Here is a possible opportunity to reduce data usage in the kernel. $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ grep -v __initdata | \ wc -l 3250 Meaning there are ~3000 declarations of arrays with what appears to be file static const content that are not marked const. So there are many static arrays that could be marked const to move the compiled object code from data to text minimizing the total amount of exposed r/w data. However, I do not know of a mechanism using coccinelle to determine whether or not any of these static declarations are ever modified. So it appears that each instance of these declarations might need manual inspection. But for arrays declared inside functions, it's much more likely that the static declaration without const is done with the intent to modify the array: (note the difference in the git grep with a leading '^\s+') $ git grep -Pn '^\s+static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ grep -v __initdata | \ wc -l 323 ------------- For instance: (head -10 of the git grep for file statics) drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; drivers/accessibility/speakup/keyhelp.c:26:static u_char funcvals[] = { drivers/accessibility/speakup/main.c:2059:static spkup_hand spkup_handler[] = { drivers/accessibility/speakup/speakup_acntpc.c:35:static unsigned int synth_portlist[] = { 0x2a8, 0 }; drivers/accessibility/speakup/speakup_decpc.c:133:static int synth_portlist[] = { 0x340, 0x350, 0x240, 0x250, 0 }; drivers/accessibility/speakup/speakup_dectlk.c:110:static int ap_defaults[] = {122, 89, 155, 110, 208, 240, 200, 106, 306}; drivers/accessibility/speakup/speakup_dectlk.c:111:static int g5_defaults[] = {86, 81, 86, 84, 81, 80, 83, 83, 73}; drivers/accessibility/speakup/speakup_dtlk.c:34:static unsigned int synth_portlist[] = { drivers/accessibility/speakup/speakup_keypc.c:34:static unsigned int synth_portlist[] = { 0x2a8, 0 }; drivers/acpi/ac.c:137:static enum power_supply_property ac_props[] = { For drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; masks is only used in static function say_key and should be const and perhaps the declaration might be better moved into that function. For drivers/accessibility/speakup/keyhelp.c:26:static u_char funcvals[] = { funcvals is only used in static function spk_handle_help and should be const and perhaps the declaration might be better moved into that function. For drivers/accessibility/speakup/main.c:2059:static spkup_hand spkup_handler[] = { spkup_handler is only used in static function do_spkup and should be const and perhaps the declaration might be better moved into that function. etc... for speakup For drivers/acpi/ac.c:137:static enum power_supply_property ac_props[] = { array ac_props is assigned as a reference in acpi_ac_add as a "const enum power_supply_property *" member of a struct power_supply_desc. ------------- For instance: (head -10 of the git grep for function statics) drivers/acpi/apei/apei-base.c:781: static u8 whea_uuid_str[] = "ed855e0c-6c90-47bf-a62a-26de0fc5ad5c"; drivers/block/amiflop.c:1051: static unsigned char CRCTable1[] = { drivers/block/amiflop.c:1070: static unsigned char CRCTable2[] = { drivers/block/drbd/drbd_nl.c:872: static char units[] = { 'K', 'M', 'G', 'T', 'P', 'E' }; drivers/block/drbd/drbd_proc.c:224: static char write_ordering_chars[] = { drivers/block/drbd/drbd_receiver.c:4363: static enum drbd_conns c_tab[] = { drivers/char/pcmcia/synclink_cs.c:3717: static unsigned char patterns[] = drivers/cpufreq/intel_pstate.c:1515: static int silvermont_freq_table[] = { drivers/cpufreq/intel_pstate.c:1530: static int airmont_freq_table[] = { drivers/dma/xgene-dma.c:360: static u8 flyby_type[] = { Some of these could be const, but some could not. For instance: For drivers/acpi/apei/apei-base.c:781: static u8 whea_uuid_str[] = "ed855e0c-6c90-47bf-a62a-26de0fc5ad5c"; whea_uuid_str is assigned as a reference in "int apei_osc_setup(void)" a struct acpi_osc_context where .uuid_str is not declared as