public inbox for u-boot@lists.denx.de
 help / color / mirror / Atom feed
From: "Pali Rohár" <pali@kernel.org>
To: Priyanka Jain <priyanka.jain@nxp.com>, Wolfgang Denk <wd@denx.de>,
	Sinan Akman <sinan@writeme.com>
Cc: u-boot@lists.denx.de
Subject: [PATCH] powerpc: mpc85xx: Simplify jump to _start_cont in flash code
Date: Wed, 11 May 2022 20:37:27 +0200	[thread overview]
Message-ID: <20220511183727.1834-1-pali@kernel.org> (raw)

After more patches code for jumping to _start_cont symbol in flash memory
involved to code with useless mathematical operations. Currently it does:

  r3 := CONFIG_SYS_MONITOR_BASE + ABS(_start_cont) - CONFIG_SYS_MONITOR_BASE
  jump to r3

Which is equivalent of just:

  r3 := ABS(_start_cont)
  jump to r3

The purpose of that code is just to jump to _start_code symbol,
independently of program counter. So branch must be done to absolute
address. Trying to write:

  ba _start_cont

just cause linker error:

    LD      u-boot
  powerpc-linux-gnuspe-ld.bfd: arch/powerpc/cpu/mpc85xx/start.o: in function `switch_as':
  (.bootpg+0x4b8): relocation truncated to fit: R_PPC_ADDR24 against symbol `_start_cont' defined in .text section in arch/powerpc/cpu/mpc85xx/start.o
  make: *** [Makefile:1801: u-boot] Error 1

Probably by the fact that absolute address cannot be expressed by 24-bits.
So write the code via mtlr+blr pattern as it was before and load general
purpose register with absolute address of the symbol:

  lis     r3,_start_cont@h
  ori     r3,r3,_start_cont@l
  mtlr    r3
  blr

Seems that gcc and gnu ld linker support symbol@h and symbol@l syntax like
number@h and number@l without any problem. And disassembling of compiler
u-boot binary proved that lis+ori instructions are called with numbers
which represent halves of absolute address of _start_cont symbol.

Signed-off-by: Pali Rohár <pali@kernel.org>
---
 arch/powerpc/cpu/mpc85xx/start.S | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/cpu/mpc85xx/start.S b/arch/powerpc/cpu/mpc85xx/start.S
index 50b23a97662c..6e3900e7923b 100644
--- a/arch/powerpc/cpu/mpc85xx/start.S
+++ b/arch/powerpc/cpu/mpc85xx/start.S
@@ -1211,9 +1211,8 @@ switch_as:
 #else
 	/* Calculate absolute address in FLASH and jump there		*/
 	/*--------------------------------------------------------------*/
-	lis	r3,CONFIG_SYS_MONITOR_BASE@h
-	ori	r3,r3,CONFIG_SYS_MONITOR_BASE@l
-	addi	r3,r3,_start_cont - CONFIG_SYS_MONITOR_BASE
+	lis	r3,_start_cont@h
+	ori	r3,r3,_start_cont@l
 	mtlr	r3
 	blr
 #endif
-- 
2.20.1


             reply	other threads:[~2022-05-11 20:54 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-11 18:37 Pali Rohár [this message]
2022-06-28 15:54 ` [PATCH v2] powerpc: mpc85xx: Simplify jump to _start_cont in flash code Pali Rohár

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=20220511183727.1834-1-pali@kernel.org \
    --to=pali@kernel.org \
    --cc=priyanka.jain@nxp.com \
    --cc=sinan@writeme.com \
    --cc=u-boot@lists.denx.de \
    --cc=wd@denx.de \
    /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