Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
On Sun, 14 Oct 2007 15:58:26 +0100, Denys Vlasenko <[EMAIL PROTECTED]> wrote: > Hi, > > Following patches debloat drivers/scsi/aic7xxx/*. > I also had to add prototypes for ahc_lookup_scb > and ahd_lookup_scb to .h files. > Sorry for the late replay, but.. working fine on annwn:~# lspci | grep Adap 03:02.0 SCSI storage controller: Adaptec ASC-29320 U320 (rev 03) 03:02.1 SCSI storage controller: Adaptec ASC-29320 U320 (rev 03) annwn:~# lsscsi -H [1]aic79xx [2]aic79xx with annwn:~# lsscsi [2:0:0:0]diskSEAGATE ST336807LW 0C01 /dev/sdb [2:0:1:0]diskSEAGATE ST336807LW 0C01 /dev/sdc [2:0:2:0]diskSEAGATE ST336807LW 0C01 /dev/sdd [2:0:3:0]diskSEAGATE ST336807LW 0C01 /dev/sde and giving up to 200Mb/s on a pretty old mobo (see cached reads) with raid5: annwn:~# hdparm -tT /dev/sd[bcde] /dev/sdb: Timing cached reads: 950 MB in 2.00 seconds = 474.23 MB/sec Timing buffered disk reads: 228 MB in 3.00 seconds = 75.93 MB/sec /dev/sdc: Timing cached reads: 1020 MB in 2.00 seconds = 510.01 MB/sec Timing buffered disk reads: 228 MB in 3.01 seconds = 75.76 MB/sec /dev/sdd: Timing cached reads: 1022 MB in 2.00 seconds = 510.69 MB/sec Timing buffered disk reads: 228 MB in 3.01 seconds = 75.79 MB/sec /dev/sde: Timing cached reads: 1016 MB in 2.00 seconds = 507.92 MB/sec Timing buffered disk reads: 228 MB in 3.02 seconds = 75.48 MB/sec annwn:~# hdparm -tT /dev/md0 /dev/md0: Timing cached reads: 1020 MB in 2.00 seconds = 509.96 MB/sec Timing buffered disk reads: 602 MB in 3.01 seconds = 199.84 MB/sec -- J.A. Magallon \ Software is like sex: \ It's better when it's free Mandriva Linux release 2008.1 (Cooker) for i586 Linux 2.6.23-jam01 (gcc 4.2.2 20070909 (4.2.2-0.RC.1mdv2008.0)) SMP PREEMPT 09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0 - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
Hi Andrew, James, On Monday 15 October 2007 14:53, Gabriel C wrote: > >> Compile tested and applies cleanly to 2.6.23. > >> I don't have this hardware anymore and cannot run test these patches. > > > > I can test these patches on an aic7892 controller later on today if you > > want. > > Works fine for me tested on : > > 03:0e.0 SCSI storage controller [0100]: Adaptec AIC-7892P U160/m [9005:008f] > (rev 02) I hope this is enough for these patches to be accepted. If it is not, please let me know what do I need to do more. -- vda - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
Denys Vlasenko wrote: > On Sunday 14 October 2007 18:47, Gabriel C wrote: >>> Compile tested and applies cleanly to 2.6.23. >>> I don't have this hardware anymore and cannot run test these patches. >> I can test these patches on an aic7892 controller later on today if you want. > > I'd appreciate that. > > Do you, by any chance, use aic94xx driver too (drivers/scsi/aic94xx/*)? > After i'm done with aic7xxx, I may attack this one. Sorry I don't have any box uses this one , but I'll ask some friends maybe someone has :) > >> BTW while you seems to care about this driver could you have a look at : >> >> http://bugzilla.kernel.org/show_bug.cgi?id=3062 ?!? > > I am a desktop Linux user, so far I don't use suspend at all. Sorry. Ok , no problem maybe one 'day' the SCSI folks cares ( where the 'day' may be never ), I dunno :( > -- > vda > Gabriel - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