const char *. _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On Tue, 2 Mar 2021, Joe Perches wrote: > Here is a possible opportunity to reduce data usage in the kernel. Does it actually reduce data usage? julia > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > grep -v __initdata | \ > wc -l > 3250 > > Meaning there are ~3000 declarations of arrays with what appears to be > file static const content that are not marked const. > > So there are many static arrays that could be marked const to move the > compiled object code from data to text minimizing the total amount of > exposed r/w data. > > However, I do not know of a mechanism using coccinelle to determine > whether or not any of these static declarations are ever modified. > > So it appears that each instance of these declarations might need > manual inspection. > > But for arrays declared inside functions, it's much more likely that > the static declaration without const is done with the intent to modify > the array: > > (note the difference in the git grep with a leading '^\s+') > > $ git grep -Pn '^\s+static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > grep -v __initdata | \ > wc -l > 323 > > ------------- For instance: (head -10 of the git grep for file statics) > > drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; > drivers/accessibility/speakup/keyhelp.c:26:static u_char funcvals[] = { > drivers/accessibility/speakup/main.c:2059:static spkup_hand spkup_handler[] = { > drivers/accessibility/speakup/speakup_acntpc.c:35:static unsigned int synth_portlist[] = { 0x2a8, 0 }; > drivers/accessibility/speakup/speakup_decpc.c:133:static int synth_portlist[] = { 0x340, 0x350, 0x240, 0x250, 0 }; > drivers/accessibility/speakup/speakup_dectlk.c:110:static int ap_defaults[] = {122, 89, 155, 110, 208, 240, 200, 106, 306}; > drivers/accessibility/speakup/speakup_dectlk.c:111:static int g5_defaults[] = {86, 81, 86, 84, 81, 80, 83, 83, 73}; > drivers/accessibility/speakup/speakup_dtlk.c:34:static unsigned int synth_portlist[] = { > drivers/accessibility/speakup/speakup_keypc.c:34:static unsigned int synth_portlist[] = { 0x2a8, 0 }; > drivers/acpi/ac.c:137:static enum power_supply_property ac_props[] = { > > For drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; > > masks is only used in static function say_key and should be const and > perhaps the declaration might be better moved into that function. > > For drivers/accessibility/speakup/keyhelp.c:26:static u_char funcvals[] = { > > funcvals is only used in static function spk_handle_help and should be const > and perhaps the declaration might be better moved into that function. > > For drivers/accessibility/speakup/main.c:2059:static spkup_hand spkup_handler[] = { > > spkup_handler is only used in static function do_spkup and should be const > and perhaps the declaration might be better moved into that function. > > etc... for speakup > > For drivers/acpi/ac.c:137:static enum power_supply_property ac_props[] = { > > array ac_props is assigned as a reference in acpi_ac_add as a > "const enum power_supply_property *" member of a struct power_supply_desc. > > ------------- For instance: (head -10 of the git grep for function statics) > > drivers/acpi/apei/apei-base.c:781: static u8 whea_uuid_str[] = "ed855e0c-6c90-47bf-a62a-26de0fc5ad5c"; > drivers/block/amiflop.c:1051: static unsigned char CRCTable1[] = { > drivers/block/amiflop.c:1070: static unsigned char CRCTable2[] = { > drivers/block/drbd/drbd_nl.c:872: static char units[] = { 'K', 'M', 'G', 'T', 'P', 'E' }; > drivers/block/drbd/drbd_proc.c:224: static char write_ordering_chars[] = { > drivers/block/drbd/drbd_receiver.c:4363: static enum drbd_conns c_tab[] = { > drivers/char/pcmcia/synclink_cs.c:3717: static unsigned char patterns[] = > drivers/cpufreq/intel_pstate.c:1515: static int silvermont_freq_table[] = { > drivers/cpufreq/intel_pstate.c:1530: static int airmont_freq_table[] = { > drivers/dma/xgene-dma.c:360: static u8 flyby_type[] = { > > Some