| Message ID | 20170407095034.27908-1-vigneshr@ti.com |
|---|---|
| State | Changes Requested |
| Delegated to: | Tom Rini |
| Headers | show |
On Fri, Apr 07, 2017 at 03:20:34PM +0530, Vignesh R wrote: > Add support for %p, %pa[p], %pM, %pm and %pI4 formats to tiny-printf. > %pM and %pI4 are widely used by SPL networking stack and is required if > networking support is desired in SPL. > > Before this patch: > $ size spl/u-boot-spl > text data bss dec hex filename > 99325 4899 218584 322808 4ecf8 spl/u-boot-spl > > After this patch (with CONFIG_SPL_NET_SUPPORT): > $ size spl/u-boot-spl > text data bss dec hex filename > 99714 4899 218584 323197 4ee7d spl/u-boot-spl > > So, this patch adds ~390 bytes to code size. > > If CONFIG_SPL_NET_SUPPORT is not enabled then only %p, %pa, %pap are > supported, this adds ~90 bytes to code size. Why do we need %p/%pa/%pap? I'm fine with adding %pM/%pm/%pI4 under SPL_NET_SUPPORT as you've done.
On 4/7/2017 8:40 PM, Tom Rini wrote: > On Fri, Apr 07, 2017 at 03:20:34PM +0530, Vignesh R wrote: > >> Add support for %p, %pa[p], %pM, %pm and %pI4 formats to tiny-printf. >> %pM and %pI4 are widely used by SPL networking stack and is required if >> networking support is desired in SPL. >> >> Before this patch: >> $ size spl/u-boot-spl >> text data bss dec hex filename >> 99325 4899 218584 322808 4ecf8 spl/u-boot-spl >> >> After this patch (with CONFIG_SPL_NET_SUPPORT): >> $ size spl/u-boot-spl >> text data bss dec hex filename >> 99714 4899 218584 323197 4ee7d spl/u-boot-spl >> >> So, this patch adds ~390 bytes to code size. >> >> If CONFIG_SPL_NET_SUPPORT is not enabled then only %p, %pa, %pap are >> supported, this adds ~90 bytes to code size. > > Why do we need %p/%pa/%pap? I'm fine with adding %pM/%pm/%pI4 under > SPL_NET_SUPPORT as you've done. > Ok, I can drop support for %p/%pa/%pap. Its just that, I see debug() prints in SPL code that use them. If the agreement is not to worry them, I am fine with it.
On Fri, Apr 07, 2017 at 11:42:56PM +0530, Vignesh R wrote: > > > On 4/7/2017 8:40 PM, Tom Rini wrote: > > On Fri, Apr 07, 2017 at 03:20:34PM +0530, Vignesh R wrote: > > > >> Add support for %p, %pa[p], %pM, %pm and %pI4 formats to tiny-printf. > >> %pM and %pI4 are widely used by SPL networking stack and is required if > >> networking support is desired in SPL. > >> > >> Before this patch: > >> $ size spl/u-boot-spl > >> text data bss dec hex filename > >> 99325 4899 218584 322808 4ecf8 spl/u-boot-spl > >> > >> After this patch (with CONFIG_SPL_NET_SUPPORT): > >> $ size spl/u-boot-spl > >> text data bss dec hex filename > >> 99714 4899 218584 323197 4ee7d spl/u-boot-spl > >> > >> So, this patch adds ~390 bytes to code size. > >> > >> If CONFIG_SPL_NET_SUPPORT is not enabled then only %p, %pa, %pap are > >> supported, this adds ~90 bytes to code size. > > > > Why do we need %p/%pa/%pap? I'm fine with adding %pM/%pm/%pI4 under > > SPL_NET_SUPPORT as you've done. > > Ok, I can drop support for %p/%pa/%pap. Its just that, I see debug() > prints in SPL code that use them. If the agreement is not to worry them, > I am fine with it. debug prints, eh? Guard the support under #ifdef DEBUG ?
Hi Vignesh, On 7 April 2017 at 12:12, Vignesh R <vigneshr@ti.com> wrote: > > > On 4/7/2017 8:40 PM, Tom Rini wrote: >> On Fri, Apr 07, 2017 at 03:20:34PM +0530, Vignesh R wrote: >> >>> Add support for %p, %pa[p], %pM, %pm and %pI4 formats to tiny-printf. >>> %pM and %pI4 are widely used by SPL networking stack and is required if >>> networking support is desired in SPL. >>> >>> Before this patch: >>> $ size spl/u-boot-spl >>> text data bss dec hex filename >>> 99325 4899 218584 322808 4ecf8 spl/u-boot-spl >>> >>> After this patch (with CONFIG_SPL_NET_SUPPORT): >>> $ size spl/u-boot-spl >>> text data bss dec hex filename >>> 99714 4899 218584 323197 4ee7d spl/u-boot-spl >>> >>> So, this patch adds ~390 bytes to code size. >>> >>> If CONFIG_SPL_NET_SUPPORT is not enabled then only %p, %pa, %pap are >>> supported, this adds ~90 bytes to code size. >> >> Why do we need %p/%pa/%pap? I'm fine with adding %pM/%pm/%pI4 under >> SPL_NET_SUPPORT as you've done. >> > > Ok, I can drop support for %p/%pa/%pap. Its just that, I see debug() > prints in SPL code that use them. If the agreement is not to worry them, > I am fine with it. With the next rev can you also please compare the code size with and without tiny printf()? It's good to know that we still have a wide margin! Regards, Simon
