at 2:53 PM Jason Wang wrote:
>
> On Thu, Mar 30, 2023 at 8:33 AM Andrew Melnychenko wrote:
> >
> > Changed eBPF map updates through mmaped array.
> > Mmaped arrays provide direct access to map data.
> > It should omit using bpf_map_update_elem() call,
> > which may require capabilities that are not present.
>
> This requires kernel support, so after this change, eBPF based RSS
> doesn't work on old kernels that only support syscall based map
> updating.
>
> I think it's better to keep the syscall path and fail the fds passing
> if the kernel doesn't support mmap().
>
> Thanks
>
> >
> > Signed-off-by: Andrew Melnychenko
> > ---
> > ebpf/ebpf_rss-stub.c | 6 +++
> > ebpf/ebpf_rss.c | 120 ++-
> > ebpf/ebpf_rss.h | 10
> > 3 files changed, 113 insertions(+), 23 deletions(-)
> >
> > diff --git a/ebpf/ebpf_rss-stub.c b/ebpf/ebpf_rss-stub.c
> > index e71e229190..8d7fae2ad9 100644
> > --- a/ebpf/ebpf_rss-stub.c
> > +++ b/ebpf/ebpf_rss-stub.c
> > @@ -28,6 +28,12 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
> > return false;
> > }
> >
> > +bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
> > + int config_fd, int toeplitz_fd, int table_fd)
> > +{
> > +return false;
> > +}
> > +
> > bool ebpf_rss_set_all(struct EBPFRSSContext *ctx, struct EBPFRSSConfig
> > *config,
> >uint16_t *indirections_table, uint8_t *toeplitz_key)
> > {
> > diff --git a/ebpf/ebpf_rss.c b/ebpf/ebpf_rss.c
> > index cee658c158..08015fecb1 100644
> > --- a/ebpf/ebpf_rss.c
> > +++ b/ebpf/ebpf_rss.c
> > @@ -27,19 +27,68 @@ void ebpf_rss_init(struct EBPFRSSContext *ctx)
> > {
> > if (ctx != NULL) {
> > ctx->obj = NULL;
> > +ctx->program_fd = -1;
> > }
> > }
> >
> > bool ebpf_rss_is_loaded(struct EBPFRSSContext *ctx)
> > {
> > -return ctx != NULL && ctx->obj != NULL;
> > +return ctx != NULL && (ctx->obj != NULL || ctx->program_fd != -1);
> > +}
> > +
> > +static bool ebpf_rss_mmap(struct EBPFRSSContext *ctx)
> > +{
> > +if (!ebpf_rss_is_loaded(ctx)) {
> > +return false;
> > +}
> > +
> > +ctx->mmap_configuration = mmap(NULL, qemu_real_host_page_size(),
> > + PROT_READ | PROT_WRITE, MAP_SHARED,
> > + ctx->map_configuration, 0);
> > +if (ctx->mmap_configuration == MAP_FAILED) {
> > +trace_ebpf_error("eBPF RSS", "can not mmap eBPF configuration
> > array");
> > +return false;
> > +}
> > +ctx->mmap_toeplitz_key = mmap(NULL, qemu_real_host_page_size(),
> > + PROT_READ | PROT_WRITE, MAP_SHARED,
> > + ctx->map_toeplitz_key, 0);
> > +if (ctx->mmap_toeplitz_key == MAP_FAILED) {
> > +trace_ebpf_error("eBPF RSS", "can not mmap eBPF toeplitz key");
> > +goto toeplitz_fail;
> > +}
> > +ctx->mmap_indirections_table = mmap(NULL, qemu_real_host_page_size(),
> > + PROT_READ | PROT_WRITE, MAP_SHARED,
> > + ctx->map_indirections_table, 0);
> > +if (ctx->mmap_indirections_table == MAP_FAILED) {
> > +trace_ebpf_error("eBPF RSS", "can not mmap eBPF indirection
> > table");
> > +goto indirection_fail;
> > +}
> > +
> > +return true;
> > +
> > +indirection_fail:
> > +munmap(ctx->mmap_toeplitz_key, qemu_real_host_page_size());
> > +toeplitz_fail:
> > +munmap(ctx->mmap_configuration, qemu_real_host_page_size());
> > +return false;
> > +}
> > +
> > +static void ebpf_rss_munmap(struct EBPFRSSContext *ctx)
> > +{
> > +if (!ebpf_rss_is_loaded(ctx)) {
> > +return;
> > +}
> > +
> > +munmap(ctx->mmap_indirections_table, qemu_real_host_page_size());
> > +munmap(ctx->mmap_toeplitz_key, qemu_real_host_page_size());
> > +munmap(ctx->mmap_configuration, qemu_real_host_page_size());
> > }
> >
> > bool ebpf_rss_load(struct EBPFRSSContext *ctx)
> > {
> > struct rss_bpf *rss_bpf_ctx;
> >
> > -if (ctx == NULL) {
> > +if (ctx == NULL || ebpf_rss_is_loaded(ctx)) {
> > return false;
> > }
> >
> > @@ -66,26 +115,51 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
> > ctx->map_toeplitz_key = bpf_map__fd(
> > rss_bpf_ctx->maps.tap_rss_map_toeplitz_key);
> >
> > +if (!ebpf_rss_mmap(ctx)) {
> > +goto error;
> > +}
> > +
> > return true;
> > error:
> > rss_bpf__destroy(rss_bpf_ctx);
> > ctx->obj = NULL;
> > +ctx->program_fd = -1;
> >
> > return false;
> > }
> >
> > -static bool ebpf_rss_set_config(struct EBPFRSSContext *ctx,
> > -struct EBPFRSSConfig *config)
> > +bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
> > + int config_fd, int toeplitz_fd, int table_fd)
> > {
> > -uint32_t map_key = 0;
Re: [RFC PATCH 1/5] ebpf: Added eBPF initialization by fds and map update.
On Thu, Mar 30, 2023 at 8:33 AM Andrew Melnychenko wrote:
>
> Changed eBPF map updates through mmaped array.
> Mmaped arrays provide direct access to map data.
> It should omit using bpf_map_update_elem() call,
> which may require capabilities that are not present.
This requires kernel support, so after this change, eBPF based RSS
doesn't work on old kernels that only support syscall based map
updating.
I think it's better to keep the syscall path and fail the fds passing
if the kernel doesn't support mmap().
Thanks
>
> Signed-off-by: Andrew Melnychenko
> ---
> ebpf/ebpf_rss-stub.c | 6 +++
> ebpf/ebpf_rss.c | 120 ++-
> ebpf/ebpf_rss.h | 10
> 3 files changed, 113 insertions(+), 23 deletions(-)
>
> diff --git a/ebpf/ebpf_rss-stub.c b/ebpf/ebpf_rss-stub.c
> index e71e229190..8d7fae2ad9 100644
> --- a/ebpf/ebpf_rss-stub.c
> +++ b/ebpf/ebpf_rss-stub.c
> @@ -28,6 +28,12 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
> return false;
> }
>
> +bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
> + int config_fd, int toeplitz_fd, int table_fd)
> +{
> +return false;
> +}
> +
> bool ebpf_rss_set_all(struct EBPFRSSContext *ctx, struct EBPFRSSConfig
> *config,
>uint16_t *indirections_table, uint8_t *toeplitz_key)
> {
> diff --git a/ebpf/ebpf_rss.c b/ebpf/ebpf_rss.c
> index cee658c158..08015fecb1 100644
> --- a/ebpf/ebpf_rss.c
> +++ b/ebpf/ebpf_rss.c
> @@ -27,19 +27,68 @@ void ebpf_rss_init(struct EBPFRSSContext *ctx)
> {
> if (ctx != NULL) {
> ctx->obj = NULL;
> +ctx->program_fd = -1;
> }
> }
>
> bool ebpf_rss_is_loaded(struct EBPFRSSContext *ctx)