of these could be const, but some could not. For instance: > > For drivers/acpi/apei/apei-base.c:781: static u8 whea_uuid_str[] = "ed855e0c-6c90-47bf-a62a-26de0fc5ad5c"; > > whea_uuid_str is assigned as a reference in "int apei_osc_setup(void)" > a struct acpi_osc_context where .uuid_str is not declared as const char *. > > > > > _______________________________________________ > Cocci mailing list > Cocci@systeme.lip6.fr > https://systeme.lip6.fr/mailman/listinfo/cocci > _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Hi all! On 02/03/2021 18:42, Joe Perches wrote: [...] > ------------- For instance: (head -10 of the git grep for file statics) > > drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; > drivers/accessibility/speakup/keyhelp.c:26:static u_char funcvals[] = { > drivers/accessibility/speakup/main.c:2059:static spkup_hand spkup_handler[] = { > drivers/accessibility/speakup/speakup_acntpc.c:35:static unsigned int synth_portlist[] = { 0x2a8, 0 }; > drivers/accessibility/speakup/speakup_decpc.c:133:static int synth_portlist[] = { 0x340, 0x350, 0x240, 0x250, 0 }; > drivers/accessibility/speakup/speakup_dectlk.c:110:static int ap_defaults[] = {122, 89, 155, 110, 208, 240, 200, 106, 306}; > drivers/accessibility/speakup/speakup_dectlk.c:111:static int g5_defaults[] = {86, 81, 86, 84, 81, 80, 83, 83, 73}; > drivers/accessibility/speakup/speakup_dtlk.c:34:static unsigned int synth_portlist[] = { > drivers/accessibility/speakup/speakup_keypc.c:34:static unsigned int synth_portlist[] = { 0x2a8, 0 }; > drivers/acpi/ac.c:137:static enum power_supply_property ac_props[] = { > > For drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; Looking at the examples: Just s/^static /static const / in the lines reported by the grep's above and see if it compiles (and save space)? MfG, Bernd -- Bernd Petrovitsch Email : bernd@petrovitsch.priv.at There is NO CLOUD, just other people's computers. - FSFE LUGA : http://www.luga.at _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On Tue, 2021-03-02 at 22:41 +0100, Julia Lawall wrote: > > On Tue, 2 Mar 2021, Joe Perches wrote: > > > Here is a possible opportunity to reduce data usage in the kernel. > > Does it actually reduce data usage? Yes, at least for gcc. For instance: $ gcc --version gcc (Ubuntu 10.2.0-13ubuntu1) 10.2.0 And with the diff below (x86-64 defconfig with hwmon/pmbus and max20730) $ diff --git a/drivers/hwmon/pmbus/max20730.c b/drivers/hwmon/pmbus/max20730.c index 9dd3dd79bc18..b824eab95456 100644 --- a/drivers/hwmon/pmbus/max20730.c +++ b/drivers/hwmon/pmbus/max20730.c @@ -434,7 +434,7 @@ static long direct_to_val(u16 w, enum pmbus_sensor_classes > return d; } -static u32 max_current[][5] = { +static const u32 max_current[][5] = { [max20710] = { 6200, 8000, 9700, 11600 }, [max20730] = { 13000, 16600, 20100, 23600 }, [max20734] = { 21000, 27000, 32000, 38000 }, $ size drivers/hwmon/pmbus/max20730.o* text data bss dec hex filename 9245 256 0 9501 251d drivers/hwmon/pmbus/max20730.o.gcc.new 9149 352 0 9501 251d drivers/hwmon/pmbus/max20730.o.gcc.old with clang-11 it doesn't seem to make a difference: $ /usr/bin/clang --version Ubuntu clang version 11.0.0-2 $ size drivers/hwmon/pmbus/max20730.o* text data bss dec hex filename 9166 256 1 9423 24cf drivers/hwmon/pmbus/max20730.o.clang-11.new 9166 256 1 9423 24cf drivers/hwmon/pmbus/max20730.o.clang-11.old _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On Tue, 2 Mar 2021, Bernd Petrovitsch wrote: > Hi all! > > On 02/03/2021 18:42, Joe Perches wrote: > [...] > > ------------- For instance: (head -10 of the git grep for file statics) > > > > drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; > > drivers/accessibility/speakup/keyhelp.c:26:static u_char funcvals[] = { > > drivers/accessibility/speakup/main.c:2059:static spkup_hand spkup_handler[] = { > > drivers/accessibility/speakup/speakup_acntpc.c:35:static unsigned int synth_portlist[] = { 0x2a8, 0 }; > > drivers/accessibility/speakup/speakup_decpc.c:133:static