>> Compile tested and applies cleanly to 2.6.23. >> I don't have this hardware anymore and cannot run test these patches. > > I can test these patches on an aic7892 controller later on today if you want. Works fine for me tested on : 03:0e.0 SCSI storage controller [0100]: Adaptec AIC-7892P U160/m [9005:008f] (rev 02) Gabriel - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
On Sunday 14 October 2007 18:47, Gabriel C wrote: > > Compile tested and applies cleanly to 2.6.23. > > I don't have this hardware anymore and cannot run test these patches. > > I can test these patches on an aic7892 controller later on today if you want. I'd appreciate that. Do you, by any chance, use aic94xx driver too (drivers/scsi/aic94xx/*)? After i'm done with aic7xxx, I may attack this one. > BTW while you seems to care about this driver could you have a look at : > > http://bugzilla.kernel.org/show_bug.cgi?id=3062 ?!? I am a desktop Linux user, so far I don't use suspend at all. Sorry. -- vda - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
Denys Vlasenko wrote: > Hi, Hi, > Compile tested and applies cleanly to 2.6.23. > I don't have this hardware anymore and cannot run test these patches. I can test these patches on an aic7892 controller later on today if you want. BTW while you seems to care about this driver could you have a look at : http://bugzilla.kernel.org/show_bug.cgi?id=3062 ?!? Regards, Gabriel C - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
[PATCH 0/3] debloat aic7xxx and aic79xx drivers
Hi, Following patches debloat drivers/scsi/aic7xxx/*. I also had to add prototypes for ahc_lookup_scb and ahd_lookup_scb to .h files. 1-debloat.patch Deinlines and moves big functions from .h to .c files. Adds prototypes for ahc_lookup_scb and ahd_lookup_scb to .h files. 2-addstatic.patch Adds statics, #ifdefs out huge amount of unused code, adds consts 3-addconst.patch Adds more consts Driver code/data size reductions: Build with debugging on (CONFIG_AIC7XXX_DEBUG_ENABLE=y): textdata bss dec hex filename 310865 499221204 361991 58607 linux-2.6.23.t/drivers/scsi/aic7xxx/built-in.o 22198727541204 225945 37299 linux-2.6.23-aic-3-addconst.t/drivers/scsi/aic7xxx/built-in.o With debugging off: textdata bss dec hex filename 298896 427541172 342822 53b26 linux-2.6.23.tt/drivers/scsi/aic7xxx/built-in.o 21606827541172 219994 35b5a linux-2.6.23-aic-3-addconst.tt/drivers/scsi/aic7xxx/built-in.o make namespacecheck goes from 400+ functions to: drivers/scsi/aic7xxx/aic79xx_core.o ahd_inq ahd_inw ahd_outq ahd_outw drivers/scsi/aic7xxx/aic79xx_osm.o ahd_insb drivers/scsi/aic7xxx/aic7xxx_core.o ahc_inq ahc_outq drivers/scsi/aic7xxx/aic7xxx_osm.o ahc_insb None of these patches touch any logic, code changes are pretty minimal. Compile tested and applies cleanly to 2.6.23. I don't have this hardware anymore and cannot run test these patches. Please apply. Signed-off-by: Denys Vlasenko <[EMAIL PROTECTED]> -- vda - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
