public inbox for bpf@vger.kernel.org
 help / color / mirror / Atom feed
From: Ilya Leoshkevich <iii@linux.ibm.com>
To: Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Andrii Nakryiko <andrii@kernel.org>
Cc: bpf@vger.kernel.org, Heiko Carstens <hca@linux.ibm.com>,
	Vasily Gorbik <gor@linux.ibm.com>,
	Alexander Gordeev <agordeev@linux.ibm.com>,
	Ilya Leoshkevich <iii@linux.ibm.com>
Subject: [PATCH bpf-next v3 5/8] s390/bpf: Implement bpf_jit_supports_subprog_tailcalls()
Date: Sun, 29 Jan 2023 20:04:58 +0100	[thread overview]
Message-ID: <20230129190501.1624747-6-iii@linux.ibm.com> (raw)
In-Reply-To: <20230129190501.1624747-1-iii@linux.ibm.com>

Allow mixing subprogs and tail calls by passing the current tail
call count to subprogs.

Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
---
 arch/s390/net/bpf_jit_comp.c | 37 ++++++++++++++++++++++++++----------
 1 file changed, 27 insertions(+), 10 deletions(-)

diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c
index 3fe082d5328d..b84c7ddb758a 100644
--- a/arch/s390/net/bpf_jit_comp.c
+++ b/arch/s390/net/bpf_jit_comp.c
@@ -58,7 +58,6 @@ struct bpf_jit {
 #define SEEN_MEM	BIT(0)		/* use mem[] for temporary storage */
 #define SEEN_LITERAL	BIT(1)		/* code uses literals */
 #define SEEN_FUNC	BIT(2)		/* calls C functions */
-#define SEEN_TAIL_CALL	BIT(3)		/* code uses tail calls */
 #define SEEN_STACK	(SEEN_FUNC | SEEN_MEM)
 
 /*
@@ -549,20 +548,23 @@ static void bpf_jit_plt(void *plt, void *ret, void *target)
  * Save registers and create stack frame if necessary.
  * See stack frame layout description in "bpf_jit.h"!
  */
-static void bpf_jit_prologue(struct bpf_jit *jit, u32 stack_depth)
+static void bpf_jit_prologue(struct bpf_jit *jit, struct bpf_prog *fp,
+			     u32 stack_depth)
 {
 	/* No-op for hotpatching */
 	/* brcl 0,prologue_plt */
 	EMIT6_PCREL_RILC(0xc0040000, 0, jit->prologue_plt);
 	jit->prologue_plt_ret = jit->prg;
 
-	if (jit->seen & SEEN_TAIL_CALL) {
+	if (fp->aux->func_idx == 0) {
+		/* Initialize the tail call counter in the main program. */
 		/* xc STK_OFF_TCCNT(4,%r15),STK_OFF_TCCNT(%r15) */
 		_EMIT6(0xd703f000 | STK_OFF_TCCNT, 0xf000 | STK_OFF_TCCNT);
 	} else {
 		/*
-		 * There are no tail calls. Insert nops in order to have
-		 * tail_call_start at a predictable offset.
+		 * Skip the tail call counter initialization in subprograms.
+		 * Insert nops in order to have tail_call_start at a
+		 * predictable offset.
 		 */
 		bpf_skip(jit, 6);
 	}
@@ -1410,6 +1412,19 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp,
 
 		REG_SET_SEEN(BPF_REG_5);
 		jit->seen |= SEEN_FUNC;
+		/*
+		 * Copy the tail call counter to where the callee expects it.
+		 *
+		 * Note 1: The callee can increment the tail call counter, but
+		 * we do not load it back, since the x86 JIT does not do this
+		 * either.
+		 *
+		 * Note 2: We assume that the verifier does not let us call the
+		 * main program, which clears the tail call counter on entry.
+		 */
+		/* mvc STK_OFF_TCCNT(4,%r15),N(%r15) */
+		_EMIT6(0xd203f000 | STK_OFF_TCCNT,
+		       0xf000 | (STK_OFF_TCCNT + STK_OFF + stack_depth));
 		/* lgrl %w1,func */
 		EMIT6_PCREL_RILB(0xc4080000, REG_W1, _EMIT_CONST_U64(func));
 		/* %r1() */
@@ -1426,10 +1441,7 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp,
 		 *  B1: pointer to ctx
 		 *  B2: pointer to bpf_array
 		 *  B3: index in bpf_array
-		 */
-		jit->seen |= SEEN_TAIL_CALL;
-
-		/*
+		 *
 		 * if (index >= array->map.max_entries)
 		 *         goto out;
 		 */
@@ -1793,7 +1805,7 @@ static int bpf_jit_prog(struct bpf_jit *jit, struct bpf_prog *fp,
 	jit->prg = 0;
 	jit->excnt = 0;
 
-	bpf_jit_prologue(jit, stack_depth);
+	bpf_jit_prologue(jit, fp, stack_depth);
 	if (bpf_set_addr(jit, 0) < 0)
 		return -1;
 	for (i = 0; i < fp->len; i += insn_count) {
@@ -2462,3 +2474,8 @@ int arch_prepare_bpf_trampoline(struct bpf_tramp_image *im, void *image,
 
 	return ret;
 }
+
+bool bpf_jit_supports_subprog_tailcalls(void)
+{
+	return true;
+}
-- 
2.39.1


  parent reply	other threads:[~2023-01-29 19:05 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-29 19:04 [PATCH bpf-next v3 0/8] Support bpf trampoline for s390x Ilya Leoshkevich
2023-01-29 19:04 ` [PATCH bpf-next v3 1/8] selftests/bpf: Fix sk_assign on s390x Ilya Leoshkevich
2023-01-29 19:04 ` [PATCH bpf-next v3 2/8] s390/bpf: Add expoline to tail calls Ilya Leoshkevich
2023-01-29 19:04 ` [PATCH bpf-next v3 3/8] s390/bpf: Implement bpf_arch_text_poke() Ilya Leoshkevich
2023-01-29 19:04 ` [PATCH bpf-next v3 4/8] s390/bpf: Implement arch_prepare_bpf_trampoline() Ilya Leoshkevich
2023-01-29 19:04 ` Ilya Leoshkevich [this message]
2023-01-29 19:04 ` [PATCH bpf-next v3 6/8] s390/bpf: Implement bpf_jit_supports_kfunc_call() Ilya Leoshkevich
2023-01-29 19:05 ` [PATCH bpf-next v3 7/8] selftests/bpf: Fix s390x vmlinux path Ilya Leoshkevich
2023-01-29 19:05 ` [PATCH bpf-next v3 8/8] selftests/bpf: Trim DENYLIST.s390x Ilya Leoshkevich
2023-01-30  3:28 ` [PATCH bpf-next v3 0/8] Support bpf trampoline for s390x Alexei Starovoitov
2023-01-30 18:56   ` [PATCH bpf-next v3 0/8] Support bpf trampoline for s390x - CI issue Ilya Leoshkevich
2023-01-31  3:13     ` Alexei Starovoitov
2023-01-31 13:36       ` Ilya Leoshkevich
2023-02-02 18:29         ` Joanne Koong
2023-01-30  3:30 ` [PATCH bpf-next v3 0/8] Support bpf trampoline for s390x patchwork-bot+netdevbpf

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=20230129190501.1624747-6-iii@linux.ibm.com \
    --to=iii@linux.ibm.com \
    --cc=agordeev@linux.ibm.com \
    --cc=andrii@kernel.org \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=gor@linux.ibm.com \
    --cc=hca@linux.ibm.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