int synth_portlist[] = { 0x340, 0x350, 0x240, 0x250, 0 }; > > drivers/accessibility/speakup/speakup_dectlk.c:110:static int ap_defaults[] = {122, 89, 155, 110, 208, 240, 200, 106, 306}; > > drivers/accessibility/speakup/speakup_dectlk.c:111:static int g5_defaults[] = {86, 81, 86, 84, 81, 80, 83, 83, 73}; > > drivers/accessibility/speakup/speakup_dtlk.c:34:static unsigned int synth_portlist[] = { > > drivers/accessibility/speakup/speakup_keypc.c:34:static unsigned int synth_portlist[] = { 0x2a8, 0 }; > > drivers/acpi/ac.c:137:static enum power_supply_property ac_props[] = { > > > > For drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { 32, 16, 8, 4, 2, 1 }; > > Looking at the examples: Just s/^static /static const / in the lines > reported by the grep's above and see if it compiles (and save space)? There is a small risk with compiling that there may be a problem for a configuration you haven't considered. julia _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On 02/03/2021 18.42, Joe Perches wrote: > Here is a possible opportunity to reduce data usage in the kernel. > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > grep -v __initdata | \ > wc -l > 3250 > > Meaning there are ~3000 declarations of arrays with what appears to be > file static const content that are not marked const. > > So there are many static arrays that could be marked const to move the > compiled object code from data to text minimizing the total amount of > exposed r/w data. You can add const if you like, but it will rarely change the generated code. gcc is already smart enough to take a static array whose contents are provably never modified within the TU and put it in .rodata: static int x = 7; static int y[2] = {13, 19}; static int p(int a, const int *foo) { return a + *foo; } int q(int a) { int b = p(a, &x); return p(b, &y[b & 1]); } $ nm c.o 0000000000000000 T q 0000000000000000 r y $ size c.o text data bss dec hex filename 111 0 0 111 6f c.o So x gets optimized away completely, but y isn't so easy to get rid of - nevertheless, it's never modified and the address doesn't escape the TU, so gcc treats is as if it was declared const. I think you'll see the same if you try adding the const on a few of your real-life examples. (Of course, the control flow may be so convoluted that gcc isn't able to infer the constness, so I'm not saying it will never make a difference - only that you shouldn't expect too much.) Rasmus _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On Wed, 2021-03-03 at 10:41 +0100, Rasmus Villemoes wrote: > On 02/03/2021 18.42, Joe Perches wrote: > > Here is a possible opportunity to reduce data usage in the kernel. > > > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > > grep -v __initdata | \ > > wc -l > > 3250 > > > > Meaning there are ~3000 declarations of arrays with what appears to be > > file static const content that are not marked const. > > > > So there are many static arrays that could be marked const to move the > > compiled object code from data to text minimizing the total amount of > > exposed r/w data. > > You can add const if you like, but it will rarely change the generated > code. gcc is already smart enough to take a static array whose contents > are provably never modified within the TU and put it in .rodata: At least some or perhaps even most of the time, true, but the gcc compiler from v5 through at least v10 seems inconsistent about when it does the appropriate conversion. See the example I posted: https://lore.kernel.org/lkml/6b8b250a06a98ce42120a14824531a8641f5e8aa.camel@perches.com/ It was a randomly chosen source file conversion btw, I had no prior knowledge of whether the text/data use would change. I'm unsure about clang consistently moving static but provably const arrays from data to text. I rarely use clang. At least for v11 it seems to be better though. I didn't try 10.1. _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On Tue, Mar 2, 2021 at 4:22 PM Joe Perches <joe@perches.com> wrote: > > Here is a possible opportunity to reduce data usage in the kernel. > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > grep -v __initdata | \ > wc -l > 3250 > > Meaning there are ~3000 declarations of arrays with what appears to be > file static const content that are not marked const. > > So there are many static arrays that could be marked const to move the > compiled object code from data to text minimizing the total amount of > exposed r/w data. > > However, I do not know of a mechanism using coccinelle to determine > whether or not any of these static declarations are ever modified. I thought it would be a fun exercise but it got tedious quick. I don't know how to ignore an attribute like __initdata. Feel free to refine it: @@ type t; identifier x; @@ ( static const struct { ... } x[]; | static + const struct { ... } x[]; | static const struct s *x[]; | static + const struct s *x[]; | static const struct s x[]; | static + const struct s x[]; | static const t *x[]; | static + const t *x[]; | static const t x[]; | static + const t x[]; ) @@ type t; identifier s, x, y, z; assignment operator xx; @@ ( static const struct { ... } x[] = { ... }; | static + const struct { ... } x[] = { ... }; | static const struct s *x[] = { ... }; | static + const struct s *x[] = { ... }; | static const struct s x[] = { ... }; | static + const struct s x[] = { ... }; | static const t *x[] = { ... }; | static + const t *x[] = { ... }; | static const t x[] = { ... }; | static + const t x[] = { ... }; ) ... when != x.y xx ... when != x[...] xx ... when != z = x _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On Wed, 3 Mar 2021, Mansour Moufid wrote: > On Tue, Mar 2, 2021 at 4:22 PM Joe Perches <joe@perches.com> wrote: > > > > Here is a possible opportunity to reduce data usage in the kernel. > > > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > > grep -v __initdata | \ > > wc -l > > 3250 > > > > Meaning there are ~3000 declarations of arrays with what appears to be > > file static const content that are not marked const. > > > > So there are many static arrays that could be marked const to move the > > compiled object code from data to text minimizing the total amount of > > exposed r/w data. > > > > However, I do not know of a mechanism using coccinelle to determine > > whether or not any of these static declarations are ever modified. > > I thought it would be a fun exercise but it got tedious quick. > > I don't know how to ignore an attribute like __initdata. > > Feel free to refine it: This adds the consts, but it doesn't check that the array is never updated, or never stored in a field from which it could get updated. In my attempt, I tried something like this for the first part: @r disable optional_qualifier@ type T; identifier x; position p != ok.p; @@ static T x@p[] = { ... }; In principle, this should match cases where there is no const. But it dones't work. It matches some cases with const too. Then I tried: @ok@ type T; identifier x; position p; @@ static const T x@p[] = { ... }; @r@ type T; identifier x; position p != ok.p; @@ static T x@p[] = { ... }; The first rule matches the cases with const, and then the second rule matches all of the cases with the declared variable at a position different than that of the first rule. It works if the type T is a simple type like int, but it doesn't work if it is something more complex, like struct foo *. Afterwards, I have the following: @bad@ position p; identifier r.x; expression e,y; @@ ( x@p[y] = e | &x@p[y] ) @good@ identifier r.x; expression y; position p != bad.p; @@ x@p[y] @notgood@ identifier r.x; position p != good.p; @@ x@p @depends on good && !notgood@ identifier r.x; type r.T; @@ static +const T x[] = { ... }; That is, * rule bad: Give up if we assign an element or take the address of an element. * rule good: Things are going well if we access an element of the array. If there is no access at all, there is something suspicious, perhaps