> {
> -return ctx != NULL && ctx->obj != NULL;
> +return ctx != NULL && (ctx->obj != NULL || ctx->program_fd != -1);
> +}
> +
> +static bool ebpf_rss_mmap(struct EBPFRSSContext *ctx)
> +{
> +if (!ebpf_rss_is_loaded(ctx)) {
> +return false;
> +}
> +
> +ctx->mmap_configuration = mmap(NULL, qemu_real_host_page_size(),
> + PROT_READ | PROT_WRITE, MAP_SHARED,
> + ctx->map_configuration, 0);
> +if (ctx->mmap_configuration == MAP_FAILED) {
> +trace_ebpf_error("eBPF RSS", "can not mmap eBPF configuration
> array");
> +return false;
> +}
> +ctx->mmap_toeplitz_key = mmap(NULL, qemu_real_host_page_size(),
> + PROT_READ | PROT_WRITE, MAP_SHARED,
> + ctx->map_toeplitz_key, 0);
> +if (ctx->mmap_toeplitz_key == MAP_FAILED) {
> +trace_ebpf_error("eBPF RSS", "can not mmap eBPF toeplitz key");
> +goto toeplitz_fail;
> +}
> +ctx->mmap_indirections_table = mmap(NULL, qemu_real_host_page_size(),
> + PROT_READ | PROT_WRITE, MAP_SHARED,
> + ctx->map_indirections_table, 0);
> +if (ctx->mmap_indirections_table == MAP_FAILED) {
> +trace_ebpf_error("eBPF RSS", "can not mmap eBPF indirection table");
> +goto indirection_fail;
> +}
> +
> +return true;
> +
> +indirection_fail:
> +munmap(ctx->mmap_toeplitz_key, qemu_real_host_page_size());
> +toeplitz_fail:
> +munmap(ctx->mmap_configuration, qemu_real_host_page_size());
> +return false;
> +}
> +
> +static void ebpf_rss_munmap(struct EBPFRSSContext *ctx)
> +{
> +if (!ebpf_rss_is_loaded(ctx)) {
> +return;
> +}
> +
> +munmap(ctx->mmap_indirections_table, qemu_real_host_page_size());
> +munmap(ctx->mmap_toeplitz_key, qemu_real_host_page_size());
> +munmap(ctx->mmap_configuration, qemu_real_host_page_size());
> }
>
> bool ebpf_rss_load(struct EBPFRSSContext *ctx)
> {
> struct rss_bpf *rss_bpf_ctx;
>
> -if (ctx == NULL) {
> +if (ctx == NULL || ebpf_rss_is_loaded(ctx)) {
> return false;
> }
>
> @@ -66,26 +115,51 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
> ctx->map_toeplitz_key = bpf_map__fd(
> rss_bpf_ctx->maps.tap_rss_map_toeplitz_key);
>
> +if (!ebpf_rss_mmap(ctx)) {
> +goto error;
> +}
> +
> return true;
> error:
> rss_bpf__destroy(rss_bpf_ctx);
> ctx->obj = NULL;
> +ctx->program_fd = -1;
>
> return false;
> }
>
> -static bool ebpf_rss_set_config(struct EBPFRSSContext *ctx,
> -struct EBPFRSSConfig *config)
> +bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
> + int config_fd, int toeplitz_fd, int table_fd)
> {
> -uint32_t map_key = 0;
> +if (ctx == NULL || ebpf_rss_is_loaded(ctx)) {
> +return false;
> +}
>
> -if (!ebpf_rss_is_loaded(ctx)) {
> +if (program_fd < 0 || config_fd < 0 || toeplitz_fd < 0 || table_fd < 0) {
> return false;
> }
> -if (bpf_map_update_elem(ctx->map_configuration,
> -_key,
[RFC PATCH 1/5] ebpf: Added eBPF initialization by fds and map update.
Changed eBPF map updates through mmaped array.
Mmaped arrays provide direct access to map data.
It should omit using bpf_map_update_elem() call,
which may require capabilities that are not present.
