Linux MIPS Architecture development
 help / color / mirror / Atom feed
From: David Daney <ddaney.cavm@gmail.com>
To: "Steven J. Hill" <sjhill@mips.com>
Cc: linux-mips@linux-mips.org, ralf@linux-mips.org, cernekee@gmail.com
Subject: Re: [PATCH][RFC] MIPS: microMIPS: Add support to micro-assembler.
Date: Wed, 30 Jan 2013 10:04:41 -0800	[thread overview]
Message-ID: <510960B9.1060800@gmail.com> (raw)
In-Reply-To: <1359514228-2161-1-git-send-email-sjhill@mips.com>

On 01/29/2013 06:50 PM, Steven J. Hill wrote:
> From: "Steven J. Hill" <sjhill@mips.com>
>
> This adds microMIPS instruction support to the micro-assembler.
>
> Signed-off-by: Steven J. Hill <sjhill@mips.com>
> ---
>   arch/mips/mm/uasm.c |  121 +++++++++++++++++++++++++++++++++------------------
>   1 file changed, 78 insertions(+), 43 deletions(-)
>
> diff --git a/arch/mips/mm/uasm.c b/arch/mips/mm/uasm.c
> index 942ff6c..13f0ca7 100644
> --- a/arch/mips/mm/uasm.c
> +++ b/arch/mips/mm/uasm.c
> @@ -10,6 +10,7 @@
>    * Copyright (C) 2004, 2005, 2006, 2008	 Thiemo Seufer
>    * Copyright (C) 2005, 2007  Maciej W. Rozycki
>    * Copyright (C) 2006  Ralf Baechle (ralf@linux-mips.org)
> + * Copyright (C) 2012 MIPS Technologies, Inc.  All rights reserved.
>    */
>
>   #include <linux/kernel.h>
> @@ -35,27 +36,6 @@ enum fields {
>   	SCIMM = 0x400
>   };
>
> -#define OP_MASK		0x3f
> -#define OP_SH		26
> -#define RS_MASK		0x1f
> -#define RS_SH		21
> -#define RT_MASK		0x1f
> -#define RT_SH		16
> -#define RD_MASK		0x1f
> -#define RD_SH		11
> -#define RE_MASK		0x1f
> -#define RE_SH		6
> -#define IMM_MASK	0xffff
> -#define IMM_SH		0
> -#define JIMM_MASK	0x3ffffff
> -#define JIMM_SH		0
> -#define FUNC_MASK	0x3f
> -#define FUNC_SH		0
> -#define SET_MASK	0x7
> -#define SET_SH		0
> -#define SCIMM_MASK	0xfffff
> -#define SCIMM_SH	6
> -
>   enum opcode {
>   	insn_invalid,
>   	insn_addiu, insn_addu, insn_and, insn_andi, insn_bbit0, insn_bbit1,
> @@ -77,14 +57,17 @@ struct insn {
>   	enum fields fields;
>   };
>
> +/* Register length mask. */
> +#define RX_MASK		0x1f
> +
>   /* This macro sets the non-variable bits of an instruction. */
>   #define M(a, b, c, d, e, f)					\
> -	((a) << OP_SH						\
> -	 | (b) << RS_SH						\
> -	 | (c) << RT_SH						\
> -	 | (d) << RD_SH						\
> -	 | (e) << RE_SH						\
> -	 | (f) << FUNC_SH)
> +	((a) << 26						\
> +	 | (b) << 21						\
> +	 | (c) << 16						\
> +	 | (d) << 11						\
> +	 | (e) << 6						\
> +	 | (f) << 0)
>
>   static struct insn insn_table[] __uasminitdata = {
>   	{ insn_addiu, M(addiu_op, 0, 0, 0, 0, 0), RS | RT | SIMM },
> @@ -158,30 +141,46 @@ static struct insn insn_table[] __uasminitdata = {
>
>   static inline __uasminit u32 build_rs(u32 arg)
>   {
> -	WARN(arg & ~RS_MASK, KERN_WARNING "Micro-assembler field overflow\n");
> +#ifdef CONFIG_CPU_MICROMIPS
> +#define RS_SH		16
> +#else
> +#define RS_SH		21
> +#endif

I wonder if we could get rid of most of these #ifdefs by taking the 
approach used in scripts/recordmcount.{c,h} or scripts/sortextable.{c,h}

That is, move all the code that varies between microMIPS and MIPS to an 
included file.  At the top of that file have a single 
#if...#else...#endif block that defines the constants for each variety.

Then include this file as many times as needed (once or twice) to 
produce the definitions required.

This would allow both a MIPS and microMIPS uasm to exist in the same kernel.

Use Case: Synthesizing VDSO code in a microMIPS kernel.  The kernel 
might be build for microMIPS, but the user space code in the VDSO must 
(by the kernel ABI) be standard MIPS instructions.



>
> -	return (arg & RS_MASK) << RS_SH;
> +	WARN(arg & ~RX_MASK, KERN_WARNING "Micro-assembler field overflow\n");
> +
> +	return (arg & RX_MASK) << RS_SH;
>   }
>
>   static inline __uasminit u32 build_rt(u32 arg)
>   {
> -	WARN(arg & ~RT_MASK, KERN_WARNING "Micro-assembler field overflow\n");
> +#ifdef CONFIG_CPU_MICROMIPS
> +#define RT_SH		21
> +#else
> +#define RT_SH		16
> +#endif
> +
> +	WARN(arg & ~RX_MASK, KERN_WARNING "Micro-assembler field overflow\n");
>
> -	return (arg & RT_MASK) << RT_SH;
> +	return (arg & RX_MASK) << RT_SH;
>   }
>
>   static inline __uasminit u32 build_rd(u32 arg)
>   {
> -	WARN(arg & ~RD_MASK, KERN_WARNING "Micro-assembler field overflow\n");
> +#define RD_SH		11
>
> -	return (arg & RD_MASK) << RD_SH;
> +	WARN(arg & ~RX_MASK, KERN_WARNING "Micro-assembler field overflow\n");
> +
> +	return (arg & RX_MASK) << RD_SH;
>   }
>
>   static inline __uasminit u32 build_re(u32 arg)
>   {
> -	WARN(arg & ~RE_MASK, KERN_WARNING "Micro-assembler field overflow\n");
> +#define RE_SH		6
>
> -	return (arg & RE_MASK) << RE_SH;
> +	WARN(arg & ~RX_MASK, KERN_WARNING "Micro-assembler field overflow\n");
> +
> +	return (arg & RX_MASK) << RE_SH;
>   }
>
>   static inline __uasminit u32 build_simm(s32 arg)
> @@ -194,6 +193,8 @@ static inline __uasminit u32 build_simm(s32 arg)
>
>   static inline __uasminit u32 build_uimm(u32 arg)
>   {
> +#define IMM_MASK	0xffff
> +
>   	WARN(arg & ~IMM_MASK, KERN_WARNING "Micro-assembler field overflow\n");
>
>   	return arg & IMM_MASK;
> @@ -201,32 +202,54 @@ static inline __uasminit u32 build_uimm(u32 arg)
>
>   static inline __uasminit u32 build_bimm(s32 arg)
>   {
> -	WARN(arg > 0x1ffff || arg < -0x20000,
> +#ifdef CONFIG_CPU_MICROMIPS
> +#define BIMM_SHIFT	1
> +#else
> +#define BIMM_SHIFT	2
> +#endif
> +
> +	WARN(arg > (0x1ffff >> (2 - BIMM_SHIFT)) ||
> +	     arg < (-0x20000 >> (2 - BIMM_SHIFT)),
>   	     KERN_WARNING "Micro-assembler field overflow\n");
>
>   	WARN(arg & 0x3, KERN_WARNING "Invalid micro-assembler branch target\n");
>
> -	return ((arg < 0) ? (1 << 15) : 0) | ((arg >> 2) & 0x7fff);
> +	return ((arg < 0) ? (1 << 15) : 0) | ((arg >> BIMM_SHIFT) & 0x7fff);
>   }
>
>   static inline __uasminit u32 build_jimm(u32 arg)
>   {
> -	WARN(arg & ~(JIMM_MASK << 2),
> +#define JIMM_MASK	0x3ffffff
> +#ifdef CONFIG_CPU_MICROMIPS
> +	arg >>= 1;
> +#else
> +	arg >>= 2;
> +#endif
> +
> +	WARN(arg & ~JIMM_MASK,
>   	     KERN_WARNING "Micro-assembler field overflow\n");
>
> -	return (arg >> 2) & JIMM_MASK;
> +	return (arg & JIMM_MASK);
>   }
>
>   static inline __uasminit u32 build_scimm(u32 arg)
>   {
> -	WARN(arg & ~SCIMM_MASK,
> +#ifdef CONFIG_CPU_MICROMIPS
> +#define SCIMM_SH	6
> +#else
> +#define SCIMM_SH	0
> +#endif
> +
> +	WARN(arg & ~(0xfffff >> SCIMM_SH),
>   	     KERN_WARNING "Micro-assembler field overflow\n");
>
> -	return (arg & SCIMM_MASK) << SCIMM_SH;
> +	return (arg & ((0xfffff >> SCIMM_SH) << (SCIMM_SH + 10)));
>   }
>
>   static inline __uasminit u32 build_func(u32 arg)
>   {
> +#define FUNC_MASK	0x3f
> +
>   	WARN(arg & ~FUNC_MASK, KERN_WARNING "Micro-assembler field overflow\n");
>
>   	return arg & FUNC_MASK;
> @@ -234,6 +257,8 @@ static inline __uasminit u32 build_func(u32 arg)
>
>   static inline __uasminit u32 build_set(u32 arg)
>   {
> +#define SET_MASK	0x7
> +
>   	WARN(arg & ~SET_MASK, KERN_WARNING "Micro-assembler field overflow\n");
>
>   	return arg & SET_MASK;
> @@ -245,6 +270,11 @@ static inline __uasminit u32 build_set(u32 arg)
>    */
>   static void __uasminit build_insn(u32 **buf, enum opcode opc, ...)
>   {
> +#if defined(CONFIG_CPU_MICROMIPS) && defined(CONFIG_CPU_LITTLE_ENDIAN)
> +#define OP_SHIFT	16
> +#else
> +#define OP_SHIFT	0
> +#endif
>   	struct insn *ip = NULL;
>   	unsigned int i;
>   	va_list ap;
> @@ -285,7 +315,7 @@ static void __uasminit build_insn(u32 **buf, enum opcode opc, ...)
>   		op |= build_scimm(va_arg(ap, u32));
>   	va_end(ap);
>
> -	**buf = op;
> +	**buf = ((op & 0xffff) << OP_SHIFT) | (op >> OP_SHIFT);
>   	(*buf)++;
>   }
>
> @@ -555,12 +585,17 @@ UASM_EXPORT_SYMBOL(uasm_r_mips_pc16);
>   static inline void __uasminit
>   __resolve_relocs(struct uasm_reloc *rel, struct uasm_label *lab)
>   {
> +#if defined(CONFIG_CPU_MICROMIPS) && defined(CONFIG_CPU_LITTLE_ENDIAN)
> +#define REL_SHIFT	16
> +#else
> +#define REL_SHIFT	0
> +#endif
>   	long laddr = (long)lab->addr;
>   	long raddr = (long)rel->addr;
>
>   	switch (rel->type) {
>   	case R_MIPS_PC16:
> -		*rel->addr |= build_bimm(laddr - (raddr + 4));
> +		*rel->addr |= (build_bimm(laddr - (raddr + 4)) << REL_SHIFT);
>   		break;
>
>   	default:
>

  parent reply	other threads:[~2013-01-30 18:04 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-01-30  2:50 [PATCH][RFC] MIPS: microMIPS: Add support to micro-assembler Steven J. Hill
2013-01-30  2:52 ` Hill, Steven
2013-01-30 18:04 ` David Daney [this message]
2013-02-04 16:06   ` Hill, Steven

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=510960B9.1060800@gmail.com \
    --to=ddaney.cavm@gmail.com \
    --cc=cernekee@gmail.com \
    --cc=linux-mips@linux-mips.org \
    --cc=ralf@linux-mips.org \
    --cc=sjhill@mips.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