uses in macros. * rule notgood: A montion of the array that is not an element access is not good either In the end, if we match good and we don't match notgood, then we can make the change. I got stuck on the const problem, and didn't try the __initdata issue. But one could address with a rule like the rule ok above. The const problem is at least something worth looking into. julia > > > @@ > type t; > identifier x; > @@ > ( > static const struct { ... } x[]; > | > static > + const > struct { ... } x[]; > | > static const struct s *x[]; > | > static > + const > struct s *x[]; > | > static const struct s x[]; > | > static > + const > struct s x[]; > | > static const t *x[]; > | > static > + const > t *x[]; > | > static const t x[]; > | > static > + const > t x[]; > ) > > @@ > type t; > identifier s, x, y, z; > assignment operator xx; > @@ > ( > static const struct { ... } x[] = { ... }; > | > static > + const > struct { ... } x[] = { ... }; > | > static const struct s *x[] = { ... }; > | > static > + const > struct s *x[] = { ... }; > | > static const struct s x[] = { ... }; > | > static > + const > struct s x[] = { ... }; > | > static const t *x[] = { ... }; > | > static > + const > t *x[] = { ... }; > | > static const t x[] = { ... }; > | > static > + const > t x[] = { ... }; > ) > ... when != x.y xx ... > when != x[...] xx ... > when != z = x > _______________________________________________ > Cocci mailing list > Cocci@systeme.lip6.fr > https://systeme.lip6.fr/mailman/listinfo/cocci > _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
[-- Attachment #1.1: Type: text/plain, Size: 2179 bytes --] On Wednesday, March 3, 2021, Bernd Petrovitsch <bernd@petrovitsch.priv.at> wrote: > Hi all! > > On 02/03/2021 18:42, Joe Perches wrote: > [...] > > ------------- For instance: (head -10 of the git grep for file statics) > > > > drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = { > 32, 16, 8, 4, 2, 1 }; > > drivers/accessibility/speakup/keyhelp.c:26:static u_char funcvals[] = { > > drivers/accessibility/speakup/main.c:2059:static spkup_hand > spkup_handler[] = { > > drivers/accessibility/speakup/speakup_acntpc.c:35:static unsigned int > synth_portlist[] = { 0x2a8, 0 }; > > drivers/accessibility/speakup/speakup_decpc.c:133:static int > synth_portlist[] = { 0x340, 0x350, 0x240, 0x250, 0 }; > > drivers/accessibility/speakup/speakup_dectlk.c:110:static int > ap_defaults[] = {122, 89, 155, 110, 208, 240, 200, 106, 306}; > > drivers/accessibility/speakup/speakup_dectlk.c:111:static int > g5_defaults[] = {86, 81, 86, 84, 81, 80, 83, 83, 73}; > > drivers/accessibility/speakup/speakup_dtlk.c:34:static unsigned int > synth_portlist[] = { > > drivers/accessibility/speakup/speakup_keypc.c:34:static unsigned int > synth_portlist[] = { 0x2a8, 0 }; > > drivers/acpi/ac.c:137:static enum power_supply_property ac_props[] = { > > > > For drivers/accessibility/speakup/keyhelp.c:18:static u_short masks[] = > { 32, 16, 8, 4, 2, 1 }; > > Looking at the examples: Just s/^static /static const / in the lines > reported by the grep's above and see if it compiles (and save space)? I did two reverts and reported at least one issue with blind constification. Besides that we have a lot of data structures that require to drop const spécifier and the consumer won’t actually know if it’s possible to write there or not. I’m talking about driver data fields where they are defined as type of kernel_ulong_t. So, first you need to fix that, > > MfG, > Bernd > -- > Bernd Petrovitsch Email : bernd@petrovitsch.priv.at > There is NO CLOUD, just other people's computers. - FSFE > LUGA : http://www.luga.at > -- With Best Regards, Andy Shevchenko [-- Attachment #1.2: Type: text/html, Size: 2791 bytes --] [-- Attachment #2: Type: text/plain, Size: 136 bytes --] _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