Signed-off-by: Andrew Melnychenko
---
ebpf/ebpf_rss-stub.c | 6 +++
ebpf/ebpf_rss.c | 120 ++-
ebpf/ebpf_rss.h | 10
3 files changed, 113 insertions(+), 23 deletions(-)
diff --git a/ebpf/ebpf_rss-stub.c b/ebpf/ebpf_rss-stub.c
index e71e229190..8d7fae2ad9 100644
--- a/ebpf/ebpf_rss-stub.c
+++ b/ebpf/ebpf_rss-stub.c
@@ -28,6 +28,12 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
return false;
}
+bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
+ int config_fd, int toeplitz_fd, int table_fd)
+{
+return false;
+}
+
bool ebpf_rss_set_all(struct EBPFRSSContext *ctx, struct EBPFRSSConfig *config,
uint16_t *indirections_table, uint8_t *toeplitz_key)
{
diff --git a/ebpf/ebpf_rss.c b/ebpf/ebpf_rss.c
index cee658c158..08015fecb1 100644
--- a/ebpf/ebpf_rss.c
+++ b/ebpf/ebpf_rss.c
@@ -27,19 +27,68 @@ void ebpf_rss_init(struct EBPFRSSContext *ctx)
{
if (ctx != NULL) {
ctx->obj = NULL;
+ctx->program_fd = -1;
}
}
bool ebpf_rss_is_loaded(struct EBPFRSSContext *ctx)
{
-return ctx != NULL && ctx->obj != NULL;
+return ctx != NULL && (ctx->obj != NULL || ctx->program_fd != -1);
+}
+
+static bool ebpf_rss_mmap(struct EBPFRSSContext *ctx)
+{
+if (!ebpf_rss_is_loaded(ctx)) {
+return false;
+}
+
+ctx->mmap_configuration = mmap(NULL, qemu_real_host_page_size(),
+ PROT_READ | PROT_WRITE, MAP_SHARED,
+ ctx->map_configuration, 0);
+if (ctx->mmap_configuration == MAP_FAILED) {
+trace_ebpf_error("eBPF RSS", "can not mmap eBPF configuration array");
+return false;
+}
+ctx->mmap_toeplitz_key = mmap(NULL, qemu_real_host_page_size(),
+ PROT_READ | PROT_WRITE, MAP_SHARED,
+ ctx->map_toeplitz_key, 0);
+if (ctx->mmap_toeplitz_key == MAP_FAILED) {
+trace_ebpf_error("eBPF RSS", "can not mmap eBPF toeplitz key");
+goto toeplitz_fail;
+}
+ctx->mmap_indirections_table = mmap(NULL, qemu_real_host_page_size(),
+ PROT_READ | PROT_WRITE, MAP_SHARED,
+ ctx->map_indirections_table, 0);
+if (ctx->mmap_indirections_table == MAP_FAILED) {
+trace_ebpf_error("eBPF RSS", "can not mmap eBPF indirection table");
+goto indirection_fail;
+}
+
+return true;
+
+indirection_fail:
+munmap(ctx->mmap_toeplitz_key, qemu_real_host_page_size());
+toeplitz_fail:
+munmap(ctx->mmap_configuration, qemu_real_host_page_size());
+return false;
+}
+
+static void ebpf_rss_munmap(struct EBPFRSSContext *ctx)
+{
+if (!ebpf_rss_is_loaded(ctx)) {
+return;
+}
+
+munmap(ctx->mmap_indirections_table, qemu_real_host_page_size());
+munmap(ctx->mmap_toeplitz_key, qemu_real_host_page_size());
+munmap(ctx->mmap_configuration, qemu_real_host_page_size());
}
bool ebpf_rss_load(struct EBPFRSSContext *ctx)
{
struct rss_bpf *rss_bpf_ctx;
-if (ctx == NULL) {
+if (ctx == NULL || ebpf_rss_is_loaded(ctx)) {
return false;
}
@@ -66,26 +115,51 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
ctx->map_toeplitz_key = bpf_map__fd(
rss_bpf_ctx->maps.tap_rss_map_toeplitz_key);
+if (!ebpf_rss_mmap(ctx)) {
+goto error;
+}
+
return true;
error:
rss_bpf__destroy(rss_bpf_ctx);
ctx->obj = NULL;
+ctx->program_fd = -1;
return false;
}
-static bool ebpf_rss_set_config(struct EBPFRSSContext *ctx,
-struct EBPFRSSConfig *config)
+bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
+ int config_fd, int toeplitz_fd, int table_fd)
{
-uint32_t map_key = 0;
+if (ctx == NULL || ebpf_rss_is_loaded(ctx)) {
+return false;
+}
-if (!ebpf_rss_is_loaded(ctx)) {
+if (program_fd < 0 || config_fd < 0 || toeplitz_fd < 0 || table_fd < 0) {
return false;
}
-if (bpf_map_update_elem(ctx->map_configuration,
-_key, config, 0) < 0) {
+
+ctx->program_fd = program_fd;
+ctx->map_configuration = config_fd;
+ctx->map_toeplitz_key = toeplitz_fd;
+ctx->map_indirections_table = table_fd;
+
+if (!ebpf_rss_mmap(ctx)) {
+ctx->program_fd = -1;
return false;
}
+
+return true;
+}
+
+static bool ebpf_rss_set_config(struct EBPFRSSContext *ctx,
+struct EBPFRSSConfig *config)
+{
+if (!ebpf_rss_is_loaded(ctx)) {
+return false;
+}
+
+memcpy(ctx->mmap_configuration, config, sizeof(*config));
return true;
}
[RFC PATCH 1/5] ebpf: Added eBPF initialization by fds.
eBPF RSS context may be initialized by program fd and map fds.
virtio-net may provide fds passed by libvirt.