On Friday 31 August 2007 17:27, [EMAIL PROTECTED] wrote: > On Fri, 31 Aug 2007 16:13:59 BST, Denys Vlasenko said: > > > >textdata bss dec hex filename > > 261433 500181172 312623 4c52f > > linux-2.6.23-rc1.org.t/drivers/scsi/aic7xxx/built-in.o > > 199654 500181172 250844 3d3dc > > linux-2.6.23-rc1.aic.t/drivers/scsi/aic7xxx/built-in.o > > 184014 213141172 206500 326a4 > > linux-2.6.23-rc1.aic1.t/drivers/scsi/aic7xxx/built-in.o > > 20237828501172 206400 32640 > > linux-2.6.23-rc1.aic2.t/drivers/scsi/aic7xxx/built-in.o > > > > 1-debloat.patchdeinlines a lot of functions > > 2-addstatic.patch adds statics, #ifdefs out huge amount of unused code, > > adds consts > > 3-addconst.patch adds more consts > > Yowza. Looking at aic1->aic2, it looks like 20K became 'const', and only > 3K *wasn't* 'const'? Exactly. There are firmware images/patches and a lot of structures with char* members pointing to text/messages. Btw, aic94xx driver needs the same treatment. -- vda - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH 0/3] debloat aic7xxx and aic79xx drivers
On Fri, 31 Aug 2007 16:13:59 BST, Denys Vlasenko said: > >textdata bss dec hex filename > 261433 500181172 312623 4c52f > linux-2.6.23-rc1.org.t/drivers/scsi/aic7xxx/built-in.o > 199654 500181172 250844 3d3dc > linux-2.6.23-rc1.aic.t/drivers/scsi/aic7xxx/built-in.o > 184014 213141172 206500 326a4 > linux-2.6.23-rc1.aic1.t/drivers/scsi/aic7xxx/built-in.o > 20237828501172 206400 32640 > linux-2.6.23-rc1.aic2.t/drivers/scsi/aic7xxx/built-in.o > > 1-debloat.patchdeinlines a lot of functions > 2-addstatic.patch adds statics, #ifdefs out huge amount of unused code, adds > consts > 3-addconst.patch adds more consts Yowza. Looking at aic1->aic2, it looks like 20K became 'const', and only 3K *wasn't* 'const'? pgpaOhlvqBmTy.pgp Description: PGP signature
[PATCH 0/3] debloat aic7xxx and aic79xx drivers
On Friday 31 August 2007 14:42, Hannes Reinecke wrote: > Jan Engelhardt wrote: > > On Aug 30 2007 13:02, Matthew Wilcox wrote: > >>> Well, you can send it to Linus/Andrew, that will usually upset people > >>> and they start commenting on it. Or they don't, and everything is fine. > >>> (The "default y" approach so to speak ;-) > >> > >> The problem is that we don't really have a maintainer for the aic7xyz > >> drivers any more. Volunteers welcome. NOT IT! > > > > Take it before someone else does! > > Well, the semi-official maintainers are James B. and me. > > So I might as well do it officially. Cool. Thanks to Arjan's insistence, I also took a look at adding statics and unexpectedly discovered yet another 50 kbytes of dead code (I'm not kidding). Attached are three patches which fix that: textdata bss dec hex filename 261433 500181172 312623 4c52f linux-2.6.23-rc1.org.t/drivers/scsi/aic7xxx/built-in.o 199654 500181172 250844 3d3dc linux-2.6.23-rc1.aic.t/drivers/scsi/aic7xxx/built-in.o 184014 213141172 206500 326a4 linux-2.6.23-rc1.aic1.t/drivers/scsi/aic7xxx/built-in.o 20237828501172 206400 32640 linux-2.6.23-rc1.aic2.t/drivers/scsi/aic7xxx/built-in.o 1-debloat.patchdeinlines a lot of functions 2-addstatic.patch adds statics, #ifdefs out huge amount of unused code, adds consts 3-addconst.patch adds more consts make namespacecheck goes from 400+ functions to: drivers/scsi/aic7xxx/aic79xx_core.o ahd_inq ahd_inw ahd_outq ahd_outw drivers/scsi/aic7xxx/aic79xx_osm.o ahd_insb drivers/scsi/aic7xxx/aic7xxx_core.o ahc_inq ahc_outq drivers/scsi/aic7xxx/aic7xxx_osm.o ahc_insb None of these patches actually touch any logic, code changes are pretty minimal. Compile tested and applies cleanly to 2.6.23-rc1, applies with some fuzz to 2.6.23-rc3. Please propagate to mainline. -- vda - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