[-- Attachment #1: Type: text/plain, Size: 2086 bytes --] On Wed, 3 Mar 2021, Joe Perches wrote: > On Wed, 2021-03-03 at 10:41 +0100, Rasmus Villemoes wrote: > > On 02/03/2021 18.42, Joe Perches wrote: > > > Here is a possible opportunity to reduce data usage in the kernel. > > > > > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > > > grep -v __initdata | \ > > > wc -l > > > 3250 > > > > > > Meaning there are ~3000 declarations of arrays with what appears to be > > > file static const content that are not marked const. > > > > > > So there are many static arrays that could be marked const to move the > > > compiled object code from data to text minimizing the total amount of > > > exposed r/w data. > > > > You can add const if you like, but it will rarely change the generated > > code. gcc is already smart enough to take a static array whose contents > > are provably never modified within the TU and put it in .rodata: > > At least some or perhaps even most of the time, true, but the gcc compiler > from v5 through at least v10 seems inconsistent about when it does the > appropriate conversion. > > See the example I posted: > https://lore.kernel.org/lkml/6b8b250a06a98ce42120a14824531a8641f5e8aa.camel@perches.com/ > > It was a randomly chosen source file conversion btw, I had no prior > knowledge of whether the text/data use would change. > > I'm unsure about clang consistently moving static but provably const arrays > from data to text. I rarely use clang. At least for v11 it seems to be > better though. I didn't try 10.1. I tried the relevnt drivers in drivers/input/joystick. I got only one driver that changed with gcc 9.3, which was drivers/input/joystick/analog.c. It actually got larger: original: text data bss dec hex filename 22607 10560 320 33487 82cf drivers/input/joystick/analog.o after adding const: text data bss dec hex filename 22728 10816 320 33864 8448 drivers/input/joystick/analog.o This was the only case where bss was not 0, but I don't know if there is a connection. julia [-- Attachment #2: Type: text/plain, Size: 136 bytes --] _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
On Sun, 2021-03-07 at 20:14 +0100, Julia Lawall wrote: > > On Wed, 3 Mar 2021, Joe Perches wrote: > > > On Wed, 2021-03-03 at 10:41 +0100, Rasmus Villemoes wrote: > > > On 02/03/2021 18.42, Joe Perches wrote: > > > > Here is a possible opportunity to reduce data usage in the kernel. > > > > > > > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > > > > grep -v __initdata | \ > > > > wc -l > > > > 3250 > > > > > > > > Meaning there are ~3000 declarations of arrays with what appears to be > > > > file static const content that are not marked const. > > > > > > > > So there are many static arrays that could be marked const to move the > > > > compiled object code from data to text minimizing the total amount of > > > > exposed r/w data. > > > > > > You can add const if you like, but it will rarely change the generated > > > code. gcc is already smart enough to take a static array whose contents > > > are provably never modified within the TU and put it in .rodata: > > > > At least some or perhaps even most of the time, true, but the gcc compiler > > from v5 through at least v10 seems inconsistent about when it does the > > appropriate conversion. > > > > See the example I posted: > > https://lore.kernel.org/lkml/6b8b250a06a98ce42120a14824531a8641f5e8aa.camel@perches.com/ > > > > It was a randomly chosen source file conversion btw, I had no prior > > knowledge of whether the text/data use would change. > > > > I'm unsure about clang consistently moving static but provably const arrays > > from data to text. I rarely use clang. At least for v11 it seems to be > > better though. I didn't try 10.1. > > I tried the relevnt drivers in drivers/input/joystick. I got only one > driver that changed with gcc 9.3, which was > drivers/input/joystick/analog.c. It actually got larger: > > original: > > text data bss dec hex filename > 22607 10560 320 33487 82cf drivers/input/joystick/analog.o > > after adding const: > > text data bss dec hex filename > 22728 10816 320 33864 8448 drivers/input/joystick/analog.o > > This was the only case where bss was not 0, but I don't know if there is a > connection. You really need consider using defconfig so whatever object code does not have tracing/debugging support. For instance, this code with defconfig and analog joystick: Original: $ size drivers/input/joystick/analog.o text data bss dec hex filename 8115 261 224 8600 2198 drivers/input/joystick/analog.o with const: $ size drivers/input/joystick/analog.o text data bss dec hex filename 8179 201 224 8604 219c drivers/input/joystick/analog.o _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