Signed-off-by: Andrew Melnychenko
---
ebpf/ebpf_rss-stub.c | 6 ++
ebpf/ebpf_rss.c | 31 ---
ebpf/ebpf_rss.h | 5 +
3 files changed, 39 insertions(+), 3 deletions(-)
diff --git a/ebpf/ebpf_rss-stub.c b/ebpf/ebpf_rss-stub.c
index e71e229190..8d7fae2ad9 100644
--- a/ebpf/ebpf_rss-stub.c
+++ b/ebpf/ebpf_rss-stub.c
@@ -28,6 +28,12 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
return false;
}
+bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
+ int config_fd, int toeplitz_fd, int table_fd)
+{
+return false;
+}
+
bool ebpf_rss_set_all(struct EBPFRSSContext *ctx, struct EBPFRSSConfig *config,
uint16_t *indirections_table, uint8_t *toeplitz_key)
{
diff --git a/ebpf/ebpf_rss.c b/ebpf/ebpf_rss.c
index 118c68da83..6e9b88efed 100644
--- a/ebpf/ebpf_rss.c
+++ b/ebpf/ebpf_rss.c
@@ -27,19 +27,20 @@ void ebpf_rss_init(struct EBPFRSSContext *ctx)
{
if (ctx != NULL) {
ctx->obj = NULL;
+ctx->program_fd = -1;
}
}
bool ebpf_rss_is_loaded(struct EBPFRSSContext *ctx)
{
-return ctx != NULL && ctx->obj != NULL;
+return ctx != NULL && (ctx->obj != NULL || ctx->program_fd != -1);
}
bool ebpf_rss_load(struct EBPFRSSContext *ctx)
{
struct rss_bpf *rss_bpf_ctx;
-if (ctx == NULL) {
+if (ctx == NULL || ebpf_rss_is_loaded(ctx)) {
return false;
}
@@ -70,10 +71,26 @@ bool ebpf_rss_load(struct EBPFRSSContext *ctx)
error:
rss_bpf__destroy(rss_bpf_ctx);
ctx->obj = NULL;
+ctx->program_fd = -1;
return false;
}
+bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
+ int config_fd, int toeplitz_fd, int table_fd)
+{
+if (ctx == NULL || ebpf_rss_is_loaded(ctx)) {
+return false;
+}
+
+ctx->program_fd = program_fd;
+ctx->map_configuration = config_fd;
+ctx->map_toeplitz_key = toeplitz_fd;
+ctx->map_indirections_table = table_fd;
+
+return true;
+}
+
static bool ebpf_rss_set_config(struct EBPFRSSContext *ctx,
struct EBPFRSSConfig *config)
{
@@ -160,6 +177,14 @@ void ebpf_rss_unload(struct EBPFRSSContext *ctx)
return;
}
-rss_bpf__destroy(ctx->obj);
+if (ctx->obj != NULL) {
+rss_bpf__destroy(ctx->obj);
+} else {
+close(ctx->program_fd);
+close(ctx->map_configuration);
+close(ctx->map_toeplitz_key);
+close(ctx->map_indirections_table);
+}
ctx->obj = NULL;
+ctx->program_fd = -1;
}
diff --git a/ebpf/ebpf_rss.h b/ebpf/ebpf_rss.h
index bf3f2572c7..363859c8bb 100644
--- a/ebpf/ebpf_rss.h
+++ b/ebpf/ebpf_rss.h
@@ -14,6 +14,8 @@
#ifndef QEMU_EBPF_RSS_H
#define QEMU_EBPF_RSS_H
+#define EBPF_RSS_MAX_FDS 4
+
struct EBPFRSSContext {
void *obj;
int program_fd;
@@ -36,6 +38,9 @@ bool ebpf_rss_is_loaded(struct EBPFRSSContext *ctx);
bool ebpf_rss_load(struct EBPFRSSContext *ctx);
+bool ebpf_rss_load_fds(struct EBPFRSSContext *ctx, int program_fd,
+ int config_fd, int toeplitz_fd, int table_fd);
+
bool ebpf_rss_set_all(struct EBPFRSSContext *ctx, struct EBPFRSSConfig *config,
uint16_t *indirections_table, uint8_t *toeplitz_key);
--
2.31.1