diff --git a/lib/tiny-printf.c b/lib/tiny-printf.c index 6def8f98aa41..e524ffb657e6 100644 --- a/lib/tiny-printf.c +++ b/lib/tiny-printf.c @@ -12,6 +12,7 @@ #include <common.h> #include <stdarg.h> #include <serial.h> +#include <linux/ctype.h> struct printf_info { char *bf; /* Digit buffer */ @@ -52,6 +53,148 @@ static void div_out(struct printf_info *info, unsigned long *num, out_dgt(info, dgt); } +#ifdef CONFIG_SPL_NET_SUPPORT +static void string(struct printf_info *info, char *s) +{ + char ch; + + while ((ch = *s++)) + out(info, ch); +} + +static const char hex_asc[] = "0123456789abcdef"; +#define hex_asc_lo(x) hex_asc[((x) & 0x0f)] +#define hex_asc_hi(x) hex_asc[((x) & 0xf0) >> 4] + +static inline char *pack_hex_byte(char *buf, u8 byte) +{ + *buf++ = hex_asc_hi(byte); + *buf++ = hex_asc_lo(byte); + return buf; +} + +static void mac_address_string(struct printf_info *info, u8 *addr, + bool separator) +{ + /* (6 * 2 hex digits), 5 colons and trailing zero */ + char mac_addr[6 * 3]; + char *p = mac_addr; + int i; + + for (i = 0; i < 6; i++) { + p = pack_hex_byte(p, addr[i]); + if (separator && i != 5) + *p++ = ':'; + } + *p = '\0'; + + string(info, mac_addr); +} + +static char *put_dec_trunc(char *buf, unsigned int q) +{ + unsigned int d3, d2, d1, d0; + d1 = (q >> 4) & 0xf; + d2 = (q >> 8) & 0xf; + d3 = (q >> 12); + + d0 = 6 * (d3 + d2 + d1) + (q & 0xf); + q = (d0 * 0xcd) >> 11; + d0 = d0 - 10 * q; + *buf++ = d0 + '0'; /* least significant digit */ + d1 = q + 9 * d3 + 5 * d2 + d1; + if (d1 != 0) { + q = (d1 * 0xcd) >> 11; + d1 = d1 - 10 * q; + *buf++ = d1 + '0'; /* next digit */ + + d2 = q + 2 * d2; + if ((d2 != 0) || (d3 != 0)) { + q = (d2 * 0xd) >> 7; + d2 = d2 - 10 * q; + *buf++ = d2 + '0'; /* next digit */ + + d3 = q + 4 * d3; + if (d3 != 0) { + q = (d3 * 0xcd) >> 11; + d3 = d3 - 10 * q; + *buf++ = d3 + '0'; /* next digit */ + if (q != 0) + *buf++ = q + '0'; /* most sign. digit */ + } + } + } + return buf; +} + +static void ip4_addr_string(struct printf_info *info, u8 *addr) +{ + /* (4 * 3 decimal digits), 3 dots and trailing zero */ + char ip4_addr[4 * 4]; + char temp[3]; /* hold each IP quad in reverse order */ + char *p = ip4_addr; + int i, digits; + + for (i = 0; i < 4; i++) { + digits = put_dec_trunc(temp, addr[i]) - temp; + /* reverse the digits in the quad */ + while (digits--) + *p++ = temp[digits]; + if (i != 3) + *p++ = '.'; + } + *p = '\0'; + + string(info, ip4_addr); +} +#endif + +/* + * Show a '%p' thing. A kernel extension is that the '%p' is followed + * by an extra set of characters that are extended format + * specifiers. + * + * Right now we handle: + * + * - 'M' For a 6-byte MAC address, it prints the address in the + * usual colon-separated hex notation. + * - 'm' Same as above except there is no colon-separator. + * - 'I4'for IPv4 addresses printed in the usual way (dot-separated + * decimal). + */ + +static void pointer(struct printf_info *info, const char *fmt, void *ptr) +{ + unsigned long num = (uintptr_t)ptr; + unsigned long div; + + switch (*fmt) { + case 'a': + + switch (fmt[1]) { + case 'p': + default: + num = *(phys_addr_t *)ptr; + break; + } + break; +#ifdef CONFIG_SPL_NET_SUPPORT + case 'm': + return mac_address_string(info, ptr, false); + case 'M': + return mac_address_string(info, ptr, true); + case 'I': + if (fmt[1] == '4') + return ip4_addr_string(info, ptr); +#endif + default: + break; + } + div = 1UL << (sizeof(long) * 8 - 4); + for (; div; div /= 0x10) + div_out(info, &num, div); +} + static int _vprintf(struct printf_info *info, const char *fmt, va_list va) { char ch; @@ -144,6 +287,11 @@ static int _vprintf(struct printf_info *info, const char *fmt, va_list va) case 's': p = va_arg(va, char*); break; + case 'p': + pointer(info, fmt, va_arg(va, void *)); + while (isalnum(fmt[0])) + fmt++; + break; case '%': out(info, '%'); default:
Add support for %p, %pa[p], %pM, %pm and %pI4 formats to tiny-printf. %pM and %pI4 are widely used by SPL networking stack and is required if networking support is desired in SPL. Before this patch: $ size spl/u-boot-spl text data bss dec hex filename 99325 4899 218584 322808 4ecf8 spl/u-boot-spl After this patch (with CONFIG_SPL_NET_SUPPORT): $ size spl/u-boot-spl text data bss dec hex filename 99714 4899 218584 323197 4ee7d spl/u-boot-spl So, this patch adds ~390 bytes to code size. If CONFIG_SPL_NET_SUPPORT is not enabled then only %p, %pa, %pap are supported, this adds ~90 bytes to code size. Compiler used is: arm-linux-gnueabihf-gcc (Linaro GCC 6.2-2016.11) 6.2.1 20161016 Signed-off-by: Vignesh R <vigneshr@ti.com> --- lib/tiny-printf.c | 148 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 148 insertions(+)