[-- Attachment #1: Type: text/plain, Size: 3333 bytes --] On Sun, 7 Mar 2021, Joe Perches wrote: > On Sun, 2021-03-07 at 20:14 +0100, Julia Lawall wrote: > > > > On Wed, 3 Mar 2021, Joe Perches wrote: > > > > > On Wed, 2021-03-03 at 10:41 +0100, Rasmus Villemoes wrote: > > > > On 02/03/2021 18.42, Joe Perches wrote: > > > > > Here is a possible opportunity to reduce data usage in the kernel. > > > > > > > > > > $ git grep -P -n '^static\s+(?!const|struct)(?:\w+\s+){1,3}\w+\s*\[\s*\]' drivers/ | \ > > > > > grep -v __initdata | \ > > > > > wc -l > > > > > 3250 > > > > > > > > > > Meaning there are ~3000 declarations of arrays with what appears to be > > > > > file static const content that are not marked const. > > > > > > > > > > So there are many static arrays that could be marked const to move the > > > > > compiled object code from data to text minimizing the total amount of > > > > > exposed r/w data. > > > > > > > > You can add const if you like, but it will rarely change the generated > > > > code. gcc is already smart enough to take a static array whose contents > > > > are provably never modified within the TU and put it in .rodata: > > > > > > At least some or perhaps even most of the time, true, but the gcc compiler > > > from v5 through at least v10 seems inconsistent about when it does the > > > appropriate conversion. > > > > > > See the example I posted: > > > https://lore.kernel.org/lkml/6b8b250a06a98ce42120a14824531a8641f5e8aa.camel@perches.com/ > > > > > > It was a randomly chosen source file conversion btw, I had no prior > > > knowledge of whether the text/data use would change. > > > > > > I'm unsure about clang consistently moving static but provably const arrays > > > from data to text. I rarely use clang. At least for v11 it seems to be > > > better though. I didn't try 10.1. > > > > I tried the relevnt drivers in drivers/input/joystick. I got only one > > driver that changed with gcc 9.3, which was > > drivers/input/joystick/analog.c. It actually got larger: > > > > original: > > > > text data bss dec hex filename > > 22607 10560 320 33487 82cf drivers/input/joystick/analog.o > > > > after adding const: > > > > text data bss dec hex filename > > 22728 10816 320 33864 8448 drivers/input/joystick/analog.o > > > > This was the only case where bss was not 0, but I don't know if there is a > > connection. > > You really need consider using defconfig so whatever object code > does not have tracing/debugging support. > > For instance, this code with defconfig and analog joystick: > > Original: > > $ size drivers/input/joystick/analog.o > text data bss dec hex filename > 8115 261 224 8600 2198 drivers/input/joystick/analog.o > > with const: > > $ size drivers/input/joystick/analog.o > text data bss dec hex filename > 8179 201 224 8604 219c drivers/input/joystick/analog.o Thanks for the suggestion. It occurred to me that in one case my transformation was wrong, because the array was multi-level, and a sub array was being stored in a structure field that was not declared as const. So the argument that the compiler would figure out to put the array in .rodata didn't make sense. But I still go the same sizes for that file. So I'll try the whole thing again. thanks, julia [-- Attachment #2: Type: text/plain, Size: 136 bytes --] _______________________________________________ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci